Home

Consistent Climate Scenarios User Guide- Version 2.2

1. GCM name used Representative in Consistent Future Climate Climate PCMDI CMIP3 name Expert Review Panel recommendation partition RFC Scenarios Project CCCMA 63 CGCM3 1 T63 Likely to be less reliable ECHAMS5 ECHAMS5 MPI OM A ERANG Sas More likely to produce credible projections MIROC H MIROC3 2 hires MIROC M MIROC3 2 medres CNRM CNRM CM3 Consistently under performed Not assessed but expected to be better than HP CSIRO MK35 CSIRO MK3 5 Sedo ie p GFDL 20 GFDL CM2 0 HADCM3 UKMO HadCM3 More likely to produce credible projections HADGEM1 UKMO HadGEM1 CCCMA 47 CGCMS3 1 T47 WI GISS AOM GISS AOM Likely to be less reliable MRI GCM 232 MRI CGCM2 3 2 BCCR BCCR BCM2 0 Not recommended CSIRO MK30 CSIRO Mk3 0 Consistently under performed GFDL 21 GFDL CM2 1 More likely to produce credible projections WP IAP FGOALS G10 FGOALS g1 0 Likely to be less reliable INMCM INM CM3 0 Not recommended NCAR CCSM CCSM3 Likely to be less reliable 68 Consistent Climate Scenarios User Guide Version 2 2 UK Met Office Hadley Centre models added in June 2012 8 2 Composite HI HP WI and WP climate projections data It is well Known that individual GCMs are subject to individual model bias For this reason the CCS Expert Panel recommends either 1 running the 19 GCMs supplied or 2 if this is not practicable short cutting this processing by ordering composite data based on RFCs Composite da
2. Predicted from regression Figure 9 6 Regression models y axis vs huss screen height specific humidity x axis showing each GCM plotted against the prediction if each model is regressed against only its own predictands For example MIROC3 2 medres huss predicted from coefficients derived for that GCM only This measures the internal consistency of the regression and shows that for the models included in the calculation huss is highly predictable HUSS v Regressed HUSS y 0 7941x 1 16 Figure 9 7 The same as Figure 8 6 but showing the GCM huss x axis plotted against the predicted values y axis using the coefficients in Table 9 7 81 Department of Science Information Technology and Innovation 9 3 Estimating potential evaporation and pan evaporation As pan evaporation is not a variable directly available from GCMs pan evaporation data were initially calculated using OzClim Areal Wet Area potential evaporation trends based on Morton s method Morton 1983 This initial approach was taken since trends in Areal Wet Area potential evaporation should be highly correlated with trends in pan evaporation A refinement was made in CF Version 1 1 with OzClim trends for solar radiation and vapour pressure being adopted to calculate trends in pan evaporation CF Version 1 1 daily pan evaporation has been re computed using the same synthetic pan
3. klan Manne and Richard Richels 1997 On stabilizing C02 concentrations cost effective emission reduction strategies Environmental Modelling and Assessment 2 1997 NB All climate change factors are pattern scaled off the ALB scenario NB p subsscript preliminary and IS92 are for older scenarios Year A1B AT AFI s B1 B2 Alp Kp Bip B2p I1592a IS92a SAR C02 450 CO2 SSO RCP4 5 RCP6 O RCP3 PD RCPS S 1970 325 00 325 00 325 00 325 00 325 00 325 00 325 00 325 00 325 00 325 00 325 00 325 00 325 00 325 00 325 000 325 000 325 000 325 000 1980 337 00 337 00 337 00 337 00 337 00 337 00 337 00 337 00 337 00 337 00 337 00 337 00 337 00 337 00 337 000 337 000 337 000 337 000 1990 352 00 352 00 352 00 352 00 352 00 352 00 352 00 352 00 352 00 352 00 352 00 353 00 352 00 352 00 353 000 353 000 353 000 353 000 2000 367 00 367 00 367 00 367 00 367 00 367 00 367 00 367 00 367 00 367 00 367 00 370 00 371 00 368 00 368 865 363 865 366 665 368 865 2010 388 00 386 00 386 00 336 00 366 00 385 00 390 00 366 00 385 00 367 00 387 00 391 00 393 00 389 00 369 126 389 072 369 285 389 324 2020 418 00 410 00 415 00 414 00 410 00 406 00 421 00 416 00 407 00 412 00 413 00 416 00 414 00 414 00 411 129 409 360 412 0668 415 780 2030 447 00 435 00 449 00 444 00 432 00 425 00 454 00 447 00 425 00 433 00 439 00 444 00 430 00 440 00 435 046 428 876 430 763 448 835 2040 483 00 466 00 495 00 481 00 457 00 448 00 490 00 4684 00 445 00 457 00 468 00 475 00 439 00 467 00 460 645 450 696 440 222 48
4. Century climate 3 9 Quaniile trend plots 3 10 Histograms of quantile matched climate projections 3 11 Transient climate data test set for 1889 2100 4 Change factor GF methodology issis aaa aa 4 1 Change factor definition 4 2 Background 4 3 Calculation of change factors 4 4 A worked example projecting climate data for 2050 for a specific location 5 Quantile matching QM methodology sssssssssssussnunsnnnrnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnn mnnn 5 1 Steps involved to calculate QM projections data for 2030 5 2 Variation of methodology for calculating 2050 QM projections data 5 3 Post projection clamping 5 4 Transforms applied 6 Description of daily climate Variables ccssecssseeseeeessneeessneneseneessneeensnenenenesesneeessneneneneneseeers 6 1 SILO data 6 2 Patched Point and drilled data 7 Emissions scenarios and climate warming sensitivity seusssusssunnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnn nnna 7 1 Emissions scenarios background information 7 2 Selecting emissions scenarios 7 3 Climate warming sensitivity 20 20 23 24 28 30 32 34 36 39 41 43 45 45 45 46 49 50 50 52 53 54 55 55 56 57 57 60 60 Consistent Climate Scenarios User Guide Version 2 2 8 Global Climate Models aaa aaa aaa aaa aara aaa aaa 64 8 1 Selecting Global Climate Models 66 8 2 Composite HI HP WI and WP climate projections data 69 9 Infilling of
5. Morton F 1983 Operational estimates of areal evapotranspiration and their significance to the science and practice of hydrology Journal of Hydrology 66 1 76 NOAA 2010 Trends in Atmospheric Carbon Dioxide retrieved July 21 2010 from http www esrl noaa gov gmd ccgg trends global_ data NOAA 2011 National Weather Service Climate Prediction Centre Cold amp Warm Events by Season Retrieved July 2011 from http www cpc ncep noaa gov products analysis_monitoring ensostuff ensoyears shtml Page C M and Jones D 2001 OzClim The development of a climate scenario generator for Australia In A Zerger and R M Argent eds MODSIM 2001 International Congress on Modelling and Simulation Modelling and Simulation Society of Australia and New Zealand http www mssanz org au MODSIM01 Vol 202 Page pdf Panjkov A 2012 draft Consistent Climate Scenarios Project Quantile matching for climate projections Department of Science Information Technology Innovation and the Arts Queensland Government People amp Place 2009 Glossary of Climate Adaptation and Decision Making 2009 Retrieved May 21 2010 from http www peopleandplace net media_library text 2009 5 19 glossary_ of climate adaptation and decision making Program for Climate Model Diagnosis and Intercomparison 2011 About WCRP CMIP3 Model Output Retrieved December 3 2009 from http Awww pcmdi lIn gov ipcc info_for_analysts php Queensland
6. 1960 1970 1980 1990 2000 2010 2020 2030 Year Figure 3 12 An example from a quantile trend plot for a specific showing time series for the 0 1 0 5 and 0 9 quantiles and the statistical significance i e p values of daily minimum temperature for Augusts with trends computed over the 1957 2010 period location filename T Min_040428 10 50 90 _1957_2010_2030_QuantileTrends_040428 png Plotting scales may change between stations Notes e Where observed quantile trends are statistically significant p lt 0 1 the quantile matched approach is used to produce the 2030 rainfall temperature solar radiation and specific humidity projections data and the 2050 solar radiation and specific humidity data e Where observed quantile trends are not significant the climate projections method defaults to the change factor approach for the above mentioned climate variables 40 Consistent Climate Scenarios User Guide Version 2 2 3 10 Histograms of quantile matched climate projections If diagnostics are selected when ordering through the web portal the QM 2030 datasets will include plots of histograms of QM climate projections QM 2050 quantile trend plots can be downloaded via ftp climate mft derm qld qov au Climate Scenarios QM 2050 TestData These histograms arranged by month are location specific and show frequency distributions for specified climate variables or their transforms The frequency
7. 2 1x10 7 mm deg Global aes Figure 10 2 Illustrating limitations associated with downscaling from GCMs a An extract of pre OzClim model simulated base line climate showing the mean rainfall at one location b a consequential anomalously extreme downward trend in percentage change per degree of global warming and c Bureau of Meteorology observational data from the same region showing a 2 mm average rainfall for the month as compared to 1 nanometre shown by the GCM model observed monthly average 90 Consistent Climate Scenarios User Guide Version 2 2 Table 10 2 presents an example of several obvious anomalies found in the trend files supplied by CSIRO In all these cases they are change per degree of global warming expressed as a percentage of model base line climate and known as tpc files These are derived directly from OzClim being the intermediate trend files produced by interpolation from Global Climate Model GCM tpo files and cached in OzClim In OzClim the data output using the trend and degree of global warming are clamped to plausible values so anomalies are often not visible the issue of anomalies then arises when pattern of change data is published Table 10 2 Selected examples where rainfall in a few GCMs are affected by anomalous trends in OzClim Columns are GCM model name Climate variable Month and for all trend per degree of global warming tpc data for the month the mean standard d
8. 326 200 327 400 328 600 329 800 685 200 688 000 690 800 693 600 696 400 699 200 702 000 Lp 325 000 326 200 327 400 328 600 329 800 786 800 795 000 803 200 811 400 819 600 627 800 636 000 Bip 325 000 326 200 327 400 328 600 329 800 560 600 562 000 563 400 564 800 566 200 567 600 569 000 B2p 15928 325 000 325 000 326 200 326 200 327 400 327 400 328 600 328 600 329 800 329 800 600 000 673 000 603 000 678 000 606 000 683 000 609 000 686 000 612 000 693 000 615 000 698 000 618 000 703 000 Older sce T392a SAR 325 000 326 200 327 400 328 600 329 800 679 600 684 500 689 400 694 300 699 200 704 100 709 000 CO stabilisation scenarios narios 002 450 325 000 326 200 327 400 328 600 329 800 450 000 450 000 450 000 450 000 450 000 450 000 450 000 02 S50 325 000 326 200 327 400 328 600 329 800 546 400 547 000 547 600 548 200 548 800 549 400 550 000 RCP4 S 325 000 326 200 327 400 328 600 329 800 535 588 536 049 536 511 536 973 37 435 537 896 538 358 RCP6 0 325 000 326 200 327 400 328 600 329 800 649 279 652 686 656 093 659 501 662 908 666 316 669 723 RCP3 PD 325 000 326 200 327 400 328 600 329 800 423 961 423 450 422 939 422 428 421 917 421 406 420 895 RCPS S 325 000 326 200 327 400 328 600 329 000 6831 233 690 340 699 446 908 553 917 660 926 767 935 874 AR5 RCP scenarios
9. Bureau of Meteorology Class A pan evaporation site data has been used since 1970 Bird cages were installed then to reduce errors e g evaporation readings were occasionally too high due to animals drinking or splashing water The frequency spikes that occur in the observed pan evaporation frequency distribution plots are due largely to lack of precision in some post 1970 manually recorded observations For example during measurement evaporation has been rounded down to the nearest full can which holds 4 mm giving spikes at 4 mm and 8 mm on the x axis rather than providing more precise values Spikes in the observed data also occur on high rainfall days especially if the pan overflows or there is a reading error in the rainfall which is subsequently transferred to the pan reading 3 7 Comparison of model projections plots Comparison of model projections plots are provided in the ZIP archives as part of the user information framework to assist users in GCM model selection Each plot is location specific and shows projected changes i e climate change factors at 2030 from a 1960 2010 base period climate for both annual mean temperature and rainfall The changes in temperature and rainfall are presented for each of 19 GCMs and three climate warming sensitivities low median and high The GCMs were all forced by the A1B emissions scenario The projected changes have been calculated using amounts of global warming from MAG
10. In OzClim change factors are applied to a 1975 to 2004 baseline climate CSIRO 2010 4 2 Background The CMIP3 Coupled Model Intercomparison Project 3 database includes experiments from the 23 GCM models submitted to the International Panel for Climate Change IPCC Fourth Assessment Report AR4 IPCC AR4 represents a global scientific consensus on the issue of climate change and is accepted as the highest authority for policy makers The CMIP3 database has been freely available to the global research community resulting in unprecedented levels of evaluation and analysis OzClim CSIRO 2010 is a CSIRO product used to explore climate change scenarios from 2020 to 2100 for Australia and has an interactive website at http www csiro au ozclim OzClim provides climate change projections information based on the results of experiments from 23 GCM projections which are a subset from the International Panel for Climate Change Fourth Assessment Report IPCC AR4 For any one GCM the experiments in the CMIP3 database do not cover the full range of emissions scenarios eight which were considered by the IPCC in its 4 OzClim version 3 45 Department of Science Information Technology and Innovation Special Report on Emissions Scenarios SRES IPCC 2000 The experiments also did not include a desirable level of replication The IPCC employed a relatively straight forward methodology to extrapolate AR4 model outputs
11. May 2015 Australian Government Queensland Department of Agriculture Fisheries and Forestry en Consistent Climate Scenarios User Guide Version 2 2 About this document The intent of this User Guide is to provide users previously restricted to project teams under the CCRP program see Appendix but now extended to all users with background input and guidelines for using the Consistent Climate Scenarios CCS projections data The User Guide aims to assist users in interpreting the data that has been provided This updated User Guide V2 2 which supersedes the previous draft V2 1 released in August 2012 includes additional information about the availability of CCS datasets through the Long Paddock Climate Change Projections web portal updated CF and QM file naming conventions additional projections data for four Representative Future Climate partitions RFCs maps showing projected 21st Century temperature changes for the four RFCs This User Guide was supported by the project s Chair and Expert Panel In addition User feedback has also played an important role in the development and improvement of the information that this User Guide contains Referencing Consistent Climate Scenarios CCS Data The data source should be acknowledged as the Queensland Government SILO database http www longpaddock gid gov au silo The SILO database is operated by DSITI The climate change factors used to calculate CC
12. Select one or more global warming sensitivities High Medium Low 106 Consistent Climate Scenarios User Guide Version 2 2 Data Order Step 3 Select data parameters Climate models Models HI HP WI and WP are composites of the four mean climate response patterns below MIROCS 2 hires High global warming and a warm Indian Ocean High global warming and a warm Pacific Ocean Global Warming K CSIRO MK3 0 2 0 j Smaller global warming and a warm Pacific Ocean Smaller global warming and a warm Indian Ocean 1 6 20 West Pacific East Indian index Adapted from Watterson 2011 Data Order Step 4 Delivery details Provide a unique label for your data files CFtest Up to eight characters Select the type of operating platform for your file format Windows UNIX Select the output file format Should diagnostic charts be included in the output data file package No Yes Delivery method I wish to receive my order by Delivery by post Delivery or notification address 10 0 10 20 30 40 Select one or more Global Climate Models UKMO HadCM3 MIROC3 3 hires GFDL CM2 1 GFDL CM2 0 MIROC3 2 medres ECHO G UKMO HadGEM1 ECHAM5 MPI OM MRI CGCM2 3 2 CCSM3 CGCM3 1 T63 GISS AOM INM CM3 0 CGCM3 1 T47 FGOALS g1 0 CSIRO Mk3 5 CSIRO Mk3 0 CNRM CM3 BCCR BCM2
13. Thank you for ordering Consistent Climate Scenario Data through the Long Paddock website Your order number its682 and unique job label QMtest is now ready for download at ftp climate scenarios mft derm qid gov au Climate_Scenarios CCCS_Web_Data_Outputs stuart burgess The data will be removed from the FTP server in 7 days The user guide and other supporting documents are available at ftp climate mft derm qid gov au Climate_Scenarios Documentation Please note If an error message appears when attempting to download data from the FTP server it may be that your web browser is operating behind an institutional firewall which could be interfering with the file transfer process Use of a third party FTP client such as FileZilla with a fuller range of FTP options may overcome such difficulties If you experience unresolved difficulties with accessing information from the FTP server please contact Stuart Burgess DSITI Ph 07 3170 5528 Email stuart burgess science dsiti qid gov au The Consistent Climate Scenarios Data was developed by the Queensland Department of Science Information Technology and Innovation under funding from the Department of Agriculture Fisheries and Forestry s Australia s Farming Future Climate Change Research Program http www daff gov au climatechange australias farming future climate change and productivity research 108 Consistent Climate Scenarios User Guide Version 2 2 1
14. The projected pattern of a climate variable in the future according to any GCM is essentially independent of the SRES scenario and of time The error limits of this method are not well understood for many climate variables e The Consistent Climate Scenarios CCS project utilises the pattern scaling technique to compute trends per degree of global warming at specific locations in Australia for a number of climate variables However the pattern scaling technique has only been validated for temperature and precipitation e Not all of the climate variables of interest to this project were available in all GCMs Of the 19 GCMs used in this project only seven five from OzClim and two Hadley Centre models had a complete set of trends per degree of global warming for all of the climate variables required in CCS and did not require any infilling The seven GCMs were CSIRO 91 Department of Science Information Technology and Innovation Mk3 5 GISS AOM HADCM3 HADGEM1 INM CM3 0 MIROC3 2 hires and MIROC3 2 medres Trends in vapour pressure were not directly available from OzClim so changes in specific humidity have been used to compute those Trends per degree of global warming for missing climate variables have been estimated infilled using suitable methods and these are noted in the metadata e Trends per degree of global warming are computed from GCM monthly averaged data They do not include information abou
15. e Agriculture transforming to adapt to climate change Peanut industry expansion in the NT as a blueprint CSIRO e Development of effective management strategies to adapt production to mitigate climate change challenges in the wine industry Grape and Wine Research and Development Corporation GWRDC e Developing improved on ground practices and institutional policies for managing climate variability and climate change within beef production enterprises across northern Australia DEEDI Qld e Climate Change Adaptation in the Southern Livestock Industries Meat amp Livestock Australia e Amelioration of thermal stress impacts on animal performance and welfare in southern Australian dairy beef and sheep industries University of Melbourne e Adaptation of fisheries and fisheries management to climate change in south eastern Australia a national case study DPI Vic e Consistent Climate Scenarios QOCCE DSITIA Qld 104 Consistent Climate Scenarios User Guide Version 2 2 Consistent Climate Scenarios Web Portal http www longpaddock qld gov au climateprojections access html This example shows the process required to order 2030 daily projections data for Brisbane Airport based on the Change Factor method using the UKMO HadGEM1 GCM forced by the A1FI emissions scenario with medium climate warming sensitivity Home Climate change projections Data access Consistent Climate Scenarios data are currently
16. evaporation calculation method that is used in SILO Rayner 2005 9 4 Estimating solar radiation In this project the climate variable downward short wave radiation rsds is an important input into calculation of vapour pressure deficit As rsds was not readily available on PCMDI for two GCMs of interest CCCMA 47 and ECHO G it became necessary to estimate trends for those GCMs The first step in estimating solar radiation took an approach to find candidate relationships between the available climate variables and rsds using a correlation relationship to give reasonable estimates of trends in rsds This included using the reciprocal top of atmosphere radiation Ra being added to the standard climate variables After experimentation several GCMs Table 9 8 were selected for the availability of rsds plus the nominated predictands Table 9 8 GCMs selected for the availability of downward short wave radiation rsds GCM name used in Consistent Climate Expert Review Panel recommendation Scenarios Project CSIRO MK35 Not assessed but expected to be better than CSIRO Mk3 0 MIROC H MIROC M More likely to produce credible projections CCCMA 63 GISS AOM i liabl IAP FGOALS G10 Less likely to be reliable MRI GCM 232 CNRM Consistently underperformed BCCR INMCM Not recommended 82 Consistent Climate Scenarios User Guide Version 2 2 Trends in solar radiation as a percentage of the average are comp
17. plus 10 comparison of model projections plots 10 time series plots 50 quantile trend plots 60 QM histogram plots Department of Science Information Technology and Innovation 2 1 Change factor CF data Availability Change factor CF based projections data can be ordered through the Long Paddock website s Climate Change Projections web portal http www longpaddock qld gov au climateprojections Users have a choice of CF projections data files for 2030 and 2050 19 GCMs see Section 8 eight emissions scenarios see Section 7 1 and three climate warming sensitivities see Section 7 3 Each projections file contains projections data for six climate variables see Section 6 using the CF methodology see Section 4 CF projections data can be made available for any weather station or point location latitude and longitude within Australia For small orders the average delivery time to the ftp server is 60 minutes from being submitted For large orders delivery time is usually within 24 hours Users are notified by email with a link to their data on ftp server as soon as an order has been processed and is ready for download All CCS files on the ftp server are contained in ZIP format archives Users should note that the ZIP format archives will be deleted seven days after having been processed This is to ensure that the ftp site does not reach capacity enabling new file space to be created fo
18. 0196 0 0005 5 6363 0 3182 1 2720 871 0217 Jun 0 0094 0 0325 0 0001 4 8687 0 0072 2 0298 1996 3467 Jul 0 0205 0 0212 0 0006 6 9712 0 0339 0 7876 2369 9871 Aug 0 0059 0 0042 0 0013 11 4853 0 1778 0 4918 1280 0288 Sep 0 0040 0 0043 0 0004 26 0600 0 0607 1 6455 624 5360 Oct 0 0032 0 0091 0 0004 10 1548 0 0931 1 3102 451 3216 Nov 0 0039 0 0017 0 0002 15 3375 0 1875 2 5574 326 1852 Dec 0 0179 0 0001 0 0000 16 8268 0 2668 3 9200 543 2014 83 Department of Science Information Technology and Innovation Regressed RSDS v Model RSDS Separate regression parameters for each GCM monthly 45 y 0 9429x 0 01337 ae Estimated RSDS mri_cgcem2_3 2a ncar_ccsm3_0 Correlation po roxxbda 3348 3 D Model RSDS Figure 9 8 Plot of rsds trend x axis for individual GCMs against the prediction for that GCM y axis if the regression parameters are derived for each individual GCM and month This allows us to say in this case that all GCMs show the same general relationship between climate variables and the regression model produces a linear trend It should be noted that trends in downward short wave solar radiation appear to relate to trends in cloudiness and related water variables Note also that despite the signs being shown as positive above when the signs of the coefficients are taken into account the trends in rsds are generally inve
19. 48 Consistent Climate Scenarios User Guide Version 2 2 4 4 A worked example projecting climate data for 2050 for a specific location An outline of the above processes required to generate CCSP data based on the OzClim change factor approach is presented in Table 4 1 This example shows the steps required to calculate 2050 projections data for July at Brian Pastures using the CSIRO Mk3 5 GCM forced by the A1T emissions scenario with low climate warming sensitivity Table 4 1 An outline of the steps involved to calculate change factors for applying to a baseline climatology to produce 2050 climate projections for July at Brian Pastures This example uses CSIRO Mk3 5 forced by the A1T emissions scenario with low climate warming sensitivity Step 2 Step 3 Step 4 Step 1 Extract 21 Century GCM projections of global annual average surface temperature from a selected Global Climate Model i e CSIRO Mk3 5 Select climate variable i e rainfall or temperature and month July Extract the 21 Century projections average monthly values of climate variable at GCM model grid point Climate variable infilled by DSITIA if not available through OzClim Compute linear trend mean rate of change per degree of 21 Century global warming between projections of global annual average surface temperature and projections of monthly climate variable at GCM grid point level Repeat the
20. A protocol used for transferring files from one computer to another Global Climate Model GCM A GCM is built on a sophisticated computer program which uses mathematical equations based on the physical laws governing the behaviour of the earth climate system to simulate the global climate GCMs are used to produce climate projections for the 21st Century Grass Production Model GRASP A pasture simulation and water balance model specific to point locations integrating climate soil plant animal and management processes related to perennial grasses of Northern Australia http www longpaddock qld gov au GRASP Greenhouse Gases GHG Natural and anthropogenic man made gases in the atmosphere that absorb and emit infrared or heat radiation causing the greenhouse effect The main greenhouse gases are water vapour H20 carbon dioxide COz nitrous oxide NO and methane CH3 Interdecadal Pacific Oscillation IPO The IPO which has similarities to the Pacific Decadal Oscillation is a slow background change in Pacific Ocean sea surface temperatures which fluctuates between warm and cool phases on an inter decadal time scale and affects the relationship between the El Nifio Southern Oscillation ENSO and Queensland summer rainfall During cool phases of the IPO La Ni a events tend to be more frequent during cool phases of the IPO Phase changes of the IPO occurred in 1909 to cool 1922 to warm 1945 to cool 1978 to
21. Panel GCM CHARACTERISTICS MODEL FORCINGS RESOLUTION Horizontal resolution in degrees and the number of vertical levels L Ensemble OzClim Name Members well mixed greenhouse gases Sulphate direct Sulphate indirect Organic carbon Solar irradiance Volcanic aerosol Number of forcings 18 fispLows notusedinprojecd ust fe te af GISS ER not used in project E2_4 j j j j 10 NCAR PCM1 not used in projectirun4 h 15 GISS EH not used in project E1_2_3 j ft j j 10 Note The INMCM GCM is no longer recommended for use due to unstable drift in the model The rank for this model will be adjusted 64 Consistent Climate Scenarios User Guide Version 2 2 The CCSP Expert Panel classified the available GCMs according to each model s reliability in the Australian region Table 8 2 Crimp et al 2010 based on Smith and Chiew 2009 Table 8 2 also compares this ranking with that of Suppiah et al 2007 Similar rankings provided by other studies are shown in Table 8 4 As a result the panel recommended nineteen of these models for use in the preparation of CCSP projections data Table 8 2 Twenty three GCMs categorised in terms of their overall performance in simulating climate variables derived from an aggregate of global and regional statistics adapted from Smith and Chiew 2009 Crimp et al 2010 and Suppiah et al 2007 The grey shaded panels highlig
22. above steps for each grid point across Australia to produce the pattern of change mean rate per degree of 21 Century global warming across Australia for specific climate variables Step 5 Step 6 Interpolate regional pattern of change to a uniform 25km x 25km base then select pattern of change for a specific location In this example at Brian Pastures Location Code 040428 the CSIRO Mk3 5 pattern of change for rainfall per degree of 21 Century global warming obtained from the multiplier file see Section 3 1 is 18 68 and the pattern of change for minimum temperature per degree of global warming is 1 12 C Calculate climate change factor by multiplying the above pattern of change by the projected amount of global warming at 2050 The amount of global warming from MAGGIC for 2050 using the A1T emissions scenario with low climate warming sensitivity is 1 24 C obtained from the multiplier file see Section 3 1 For rainfall the 2050 change factor 1 24 x 18 68 is 23 16 i e a 23 16 decrease For temperature the 2050 change factor 1 24 x 1 12 is 1 39 i e a 1 39 C increase Step 7 Apply clamping if the change factor is outside the acceptable range Apply the change factors to suitable 20th Century baseline climatology to produce projections for the given climate variables In this example for rainfall we take 23 16 off all SILO historical July daily
23. and 8 Details on infilling of the missing trends per degree of 21st Century global warming for selected climate variables are provided in Section 9 11 2 Historical baseline and training period The CF projections data have not been de trended in any way In the CF Version 1 1 projections change factors could be applied to daily data from 1889 to present However in considering applications model evaluation we have since adopted the recommended use of a 1960 to 2010 baseline which has been applied in the options for ordering CF Version 1 2 projections data via the web The 1960 to 2010 baseline as a default is based on the improved quality of the historical data from 1960 onwards Users should note that OzClim CSIRO 2010 uses a fixed 1975 to 2004 baseline All QM projections have been de trended In the QM projections a 1957 2010 training period has been used to compute the perturbation rules that are applied to that historical baseline 11 3 Latitude and longitude in file names In the CF file names station latitudes and longitudes which were rounded to two decimal places prior to August 2011 are now rounded to four decimal places 11 4 Changes in SILO historical data Regular updates and quality control measures are applied to the SILO climate database and some changes may affect both CF and QM projections data For example on 26 January 2012 there was a significant update to the SILO climate database improving the precisi
24. au climateprojections Web based Aug 2012 data portal ug Portal on the Long Paddock website from which requests for CCS projections data can be made Department of Science Information Technology and Innovation 2 Accessing and interpreting data What is available The 2030 and 2050 daily AR4 based climate projections data are available for six climate variables useful for biological modelling including rainfall maximum and minimum temperature solar radiation vapour pressure pan evaporation Users can order projections data based on e the Change factor or Quantile matching method see Sections 4 and 5 e eight emissions scenarios see Section 7 1 e three climate warming sensitivities see Section 7 3 e 19 global climate models see Section 8 Registration The climate projections data are password protected To access to these data a new user must complete and submit the registration form at http www longpaddock gld gov au climateprojections registration ph Login details are provided by email once registered allow three working days Ordering data The climate projections data can be ordered by clicking the REQUEST DATA link at http www longpaddock gld gov au climateprojections registration ph Home Climate change projections Data access Consistent Climate Scenarios data are currently supplied free of charge However all users must be registered A
25. common 25km x 25km base However e the downscaling produced by this interpolation provides utility in terms of application but it does not imply any increase in accuracy over the native typically 150 250km GCM resolution 10 5 The calculation of trends per degree of global warming Regression has been used to calculate trends per degree of global warming for 2000 2100 against GCM data which may e contain outliers or e may be inherently non linear and or e the base line GCM model data may be implausibly low The following may produce impossible percentage trends per degree of global warming values e an implausible trend value is divided by a plausible base line value or e a plausible trend value is divided by an implausibly low base line value In either event when the trend per degree of global warming value is then multiplied by the observed value the error is propagated Australia has a land area equivalent to 1 5 of the Earth s total area 89 Department of Science Information Technology and Innovation In cases where OzClim trends per degree of global warming for some climate variables are not available for particular GCMs the data for those climate variables have been estimated and then in filled using either ratios or multiple regression techniques Various anomalies were found in some of the trend per degree of global warming files supplied by CSIRO for this project A search of monthly
26. corresponding to the specified location year 2030 A1Fl emissions scenario and high climate sensitivity That equates to a total of five CF projections data files one for each GCM For each location requested the ZIP archives contain a range of additional files In the ZIP archives these additional files include the corresponding observed data as contained in the SILO data base see Section 2 3 CO matching files log warning files monthly multiplier files historical time series plots and comparison of model projections plots see Section 3 As at January 20 2015 by default the years in the SILO observed daily data file and the CF projected daily data file are dated from January 1 1960 to December 31 2010 However if desired the end user can select a shorter or longer period when making their web based order the latest end year can be 2013 For example in calculating the frequency of hot days the end user can select a window of years e g 1971 to 2000 from the observations and compare the results with the same window 1971 to 2000 from the 2030 projections data files The selected historical baseline for the QM data is fixed from January 1 1957 to December 31 2010 The data from each file can be imported directly into a spread sheet or other program and the user can focus in this case on just the daily temperatures It is recommended that the end user perform some basic calculations and plots of the data and then refer to
27. does apply to QM 2050 solar radiation and specific humidity for which future CDFs are estimated using trend extrapolation The quantile trend plots are png files and are typically named Climate Variable_LocationCode_PivotQuantiles_1957_2010_Projections Year_QuantileTrends_Loc ationCode png e g T Min_040428 10 50 90 _1957_2010_2030 QuantileTrends_040428 png e ClimateVariable RadnPropOfEtlogit solar radiation transformed by ground level proportion of extra terrestrial then logit transform used RainCubeRoot rainfall transformed by cube root SH specific humidity transformed from vapour pressure T Max maximum temperature no transform T Min minimum temperature no transform e LocationCode is a six digit number BoM station code if patched point i e 002012 or all Zeros if drilled from interpolated surfaces i e 000000 e PivotQuantiles 10 50 90 denotes 0 1 0 5 and 0 9 quantiles e 1957_2010 historical training period 3 Quantiles differ from percentiles in that quantiles are expressed in sample fractions rather than sample percentages Quantile 0 1 is equivalent to the 10 percentile 39 Department of Science Information Technology and Innovation e 2030 projections year e QuantileTrends denotes Quantile trend plot file A snapshot of a quantile trend plot is presented in Figure 3 12 Augusts q 0 90 p 0 069 q 0 50 p 0 182 q 0 10 p 0 064 T Min oC
28. inputs to hydrological models and extend delivery of projections across Victoria Final report for SEACI Phase 1 Project 2 2 5P http www seaci org publications documents SEACI 1 20Reports S1_FR225P pdf Suppiah R Hennessy K J Whetton P H McInnes K Macadam l Bathols J Ricketts J amp Page C M 2007 Australian climate change projections derived from simulations performed for the IPCC 4th Assessment Report Aust Met Mag 56 2007 131 152 CSIRO Marine amp Atmospheric Research Australia US National Academies 2010 prepublication copy Climate Stabilization Targets Emissions Concentrations and Impacts over Decades to Millennia Committee on Stabilization Targets for Atmospheric Greenhouse Gas Concentrations Board on Atmospheric Sciences and Climate Division on Earth and Life Studies The National Academies Press Washington D C http www nap edu catalog 12877 html Watterson I G 2011 Understanding and partitioning future climates for Australian regions from CMIP3 using ocean warming indices Climatic Change Whetton P H R N Jones K L McInnes K J Hennessy R Suppiah C M Page J Bathols and P J Durack 2005 Australian climate change projections for impact assessment and policy application a review P H Whetton et al Series CSIRO Marine and Atmospheric Research paper 1 Aspendale Vic Wigley T M L Raper S C B Hulme M and Smith S 2000 The MAGICC SCENGEN Climate Scenario
29. is OzClim offers SRESscenarios B1 A1B and A1FI as low medium and high warming sensitivity options respectively To present this possible range of outcomes in a useful manner the emissions projections were ranked and the 10 50 median and 90 percentile projections were chosen to represent respectively a low L median M and high H global warming sensitivity to a given emissions scenario The process used to select an amount of global warming from MAGICC is represented in Figure 7 2 and is as follows 1 Select a scenario e g SRES A1Fl this yields a family of three curves 2 Select a year of interest e g 2050 3 Select a climate warming sensitivity low L medium M or high H In this particular example the result for A1Fl with high climate warming sensitivity is a single global warming offset of 2 35 C 61 Department of Science Information Technology and Innovation Selecting an amount of global warming for 2050 Emissions Scenario SRES A1FI with high climate warming sensitivity 6 0 90th percentile 5 0 4 Climate warming 2 sensitivity 50th percentile Le 4 0 4 x 2 35 C D amp 10th percentile 3 0 4 i a pieces ee eee high 7 2 204 median 1 0 i 0 0 i 1980 2000 2020 2040 2060 2080 2100 2120 Year Figure 7 2 Selecting the amount of global warming for 2050 Amounts of global warming
30. least one GCM from each RFC partition will provide coverage of a range of possible outcomes DSITI the Expert Panel and DAFF see the use of projections data based on these RFC related climate response patterns as an important step in ensuring comparability between projects 66 Consistent Climate Scenarios User Guide Version 2 2 Rainfall x 4 4 uoo oe E a e e e Rainfall x MIROC3 2 hires J High global warming and High global warming and 4 0 a warm Indian Ocean a warm Pacific Ocean ECHAMSIMPI OM 3 6 HI UKMO HadGEMt E m F T e J ek v Drier Wetter a MIROC3 2 medres UKMO Had CM3 HP Composite HI Composite E 3 2 J HP EPE A 4 E 163 Gros e 4 Pon CNRM CM3 p FGoaLs BCCR BCM2 0 2 G1 0 Rainfall yj 2 MRECGOM x Rainfall x oO caai CCSM3 P GFDL CM2 1 T47 e INM CM2 0 e CSIRO Mk3 0 GISS AOM Smaller global warming Smaller global warming a and a warm Indian Ocean and a warm Pacific Ocean Drier Wetter Drier Wetter 10 0 10 20 30 40 Wi Composite WP Composite West Pacific East Indian Index Figure 8 1 A partition of CMIP3 Global Climate Models GCMs for future climate using global warming sensitivity and ocean warming indices adapted from Watterson 2011 Values for nineteen individual GCMs forced by the SRES A1B emissions scenario are represented by the small dots and labelled by their GCM model code Table 8 2 The central horizontal and vertical lines separate
31. of Class A pans in 1970 the SILO files contain a synthetic estimate of Class A pan based on monthly spatially parameterised linear equations using vapour pressure deficit and solar radiation calibrated against actual pan measurements Estimates of pan evaporation under climate change are difficult as only estimates of Morton s wet area evaporation Morton 1983 are available for a small subset of models change factors were previously applied to SILO baseline data in the initial CF VO set As this estimate of evaporation trends was for large water bodies and only available for a subset of GCM s it was decided not to use this change factor and instead use change factors for temperature vapour pressure from specific humidity and solar radiation to compute an empirical equation to estimate a new synthetic pan estimate using the SILO spatial parameters Rayner 2005 While this was consistent with the pre 1970 synthetic pan data it introduced an inconsistency between the baseline SILO data set which contains actual pan estimates post 1970 and the climate changed data which produced a synthetic pan estimate for this period This inconsistency was eliminated in data sets supplied after July 2011 by applying the ratio of climate changed synthetic pan to baseline synthetic pan to the post 1970 actual pan data It should be noted that while synthetic pan is on average good proxy for measured pan it is less variable due to lack of
32. of downscaling Coupled Model Intercomparison Project Phase 3 CMIP3 The database formed by CMIP3 includes experiments from the 23 GCM models submitted to the International Panel for Climate Change IPCC Fourth Assessment Report AR4 Cumulative distribution function CDF The cumulative distribution function gives the probability that an observation X is less than or equal to a given value x and gives the percentile or quantile rank of an observation Debiasing Removal of systematic error Department of Environmental and Resource Management DERM DERM handled environmental issues facing Queensland including securing water for the future managing land use meeting the challenges of climate change and conserving the state s natural and cultural heritage Department of Science Information Technology and Innovation DSITI Formed in February 2015 DSITI works closely with all agencies to drive the Queensland government s priorities in research science innovation technology intellectual property service delivery 98 Consistent Climate Scenarios User Guide Version 2 2 Department of Science Information Technology Innovation and the Arts DSITIA Formed in May 2012 DSITIA brought together science information and technology innovation and the arts in one portfolio to help drive Queensland s economy The department was a critical enabler to support tourism agriculture mining and construction In February 2015 DS
33. phases of the Interdecadal Pacific Oscillation IPO Consistent Climate Scenarios User Guide Version 2 2 Table 1 1 Consistent Climate Scenarios Project Versioning Version Release date Comments Vo Apr 2010 Initial CF 2030 and 2050 projections test data for format checks etc Change yi Sep 2010 Eight GCMs eight emissions scenarios three climate sensitivities six climate variables and factor two projections years 2030 and 2050 CF data 17 GCMs eight emissions scenarios three climate sensitivities six climate variables and V11 Apr 2011 two projections years 2030 and 2050 These files Some improvements on evaporation Historical baseline 1899 to current have no method tag This is the version running under the web Historical baseline 1960 2010 Includes two V1 2 Jun 2012 Hadley Centre GCMs HADCM3 and HADGEM1 Otherwise there is no difference in the data from that of V1 1 QMV2 1 0 2030 Jun 2011 Initial 2030 test set QM in filename QMV2 2 0 2030 Initially 17 GCMs eight emissions scenarios three climate sensitivities six climate variables Quantile 2030 amp QM Jul 2011 and one projections year 2030 Two extra GCMs HADCM3 HADGEM1 added for the web matching v22 ionii g in fil version in June 2012 QM data In Mename QMV3 0 2030 May 2015 Code adjusted to fix a trivial error in QMV2 2 0 for which an insignificant amount of daily These fi
34. quantile rank using quantile matching functions that have been computed for each location month and climate variable Figure 5 2 Projected values in the lower upper end of the target distribution i e beyond pivot quantile 0 1 0 9 are based on the same shift that has been applied to the 0 1 0 9 quantile QMF identity eeo pivot points Pivot quantile 0 9 Pivot quantile 0 5 Projected Tmax 20 22 Observed Tmax Figure 5 2 Example of a quantile matching function for a specific month for daily maximum temperature showing the target CDF for 2030 black line 51 Department of Science Information Technology and Innovation Step 4 Historical climate variability is projected forward by computing quantile trend residuals interpolated about the pivot quantiles and applying these to the projected data by a second QM procedure Figure 5 3 Without this step there is too little inter annual variation in the projected time series q 0 50 p 0 051 Septembers TMax This September e010 p041 had a high 90th percentile relative to the g 0 9 trend 35 30 25 Tmax deg C ANP This September had a low 90th percentile relative to the g 0 9 trend 1970 1980 1990 2000 2010 2020 2030 Figure 5 3 Quantile trend plot for September daily maximum temperature showing residuals The projected data are adjusted
35. rainfall values to obtain projected data for 2050 For temperature we add 1 39 C to all SILO historical July daily minimum temperature values to obtain projected data for 2050 49 Department of Science Information Technology and Innovation 5 Quantile matching QM methodology This section of the User Guide provides an outline of the methodology that DSITI has used to prepare QM climate projections data for 2030 and 2050 Further information describing the quantile matching approach is documented in Kokic et al 2012 and Panjkov 2012 The QM method produces projected daily data for the future by mapping historical cumulative distribution functions CDFs sourced from a 1957 to 2010 training period to a supposed future CDF A variation in the QM method is used depending on whether projections for 2030 or 2050 are required Different methods are required as there is almost no daily data for GCMs around 2030 In addition even where data do exist no daily data for surface water vapour and solar radiation are available For 2030 the future CDF is estimated by the forward projection of historical trends in monthly quantiles out to that year However beyond 2030 there is a risk that historically based quantile trends may meet or cross each other at some point in time particularly under emissions scenarios associated with high climate sensitivity to global warming Therefore to acquire 2050 projections dat
36. research and climate applications http Awww longpaddock qld gov au SILO Special Report on Emissions Scenarios SRES A report prepared by the Intergovernmental Panel on Climate Change IPCC for the Third Assessment Report TAR in 2000 on future emission scenarios to be used for driving global circulation models to develop climate change scenarios The SRES Scenarios were also used for the Fourth Assessment Report AR4 in 2007 Stationarity Stationarity occurs when the mean and variance of a statistic do not change over time Synthetic data Best estimates used as a proxy where either no or anomalous data exists Trend per degree global warming The projected mean annual rate of change for a specific climate variable per degree of global warming 100 Consistent Climate Scenarios User Guide Version 2 2 13 References Crimp S Kokic P McKeon G Smith I Syktus J Timbal B and Whetton P 2010 A review of appropriate statistical downscaling approaches to apply as part of Phase 2 of the Consistent Climate Projection project CSIRO National Research Flagships Climate Adaptation CSIRO 2009 Glossary retrieved May 23 2011 from https wiki csiro au confluence display ozclim Glossary CSIRO 2010 Welcome to OzClim Exploring climate change scenarios for Australia Retrieved May 23 2011 from http www csiro au ozclim Dai A Wigley T Meehl G and Washington W 2001 Effects of stabili
37. scanned from Galan Manne and Richard Richels 1997 On stabilizing C02 concentrations cost effective emission reduction strategies Environmental Bodelling and Assessment 2 1997 SNE All climate change factors are pattern scaled off the A18 scenario NB p subsscript preliminary and IS92 are for older scenarios FRCP scenarios are available at bttp svv iiasa ac at veb apps tnt Rcplb dsd Action htmlpagespage compare innual values interpolated from decadal values Tear 1970 1971 1972 1970 8 sm to i 2094 2100 2035 2096 2097 2098 2099 2100 Figure 3 4 A snapshot of information contained in the annual CO concentrations data file CO2_concentrations_annual dat 118 325 000 326 200 327 400 328 600 329 800 685 600 688 500 691 400 694 300 697 200 700 100 703 000 Alt 325 000 326 200 327 400 328 600 329 800 573 200 573 500 573 600 74 100 574 400 574 700 575 000 AIFI 325 000 326 200 327 400 328 600 329 800 906 400 915 000 923 600 932 200 940 800 949 400 958 000 a 325 000 326 200 327 400 328 600 329 800 786 800 795 000 803 200 811 400 819 600 827 800 836 000 Bi 325 000 326 200 327 400 328 600 329 800 538 800 539 000 539 200 539 400 539 600 39 800 540 000 B2 325 000 326 200 327 400 328 600 329 800 593 000 596 000 599 000 602 000 605 000 608 000 611 000 AR4 IPCC SRES emissions scenarios Preliminary estimates kip 325 000
38. supplied free of charge However all users must be registered A free sample of the data ZIP 1 3M last updated 03 00PM 14 May 2012 is available for testing without being registered To register for data access please fill out and submit the registration form Once we have returned your registration details you can start requesting data Only small orders for data can be accepted via this web site For example you can only download data for up to ten locations and up to three emission scenarios at any one time If you wish to order larger amounts of data please contact Grazing Land Systems REQUEST DATA Please Enter your Credentials to Access System Do not bookmark this page If you wish to create a bookmark wait until after you have logged in Email Address Password Data Order Step 1 Select order type Choose location type Weather stations Latitude and Longitude Type of data required Full data daily projections Summary data monthly mean change factors for all models scenarios etc Data Order Step 2 Select weather stations Search by station number or name Within boundary box Station number or name brisbane Min Latitude Min Longitude Within state QLD w Max Latitude Max Longitude Selected stations Station ID Station Name 40223 BRISBANE AERO This online form only allows for up to ten stations For larger requests please use our feedback form to request the informatio
39. surface To is temperature at the surface From this we can derive the following for the derivative of Pp which should be similar to dP 1 dN dh h 1 where B subsumes the constants dt B dt dt the trends in huss So trends in huss ought to be related to trends in precipitable water and surface temperature However this is a static model and climate models involve mobile parcels of air Furthermore we are dealing with monthly averaged data Technique As in the estimation of temperature trends climate model monthly trends were extracted for 13 selected sites distributed within Australia for all climate variables their present day averages ave trend in change per degree global warming and percentage trend per degree of warming tpc if available The CSIRO supplied climate variables deemed likely to be predictands of huss were precipitable water prw cloud cover clt temperature tas precipitation pr and solar radiation rsds Additionally DSITI holds the monthly values for atmospheric specific humidity at standard pressures hus and a separate investigation was carried out to determine the feasibility of predicting huss directly from trends in huss at some low pressure level In the later event the relationship seems to break down near the coast and so a measure of distance from coast was added to compensate for this After experimentation the GCMs listed in Table 9 6 were selected for the availability of
40. the CCS FTP Data collection site as at May 1 2015 Important notes for users regarding ZIP archives Windows Self extracting ZIP archives file extension zip exe only work for the Windows environment They won t work when downloaded to non Linux x86 Unix systems e g HP UX AIX Solaris DG UX IRIX TRU64 OSF 1 Unix and Linux Standard ZIP files file extension zip can be provided for Unix and Linux users Unix users can extract the projections data files by typing unzip filename zip in the Unix command line ZIP archive file sizes When ordering data the user needs to consider the size of zipped archives and the number of files that will be produced Table 2 1 shows typical zip archive file sizes and the number of files that will be produced Consistent Climate Scenarios User Guide Version 2 2 Table 2 1 Typical ZIP archive file sizes and number of files based on the selection of the default 1960 2010 climate baseline including diagnostic plots Projections type CF 2030 or 2050 projections User selection 1 location 1 GCM 1 emissions scenario 1 climate sensitivity 0 5 MB 5 files 1 projections file 1 SILO file 1 multiplier file 1 CO matching file 1 log file User selection 10 locations 1 GCM 1 emissions scenario 1 climate sensitivity 4 8 MB 50 files 10 projections files 10 SILO files 10 multiplier files 10 CO matching files 10 log files User selection 10 lo
41. the four Representative Future Climate RFC partitions The larger dots indicate the CCS composite means for GCMs within each of the four RFC responses HI high global warming and a warmer Indian Ocean HP high global warming and a warmer Pacific Ocean WI lower global warming and a warmer Indian Ocean and WP lower global warming and a warmer Pacific Ocean The maps show projected 21 Century changes in rainfall for the GCMs clustered in each of the four HI HP WI and WP RFC partitions Maximum Temperature 4 4 F Maximum Temperature L MIROC3 2 hires High global warming and High global warming and 4 0 a warm Indian Ocean a warm Pacific Ocean ECHAMSIMPILOM 3 6 F HI UKMO HadGEM1 om I Z L e a m Less warming More warming w ete MIROC3 2 medres ae E 32l UKMO HagCM3 i R M ee Composite d CGCM3 1 T63 gro e CSIRO Mk3 5 WP Composite oO lig o z Saia CNRM CM3 m Maximum Temperature 3 2 8 fesoa BCCR BCM 2 0 Maximum Temperature F MRLCGCM 232 e 7 uo CCSM3 GFDL CM2 1 24 CGCM3 1 WP T47 W INM CM3 0 a CSIRO Mk3 0 2 0 F GIS AOM m me L Smaller global warming Smaller global warming Less warming More warming w and a warm Indian Ocean and a warm Pacific Ocean N Less warming More warming 1 6 1 1 1 1 1 1 1 1 1 1 WI Composite 20 10 0 10 20 30 40 HP Composite West Pacific East Indian Index 67 Department of Science
42. trends per degree of global warming ss sssnnssnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnn nanna 70 9 1 Estimating trends in daily maximum and minimum temperature 71 9 2 Estimating vapour pressure 76 9 3 Estimating potential evaporation and pan evaporation 82 9 4 Estimating solar radiation 82 9 5 Summary of infilling 84 10 Known limitations of CF projections Gata sisisisasicsssssesssstessietesaiesscsstesatsasieariansieaatsasieaniatinreieiaieanne 86 10 1 Base period selection 86 10 2 Capture of anomalous data in Log warning files 88 10 3 Emissions and CO stabilisation scenarios 88 10 4 Downscaling from Global Climate Models 89 10 5 The calculation of trends per degree of global warming 89 10 6 Known issues related to the calculation of change factors 91 10 7 Issues important to biological modelling 92 11 Differences between CF and QM projections data including versioning sssseseeeeees 94 11 1 GCMs emissions scenarios climate sensitivities and projections years 94 11 2 Historical baseline and training period 94 11 3 Latitude and longitude in file names 94 11 4 Changes in SILO historical data 94 11 5 Quality control measures 94 11 6 Calculation of pan evaporation 95 11 7 Projected CF and QM means and standard deviations 95 11 8 Non uniformity of perturbations 96 11 9 Differences between CF and QM projections 96 11 10 Changes between QMV2 2 0 and QMV3 0 96 11 11 Summary of differences between CF and QM versionin
43. 0 HI Composite Maps represents the mean change in the climate variable i e increase decrease per degree of 21st Century global warming using the specified GCM Rainfall UKMO HadGEM1 Developed at the Hadley Centre for Climate Prediction and Research Met Office UKMO United Kingdom 38 atmospheric levels and 40 ocean layers Well mixed greenhouse gases ozone direct and indirect sulphates black carbon organic carbon land use solar irradiance and volcanic aerosols CCS Expert Panel assessment More likely to produce credible projections 107 Department of Science Information Technology and Innovation Data Order Step 5 Confirm order Please confirm and submit your order 1 weather stations selected 40223 BRISBANE AERO Perturbation method Change factor Baseline z 1960 2010 Projection year z 2030 Emission scenarios SRES Marker Scenario A1FI Warming sensitivities Medium Climate models UKMO HadGEM1 Data file details Data file label CFtest Platform Windows Output format APSIM Include charts No Delivery method FIP Estimated size of your order is 1 Mb Estimated processing time is 24 hours Data Order Order confirmation Thank you your order has been received You can view your order and check on its progress at View your orders If you would like to place a further order click here Create a new order Subject CCCS Data Order job number its682
44. 0 1 574731 68 523529 498 9 BoR ALT 11 NOV 2050 high 2 22 3 3 3 2 7 93925 Q 0 0 000000 5 Q 1 308976 8S 201961 49 9 Sock alt 12 pec 2050 high 2 22 2 2 10625 1 2 333 47 2 0 000000 0 1 072874 90 903922 493 9 Socr ALT 0 sequal 2050 high 2 22 1 02795 1 7 5 772 O 000000 Q 1 038065 774 505882 498 9 Soca ALT 1 Jan 2030 0 29 1 n 7 7 b 009009 0 975347 111 996078 437 9 Soca AlT 2 feb 2030 med 0 2 1 2 67 2000 0 974985 84 376471 437 9 S0 ALT 3 Mar 2030 med 0 1 0323 7 f 0 000000 D 0 963190 58 817647 437 9 Socr aiT 4 ser 2030 med 0 2 7 0883 7 5 5 7 0 000000 0 995511 42 356863 437 9 Socr alt 5 may 2030 med 0 1 7 326100 e a2 0 000000 0 0 956710 49 076471 437 9 BocR ALT 6 zn 2030 med 0 0 02 7 t 30 0 000000 Q 0 856519 40 072549 437 9 BoR ALT 7 Jul 2030 wed 0 0 2 37 0 000000 0 745106 46 715686 437 9 Bock alt 8 oo 2030 med 0 0 7 482471 7 Q 0 000000 o 0 0 858558 44 627451 437 9 Boca AlT 9 Sep 2030 med 2 0 x 3 0 000000 0 1 158277 48 837255 437 9 SccR alt 10 oct 2030 med 822 1 2 2 452 o b 1 574731 68 523529 437 9 Sacr alt 1 Nov 2030 med i 9 2 7 9392 Q2 0 00 0 Q 1 304976 868 201961 437 9 BoR ALT 12 ec 2030 med o 0 1 104257 752 7 47 Q 2 0 5 1 072874 90 po 437 9 E i i P i GCM Emissions Projection Climate Monthly multipliers for specific climate variables scenario year sensitivity absolute values for temperature change for others Figure 3 1 A snapshot of information co
45. 0 time sequences to suit their project requirements although this may change the statistics of runs of above and below average rainfall based on the NOAA Oceanic Nifio Index NOAA 2011 87 Department of Science Information Technology and Innovation 10 2 Capture of anomalous data in Log warning files Log warning files have been provided with the CF data for each point location and GCM model run The Log warning files hold information to alert users to any problematic data and have been described earlier see Section 3 4 of the User Guide For example a log will be generated in a Log warning file whenever pattern scaling produces negative projections or where data lie outside the bounds of what may reasonably be expected e Inthe cases where trends in climate variables have been logged as Implausible the associated projections data have been re set to more realistic values according to Table 3 1 For example this means that negative rainfall totals will not occur such values could case biological models to fail e n some cases usually individual days anomalous values may occur These values may be derived from one of three sources which are 1 the raw data 2 interpolation or 3 the modification to climate changed data e Users should also note that any precision in the values i e change rates listed in the Log warning files is for calculation purposes only and that this precision will not occur i
46. 0000 e Latitude of the station or location in decimal degrees e Longitude as above e VersionNumber V1 2 represents CF data e Multiplier this is a notepad file Variables contained in the CF V1 2 multiplier files are e Column 1 Model Model name for each of the 19 GCMs listed in Table 8 2 and four GCM composites based on the Representative Future Climate partitions WP WI HP and HI listed in Table 8 3 e Column 2 Scenario Emissions scenario A1Fl A1B A1T A2 B1 B2 CO2_450 and CO2_550 e Column 3 Mnth Month numeric 1 12 13 e Column 4 Month Month alpha numeric Jan to Dec Annual e Column 5 Year Projections year 2030 and 2050 e Column6and7 Sensitivity Climate warming sensitivity low median high refers to the 10 50 and 90 percentile values respectively and a value indicating the projected amount of global warming C at 2030 or 2050 Column 8 to 18 Projected change per degree of 21 Century global warming for e Tasmax and Tasmin Maximum and minimum temperature absolute change C e Precipitation Rainfall per cent change e RSDS Solar radiation per cent change e huss_tpc Specific humidity per cent change e RH Relative humidity per cent change if available e WVap Water vapour per cent change 0 if not available e WSP Wind speed per cent change o if not available e Taverage Mean temperature absolute change C e SlLORain SILO rainfall observed mea
47. 002_SILO_ 30 5167_151 6681_V1 2 met Notes The metadata at the start of each APSIM format SILO historical data file include station details location code name latitude longitude baseline climate statistics annual average ambient temperature and annual amplitude in mean monthly temperature In the QM historical data files an additional code column contains 6 digits outlining codes used to distinguish between actual observations or interpolated data When ordering CF 2030 or CF 2050 projections data users can select historical baseline climate data for years from 1960 onwards The default period for calculating the long term climate statistics tav and amp is currently 1960 to 2010 climate changea However if 1970 to 2000 is selected calculations will be based on that specific period climate changed In the QM historical data files the default period for both the historical baseline climate data and calculated tav and amp is 1957 to 2010 climate changed For QM the default 1957 to 2010 period can t be changed 18 Consistent Climate Scenarios User Guide Version 2 2 2 4 An end user example A typical sequence for an end user is provided by the following example An end user is interested in studying the impacts on grape vines near Mildura due to a change in the frequency of hot days days where the maximum temperature exceeds 35 C They would like to know wh
48. 030 or 2050 e Climate warming sensitivity rate of global warming i e L M H L M and H refer to the 10 50 and 90 percentile values respectively e ModelName i e CSIRO MK35 HADGEM1 HI HP etc e Latitude and longitude of the station or location in decimal degrees rounded to 4 decimal places e CF Version i e V1 2 e SILO format either met for APSIM or p51 for GRASP Emissions scenarios and climate warming sensitivities used in the project are discussed in Section 7 More information about the AR4 GCMs and Representative Future Climate partition model composites used in this project is available in Section 8 Department of Science Information Technology and Innovation CF Projections file metadata The CF projections files in APSIM format contain the following metadata in the first 21 rows of each file Station number This is the same as the Location code following the same convention as is used by the Bureau of Meteorology BoM which consists of six digits containing leading zeros For example station 51039 in the SILO database adopts the BoM station identifier convention and becomes LocationCode 051039 for this project Station name None is listed if the location is selected by latitude and longitude Latitude and longitude decimal degrees Long term annual average ambient temperature tav C Perturbed tav based on the default period curre
49. 051039_ 31 5495 147 1961_V1 2 png described in Section 3 5 o Quantile trend plots files for 5 climate variables RadnPropOfEtlogit_051039_ 10 50 90 _1957_2010_2030_QuantileTrends_051039 png RainCubeRoot_051039_ 10 50 90 _1957_2010_2030_QuantileTrends_051039 png H_051039_ 10 50 90 _1957_2010_2030_QuantileTrends_051039 png T Max_051039_ 10 50 90 _1957_2010_2030_QuantileTrends_051039 png T Min_051039_ 10 50 90 _1957_2010_2030_QuantileTrends_051039 png described in Section 3 9 o Histograms of QM projections files for 6 climate variables RadnPropOfEtlogit_051039_ 10 50 90 _1957_2010_2030_Histograms_051039_HADCM3_A1FI_high png Rain_051039_ 10 50 90 _1957_2010_2030_Histograms_051039_HADCM3_A1FI_high png RainCubeRoot_051039_ 10 50 90 _1957_2010_2030_Histograms_051039_HADCM3_A1FI_high png VP_051039_ 10 50 90 _1957_2010_2030_Histograms_051039_HADCM3_A1FI_high png T Max_051039_ 10 50 90 _1957_2010_2030_Histograms_051039_HADCM3_A1FI_high png T Min_051039_ 10 50 90 _1957_2010_2030_Histograms_051039_HADCM3_A1FI_high png described in Section 3 10 File naming syntax for Quantile trend and Histogram plots supersedes that presented in earlier User Guides QM 2030 Projections files Once a ZIP archive is opened individual QM 2030 climate projections data files and ancillary files are then accessible The QM 2030 climate projections data files are named as follows Lo
50. 09
51. 0_H_BCCR_ 30 5167_151 6681_vV1 2 apsim 447 362 056002_A2_2050_H_BCCR_ 30 5167_151 6681_vV1 2 apsim 526 355 056002_A2_2030_M_BCCR_ 30 5167_151 6681_vV1 2 apsim 447 362 056002_A2_2050_M_BCCR_ 30 5167_151 6681_vW1 2 apsim 526 355 056002_A2_2030_L_BCCR_ 30 5167_151 6681_vV1 2 apsim 447 362 056002_A2_2050_L_BCCR_ 30 5167_151 6681_v1 2 apsim 526 355 Projected CO Figure 3 2 A snapshot of information presented in a CO matching file filename 056002_ 30 5167_151 6681_NamesList txt 3 3 CO concentrations files Two CO concentrations files are available These are an annual file named CO2_concentrations_annual dat containing CO2 concentrations for each year from 1970 to 2100 a decadal file named CO2_concentrations_decadal dat containing CO2 concentrations for each decade from 1970 to 2100 These files can be downloaded from ftp climate mft derm qld gov au Climate_Scenarios Documentation The CO data used in the CCS project and listed in these CO concentrations files is sourced from the IPCC The IPCC has documented a range of emission scenarios featured in the Special Report on Emissions Scenarios SRES IPCC 2000 Six of the scenarios documented by the IPCC used in both AR3 and AR amp 4 are utilised in this project representing outcomes of distinct narratives of economic development demographic and technological change The SRES scenarios are A1Fl A1B A1T A2 B1 and B2 see additional technical details des
52. 2549 437 9 SocA alt 7 wl 2030 high 1 30 0 0 397 SL 997 0 b Q 46 715686 437 9 Bock aiT ug 2030 high 1 30 0 482471 7 0 9 0 oO 44 627451 437 9 Socr AIT 9 sep 2030 high 1 30 1 e 5 7 2 Q 1 15827 48 837255 437 9 SocR ALT 10 Oo 2030 high 1 30 00S 0 860 7 7 7 1 68 523529 437 9 Sok alr nu Nov 2030 high 1 30 1 1 2 1 308976 63 201961 437 9 Sok alt 12 Dec 2030 high 1 30 0 1 2 2 7 471318 O 0 000000 Q 1 072874 90 903922 437 9 BoR ALT 0 Armual 2030 hi 1 30 1 02795 1 7 o Q le 7900 O 000000 Q 1 038065 774 505882 437 9 Soca alt 1 Jan 2050 high 2 22 0 2 1 1 3792 7 08277 0 2 0 000000 0 0 975347 111 996078 498 9 Soca alt 2 Feb 2050 high 2 22 0 5 1 70 742 OF 63922 Q 000000 Q 0 974985 84 376471 498 9 BCCR ALT 3 Mar 2050 high 2 22 0 1 7 3 7 A 00282 0 000000 x 0 93190 58 817647 498 9 SR alt 4 apre 2050 high 2 22 0 1 1 27 le 0 7 0 000000 5 0 995521 42 356863 498 9 Soca ALT 5 May 2050 high 2 22 0 85 a 72 7 2 2 b 0 000000 5 Q 0 956710 49 076471 49 9 Sor aiT 6 Jun 2050 high 2 22 0 82 0 5 302115 0 002 e 0 000000 5 0 0 8646519 40 072549 498 9 Bock ALT 7 Jul 2050 high 2 22 0 2 0 2 397956 77 37499 0 000000 00 0 0 745106 46 715686 498 9 Boca alt s ag 2050 high 2 22 0 0 7 13 482471 7 Q 0 000000 x 0 0 866553 44 627451 498 9 Sock aiT 9 sep 2050 high 2 22 2 322522 0 10 596473 amp 2 000000 OF 1 158277 48 837255 498 9 Sock ALT 10 oct 2050 high 2 22 22 1 4812 2 9 i Q 7003 Q 000000 5
53. 27 Department of Science Information Technology and Innovation 3 4 Log warning files Log warning files see naming convention below accompany the CF data for each point location and model run The log warning files hold information to alert users to any problematic data For example data are generated for inclusion in log warning files when projections data lie outside the bounds of what may reasonably be expected The log warning files are typically named LocationCode_Latitude_Longitude_VersionNo log e g 056002 30 5167_151 6681_V1 2 log e LocationCode is a six digit number BoM station code if patched point i e 051039 or all Zeros if drilled from interpolated surfaces i e 000000 e Latitude of the station or location in decimal degrees e Longitude as above e VersionNumber V1 2 represents CF data The log warning files include e ModelName i e CSIRO MK35 HADGEM1 HI HP etc e Emissions Scenario A1Fl A1B A1T A2 B1 B2 CO2_450 or CO2_550 e Climate Warming sensitivity low median or high L M and H refer to the 10 50 and 90 percentile values respectively e Projections year 2030 or 2050 e Number of projected Tmin greater than projected Tmax Instances when the projected minimum temperature is greater than the maximum projected temperature for the day e Number of projected Vp greater than VpSat Instances where the projected vapour pressure is checked again
54. 30 projections file contains projections data for six climate variables using the QM 2030 methodology see Section 5 1 QM 2030 projections data can be requested for any weather station in Australia However drilled data derived from interpolated surfaces are not currently available QM 2030 based projections data can be ordered through the Long Paddock website s Climate Change Projections web portal http www longpaddock qld qov au climateprojections Users are notified by email once an order for QM 2030 data has been processed and is ready for download refer to page 6 The files are contained in a ZIP format archive 12 Consistent Climate Scenarios User Guide Version 2 2 QM 2030 ZIP archives containing daily projections QM climate projections data files as well as a set of historical baseline climate data files monthly multiplier files CO2 matching files for each selected climate site are contained in a ZIP format archive named as follows User name_JobNumber_FileLabel_Archivetype e g john smith_its138_FileLabel_zip exe e Username derived from your email address e JobNumber i e its138 its139 its140 etc e FileLabel 1 8 character label specified by the user i e qm2030 qmdata etc e Archivetype i e zip exe for Windows or zip for Unix Linux File types packaged in the QM ZIP archive containing daily projections including filename examples for each selected climate site are
55. 3_ 25 6550_151 7450_V1 2 met Department of Science Information Technology and Innovation p51 format Julian day Lat Long of year 7 31 5495 147 1961 syn pan pre 70 S1LO39NYNGAN AIRPO date jday tmax tmin rain evap rad vp 18890101 39 7 25 3 0 0 10 7 26 1 21 2 Source year 18890102 Z 38 2 26 3 0 0 11 0 25 1 20 1 18890103 3 38 2 21 3 0 0 9 6 27 1 19 1 18830104 4 33 7 18 3 11 3 9 0 27 1 17 0 18890105 5 30 2 16 3 0 0 8 7 27 1 13 8 18890106 6 30 7 16 3 0 0 8 9 27 1 17 0 18890107 7 31 2 18 3 0 0 9 8 26 1 14 8 18830108 8 34 2 18 3 0 0 10 7 27 1 15 9 18890109 9 36 7 19 3 0 0 11 6 27 1 17 0 18830110 10 39 2 20 8 0 0 12 2 26 1 15 9 18890111 11 41 2 22 3 0 0 13 0 25 1 17 0 18890112 12 43 2 24 3 0 0 11 6 25 1 18 0 18830113 13 41 2 26 3 0 0 12 9 23 1 20 1 18890114 14 43 7 27 3 0 0 12 3 23 1 19 1 18890115 15 41 7 23 8 0 0 11 1 25 1 18 0 18890116 16 39 2 23 8 0 0 11 2 25 1 19 1 Maximum air temperature Minimum air temperature Rainfall Pan evaporation Vapour pressure Solar radiation Figure 2 4 A snapshot of information projections for 2030 presented in a climate projections data file suitable for use in GRASP filename 051039 _A1FI_2030_M_CSIRO MK35_ 31 5495 147 1961_V1 2 p51 2 2 Quantile matched QM data Availability of QM 2030 projections data For Quantile Matched QM 2030 data users have a choice of 19 GCMs eight emissions scenarios and three climate warming sensitivities Each QM 20
56. 4 84 20 27 36 29 3 32 4 37 2 61 0 61 11 00 14 33 4 67 1983 02 2 400 min 2098 21 85 23 83 22 12 18 29 17 59 1112 977 11 52 14 87 17 82 19 83 19 95 1983 02 tmax 2098 36 29 38 40 34 64 30 99 25 84 22 46 24 07 24 14 31 42 31 81 31 87 33 53 1983 02 pee 2099 10 15 1 77 7 62 2 37 1 74 1 54 11 87 1 73 20 20 15 30 1 42 13 56 1984 02 2 420 min 2099 22 89 22 34 20 61 17 30 12 67 10 74 10 23 9 39 12 54 15 95 18 79 21 91 1984 02 tmax 2099 34 37 35 12 33 58 31 30 27 25 24 82 23 60 25 25 28 03 29 15 32 77 35 75 1984 02 a 2100 4 17 13 35 2 58 202 5 39 4 90 3 08 3 83 3 37 11 37 4 69 14 22 1985 02 min 2100 22 07 24 05 21 99 17 44 14 85 842 9 87 10 60 13 08 16 41 20 87 22 47 1985 02 2 kmax 2100 36 84 36 61 34 32 32 05 27 40 23 28 24 88 26 55 29 54 30 15 33 10 36 53 1985 0 2 430 Figure 3 14 A snapshot of information presented in a transient climate change data file suitable for use in CENTURY filename 035149 24 8353 149 8003 NCAR CCSM_CO2 550_med_CF1 2EM09 wth In addition to the transient climate change data files the zip archive includes multiplier files from which change factors have been calculated The multiplier files which are described in Section 3 1 of this User Guide are named as follows LocationCode_Latitude_Longitude_VersionNo multiplier e g 035149 24 8353 149 8003_v1 2 multiplier 44 Consistent Climate Scenarios User Guide Version 2 2 4 Change factor CF methodology This section of th
57. 9 435 2050 22 00 496 00 555 00 522 00 462 00 473 00 529 00 525 00 467 00 481 00 499 00 507 00 445 00 490 00 466 535 477 670 442 700 540 543 2060 563 00 523 00 625 00 563 00 503 00 499 00 569 00 571 00 492 00 06 00 533 00 41 00 447 00 510 00 508 871 510 634 441 673 603 520 2070 601 00 545 00 702 00 620 00 518 00 524 00 606 00 622 00 515 00 532 00 568 00 77 00 449 00 525 00 524 302 549 620 437 461 677 078 2080 639 00 563 00 786 00 682 00 530 00 552 00 642 00 683 00 537 00 559 00 607 00 616 00 449 00 536 00 531 138 594 257 431 617 758 162 2090 674 00 572 00 872 00 754 00 538 00 581 00 bg ns 754 00 555 00 588 00 653 00 660 00 450 00 544 00 533 741 635 649 426 005 844 805 2100 703 00 575 00 958 00 836 00 540 00 611 00 836 00 569 00 618 00 703 00 709 00 450 00 550 00 538 358 669 723 420 895 935 874 AR4 IPCC SRES emissions scenarios CO stabilisation scenarios Preliminary estimates Older scenarios ARS5 RCP scenarios Figure 3 3 Observed pre 2010 and projected post 2010 decadal CO concentrations contained in the decadal file CO2_concentrations_decadal dat 26 Consistent Climate Scenarios User Guide Version 2 2 Annual CO concentrations ppm AR4 forcings atmospheric COZ concentrations BERN Model medium source http www ipcec data org ancilliary tar bern txt SStabilisation scenarios 002 450 and C02 S50 Sigmoid function between 352ppm in 1990 and the target concentration in 2100 SStabilisation function shape based on data
58. 92 X PIW_APC p Gm XClt trend cm Fy Xtas _trend cm x hus _ trend 7 m 92500 Pom X hus_trend PomXrsds tpc p 6 mm Table 9 7 Final values for the multiple linear regression coefficients P for each month Climate variables are precipitable water prw mean temperature at surface tas cloud cover clf air column specific humidity at 925 HP hus_92500 and solar radiation rsds Month prw_tpc Tas_trend prw_tpc clt_ave prw_tpc hus_92500_trend hus_92500_trend rsds_tpc tas_trend clt_trend tas_trend Jan 0 8158 0 3472 0 0474 0 0642 0 0302 558 9261 607 9183 0 1705 Feb 0 7709 0 2694 0 0048 0 3657 0 0418 588 6248 564 2603 0 2041 Mar 0 9153 1 1877 0 0368 0 1795 0 0086 994 6107 971 3263 0 0919 Apr 0 9238 0 2712 0 1711 0 2522 0 0566 671 7763 636 5437 0 1546 May 1 0603 0 1438 0 2990 0 3151 0 0036 762 1897 575 5986 0 1726 Jun 1 0133 0 5494 0 3136 0 5790 0 0198 1597 5352 1268 8066 0 8551 Jul 1 0176 0 0242 0 1713 0 7287 0 0034 2122 3183 2320 9253 0 7793 Aug 0 9297 1 3073 0 3551 0 6341 0 0028 1612 0010 1155 6561 0 7550 Sep 0 9173 0 1568 0 1663 0 2789 0 0643 384 0680 315 6471 0 4853 Oct 0 9639 2 0256 0 0366 1 1688 0 0234 774 2827 888 5436 0 6362 Nov 0 9230 1 8634 0 0214 0 1456 0 0265 344 7397 422 3786 0 2942 Dec 0 9806 2 2829 0 0658 0 4819 0 0318 411 5138 446 5273 0 3866 80 Consistent Climate Scenarios User Guide Version 2 2 HUSS v Regressed HUSS
59. 9481 0 0171 0 0367 0 0028 0 0064 Mar 0 9205 0 0267 0 0475 0 0056 0 0126 Apr 0 9751 0 0402 0 0168 0 0229 0 0062 May 1 0215 0 0290 0 0258 0 0157 0 0205 Jun 0 9719 0 1185 0 0321 0 0142 0 0673 Jul 0 9662 0 1829 0 0097 0 0399 0 1136 Aug 0 9115 0 0800 0 0056 0 0409 0 0046 Sep 0 9573 0 0101 0 0216 0 0203 0 0534 Oct 0 8802 0 0199 0 0101 0 0077 0 0530 Nov 0 9045 0 0218 0 0464 0 0058 0 0272 Dec 0 8893 0 0114 0 0384 0 0029 0 0340 73 Department of Science Information Technology and Innovation Figure 9 1 presents a comparison of minimum temperature 7asmin trends to regression derived estimates of the trends for all months for each of the six GCMs Application of Regression Modelling of Tasmin Trends for SIX GCMs Single regression derived from pool of all data 3 2 5 4 2d z Baad x Miroc M o Miroc H g BCCR ae Fee CSIRO Mk3 0 x CSIRO Mk3 5 GISS AOM 0 5 4 0 T T v r o 0 5 1 1 5 2 2 5 3 Tasmin Trend Figure 9 2 Comparison of minimum temperature Tasmin trends x axis to regression derived estimates of the trends y axis for all months for each of the six GCMs Application of Regression Modelling of Tasmax Trends for SIX GCMs Single regression derived from pool of all data 3 y 0 97 44x 0 032 gt gt 2 5 v 2 e gt s Sas ce g x Mire M k we a 5 Miroc H w 1 che se BCCR 5i CSIRO MK3 0 x CSIRO Mk3 5 GISS AOM 0 5 S
60. Clim tool CF Version 1 projections data were released for eight Global Climate Models GCMs in September 2010 followed by CF Version 1 1 projections data for 17 GCMs in April 2011 As at April 30 2012 a considerable amount of CF Version 1 1 projections data had been provided to end users exceeding 650 individual Australian locations 1 3 million data files and 850GB data volume The latest version of the CF projections data V1 2 released in June 2012 is the same as V1 1 except that it now includes data for two highly ranked Hadley Centre GCMs HADCM3 and HADGEM1 and has been adapted for delivery via the Long Paddock website s Climate Change Projections web portal The second phase of the project which incorporated a more sophisticated approach called quantile matching QM see Section 5 supplied by Dr Phil Kokic and Mr Steven Crimp CSIRO was implemented and enhanced by Dr Andrej Panjkov QCCCE QM considers projected changes in the cumulative distribution function of the climate projections Kokic et a 2012 The method used to calculate the 2030 projections data Version QMV2 2 0 released in June 2011 updated to Version QMV3 0 in May 2015 does this by incorporating significant observed trends in 10th 50th and 90th percentile values that have been extrapolated to 2030 QM 2030 projections are available for the same GCMs as the CF data QM 2030 projections data are also available via the Climate Change Projections web p
61. Climate Change Centre of Excellence 2008 in preparation AussieGRASS Environmental Calculator Product Descriptions Queensland Climate Change Centre of Excellence 2009 in preparation AussieGRASS Environmental Calculator amp FORAGE User Guide Rahmstorf S 2008 Anthropogenic Climate Change Revisiting the Facts In Zedillo E PDF Global Warming Looking Beyond Kyoto Brookings Institution Press pp 34 53 http Awww pik potsdam de stefan Publications Book chapters Rahmstorf Zedillo 2008 pdf Rayner D P 2005 Australian synthetic daily Class A pan evaporation Queensland Department of Natural Resources and Mines Technical Report December 2005 from http www longpaddock qgld gov au silo documentation AustralianSyntheticDailyClassAPanEvaporation pdf 101 Department of Science Information Technology and Innovation Reichler T and J Kim 2008 How well do coupled models simulate today s climate Bull American Meteorological Society 89 303 311 Reichler T and J Kim 2008 Uncertainties in the climate mean state of global observations reanalyses and the GFDL climate model Journal of Geophysical Research vol 113 D05106 doi 10 1029 2007JD009278 2008 Ricketts J H 2009 OzClim for the MTSRF region 18th World IMACS MODSIM Congress Cairns Australia 13 17 July 2009 http Awww mssanz org modsim09 Ricketts J H Kokic P N and Carter J O 2011 Estimating trends in monthly ma
62. Department of Science Information Technology and Innovation Consistent Climate Scenarios Data Consistent Climate Scenarios User Guide AR4 Change factor and Quantile matching based climate projections data Grazing Land Systems Science Delivery May 2015 Version 2 2 Department of Science Information Technology and Innovation Prepared by Grazing Land Systems Science Division Department of Science Information Technology and Innovation PO Box 5078 Brisbane QLD 4001 The State of Queensland Department of Science Information Technology and Innovation 2015 The Queensland Government supports and encourages the dissemination and exchange of its information The copyright in this publication is licensed under a Creative Commons Attribution 3 0 Australia CC BY licence Under this licence you are free without having to seek permission from DSITI to use this publication in accordance with the licence terms You must keep intact the copyright notice and attribute the State of Queensland Department of Science Information Technology and Innovation as the source of the publication For more information on this licence visit http creativecommons org licenses by 3 0 au deed en Disclaimer This document has been prepared with all due diligence and care based on the best available information at the time of publication The department holds no responsibility for any errors or omissions wi
63. Generator Version 2 4 Technical Manual Climatic Research Unit UEA Norwich UK 48pp Wigley T M L Richels R amp Edmonds J A 1996 Economic and environmental choices in the stabilization of atmospheric CO concentrations Nature 379 240 243 Wikipedia 2010 Special Report on Emissions Scenarios Retrieved May 24 2010 from http en wikipedia org wiki Special Report on Emissions Scenarios World Meteorological Organization 2008 Guide to Meteorological Instruments and Methods of Observation Appendix 4B WMO No 8 CIMO Guide Geneva 2008 102 Consistent Climate Scenarios User Guide Version 2 2 14 Contact details Feedback about the contents of this document is encouraged and most welcome All comments related to this document including data access and data derivation should be directed to cccs dsitia qld gov au 103 Department of Science Information Technology and Innovation 15 Appendix DAFF Climate Change Research Program Projects e A national research program for climate ready cereals performance of wheat and sorghum under current and future climates CSIRO e Adaptation of a range of wheat types to elevated atmospheric CO concentration University of Melbourne e Developing climate change resilient cropping and mixed cropping grazing businesses in Australia CSIRO e Relocation of intensive crop production systems to northern Australia Costs and opportunities DEEDI Qld
64. Guide for AR5 projections data 96 Consistent Climate Scenarios User Guide Version 2 2 11 10 Summary of differences between CF and QM versioning The differences between CF and QM versioning are summarised in Table 11 2 Table 11 2 Difference between Consistent Climate Scenarios projections data versioning Product Version Rerease Comments date vo Apr 2010 Initial CF 2030 and 2050 projections test data for format checks etc V1 Sep 2010 Eight GCMs eight emissions scenarios three climate sensitivities six climate variables and two projections years 2030 and 2050 Change factor Improvement over Vl including 17 GCMs eight CF data emissions scenarios three climate sensitivities six climate variables and two projections years 2030 and 2050 The These V1 1 Apr 2011 additional GCMs required more infilling for files have trends per degree of 21 Century global no method warming than the V1 set Incorporates improved tag infilling techniques and additional diagnostic information Some improvements on evaporation Historical baseline option 1899 to current This is the version running under the web V1 2 Jun 2012 Historical baseline 1960 2010 Includes two Hadley Centre GCMs HADCM3 and HADGEM1 Otherwise there is no difference in the data from that of V1 1 QMV2 1 0 2030 Jun 2011 Initial 2030 test set QM in filename Quantile Initially 17 GCMs ei
65. ICC applied to the mean annual rate of change for rainfall and temperature per degree of global warming On each plot e change in annual mean rainfall is shown on the x axis e change in annual mean temperature is shown on the y axis e GCM model names are abbreviated i e CCCMA GISS e the mean change from the 1960 to 2010 base period for the aggregate of all seventeen GCMs models is shown by the blue circle e changes in the median from the base period for individual models are shown by red squares e changes in rainfall based on 10 and 90 percentile climate warming sensitivities are indicated by the horizontal barbs e changes in temperature for 10 and 90 percentile climate warming sensitivities are indicated by the vertical barbs 34 Consistent Climate Scenarios User Guide Version 2 2 The comparison of model projections plots are gif files and are typically named LocationCode_Scenario_ClimateSensitvity_Projections Year_Lat_Long_mdlperf_V1 2 File Type e g 014901 _A1B_M_2030 13 8345_131 1872_mdlperf_V1 2 png e LocationCode is a six digit number BoM station code if patched point i e 014901 or all zeros if drilled from interpolated surfaces i e 000000 Scenario emissions scenario A1B Climate Warming sensitivity default M place holder only Projections year 2030 Latitude of the station or location in decimal degrees Longitude as above mdlperf model performance comparison of mode
66. ITIA transferred to DSITI Downscaling The process of transforming numerical output from a coarse to a finer scale i e from GCM model grid point to regional scale EI Ni o EI Ni o represents the warm phase of the El Nifo Southern Oscillation ENSO cycle the opposite of a La Ni a This large scale periodic warming of the central and central east tropical Pacific results in changes in the atmosphere that affect weather patterns across much of the Pacific Basin including Australia During El Ni o episodes the SOI is negative due to lower than average air pressure at Tahiti and higher than average pressure at Darwin El Nino Southern Oscillation ENSO The coupled ocean atmosphere phenomenon that produces year to year oscillations between opposite states of atmospheric pressure and rainfall associated with the large scale warming and cooling of the oceans in the central and east central tropical Pacific ENSO has three phases warm EI Ni o cold La Ni a and neutral Common measures of ENSO are the Southern Oscillation Index SOI and Ni o region ocean temperatures Emissions scenarios Categorised outcomes of greenhouse gas emissions based on potential economic development demographic and technological changes Ensemble A group of realisations runs from a Global Climate Model GCM Expert Review Panel This panel provides expert scientific advice to the Consistent Climate Scenarios project File Transfer Protocol FTP
67. Information Technology and Innovation Figure 8 2 The maps show projected 21 Century changes in maximum temperature for the GCMs clustered in each of the four HI HP WI and WP RFC partitions 4 4 T T T T T T T e MIROC3 2 hires Minimum Temperature Minimum Temperature High global warming and 4 0 a warm Indian Ocean High global warming and a warm Pacific Ocean 4 ECHAMS MPI OM 3 6 F UKMO HadGEM1 HI o Z e p a ae F 1 a Less warming More warming w ap MIROC2 2 medres Less warming More warming Y c UKMO HadCM3 HI C E Sef on oroS CSIRO MKa 5 HP Composite Composite z T83 oe oO p 4 cs CNRM CM3 Minimum Temperature z 2 8 Foos ETET Minimum Temperature 2 G1 0 2 o 7 F MRI CGCM 232 ES i CCSM3 GFDL CM2 1 24 CGCM3 1 Py WP 747 e INM CM3 0 i S mka CSIRO Mk3 0 2 0 GISS AOM gi Smaller global warming Smaller global warming Warmerand a warm Pacific Ocean 1 1 6 1 fi fi fi 1 fi 1 L 1 L L 20 10 0 10 20 30 40 West Pacific East Indian Index and a warm Indian Ocean Calder Less warming More warming Less warming More warming WI Composite WP Composite Figure 8 3 Projected 21 Century changes in minimum temperature for the GCMs clustered in each of the four HI HP WI and WP RFC partitions Table 8 3 Representative Future Climate partitions RFC and associated GCMs
68. Ms that did not require any infilling as trends per degree of global warming were available for all the climate variables needed to produce the change factors used in the calculation of the Consistent Climate Scenarios projections data GCM name used in Consistent Climate Expert Review Panel recommendation Scenarios Project HADCM3 HADGEM1 ae More likely to produce credible projections MIROC M GISS AOM Less likely to be reliable INMCM Not recommended CSIRO MK35 Not assessed but expected to be better than CSIRO Mk3 0 Most of the infilling required to produce the CF projections data used quite complex methods For some GCMs infilling was almost entirely derived from single model runs rather than ensembles The following information describes the infilling methods used for estimating trends in daily maximum and minimum temperature vapour pressure potential evaporation pan evaporation and solar radiation 70 Consistent Climate Scenarios User Guide Version 2 2 9 1 Estimating trends in daily maximum and minimum temperature Trends per degree of global warming for maximum and minimum temperature Tasmax and Tasmin were only available from OzClim for seven GCMs Table 9 2 Trends for these two climate variables have been estimated for all of the other GCMs Table 9 3 Table 9 2 GCMs that did not require any infilling for trends per degree of global warming for daily maximum and minimu
69. Oo oOo O O O co a ao q N ap bod WO oO a2 ap oOo oa Oo oO Oo oO o Oo qm N N N N N N N N N Year 2090 2100 Figure 7 1 Historical and projected CO concentrations for the emissions scenarios used in the CCSP data source http www ipcc data org ancilliary tar bern txt 59 Department of Science Information Technology and Innovation For all GCMs the experiments in the CMIP3 database do not cover the full range of emissions scenarios which were considered by the IPCC in its Special Report on Emissions Scenarios SRES The suite of model runs in CMIP3 are forced by a limited number of SRES emissions scenarios including SRES A1B a slightly smaller set from SRES A2 and a still smaller set from SRES B1 The A1FI high emissions scenario which is increasingly being adopted in policy development was not run for CMIP3 The experiments also did not include a desirable level of replication The IPCC employed a relatively straight forward methodology to extrapolate AR4 model outputs across all SRES scenarios and to quantify the sensitivity of the warming response to a particular emissions scenario This method involved calculating projected global warming for the eight SRES emissions scenarios using a single climate model MAGICC For each emissions scenario the IPCC provides three warming responses low moderate and high representing the range of sensitivity of global temperature rise to each emissions sc
70. S data have been estimated using e Coupled Model Intercomparison Research Program 3 CMIP3 patterns of change data projected changes per degree of 21st Century global warming supplied by the CSIRO and the UK Met Office Hadley Centre and e data from AR4 SRES scenario temperature response curves projected amounts of global warming supplied by the CSIRO As such the following data sources should also be acknowledged e The CMIP3 global model database http www pcmdi In gov ipcc about_ipcc php e O2zClim http www csiro au ozclim e UK Met Office Hadley Centre http www metoffice gov uk climate change resources hadley Related publications Further information describing the infilling of trends per degree of global warming for missing climate variables is documented in Ricketts J H Kokic P N and Carter J O 2011 Estimating trends in monthly maximum and minimum temperatures in GCMs for which these data are not archived Queensland Climate Change Centre of Excellence Queensland Government CSIRO Mathematics Informatics and Statistics 19 International Congress on Modelling and Simulation MODSIM Perth Australia 12 16 December 2011 http www mssanz org au modsim201 1 F5 ricketts pdf Further information describing the quantile matching approach is documented in Kokic P Jin H and Crimp S 2012 Statistical Forecasts of Observational Climate Data Extended Abstract International conference o
71. SIM or p51 for GRASP Data contained in the QM 2050 projections data files are formatted the same as those in the QM projections data files Note applies in addition to caveats associated with the QM 2030 data e The QM projections data for 2050 are available for a single GCM ECHAM5 one emissions scenario A1B and one climate warming sensitivity median Each file contains projections data for six climate variables rainfall maximum and minimum temperature projections are based on the QM 2050 methodology see Section 5 and vapour pressure evaporation and solar radiation projections are based on the QM 2030 methodology ECHAM5 is unique in that it is the only GCM that has both a high rank as assessed by the Expert Review Panel and a complete set of raw GCM daily data from 1900 to 2100 16 Consistent Climate Scenarios User Guide Version 2 2 2 3 Historical baseline climate data files Along with projections data the ZIP archive also contains files of historical climate data for baseline comparison The historical data have been extracted from the SILO database The SILO historical database as it currently exists is relied upon by the scientific community across Australia and provides researchers and modellers with seamless spatially and temporally complete Australia wide daily climate data from 1889 to current CCS historical baseline climate datasets are available from 1960 onwards since this is the period
72. SRES scenarios Further reading related to aspects of climate warming sensitivity is available in Wigley 2000 as well as https wiki csiro au confluence display ozclim Science Science Development http www nap edu catalog 12877 html http www ipcc ch pdf assessment report ar4 syr ar4_syr_spm pdf http en wikipedia org wiki Climate_sensitivity 63 Department of Science Information Technology and Innovation 8 Global Climate Models To better inform users about the use of the CF datasets and dataset selection this section provides users with brief background information about the range of Global Climate Models GCMs that we have used including recommendations to assist users on GCM selection Table 8 1 indicates the full set of CMIP3 GCM model runs that were available to the CCSP to consider including climate forcings and spatial resolution of which 19 have been selected for use The CSIRO Mk3 5 GCM hasn t been assessed with a rank in terms of skill by the CCSP Expert Review Panel However it is expected to be an improvement over CSIRO Mk3 0 Detailed information on all available IPCC AR4 listed GCMs is documented on the PCMDI website at http www pemdi lIn gov ipcc about_ipcc php Table 8 1 IPCC AR4 GCM characteristics including their official PCMDI CMIP3 name number of ensemble members specific climate forcings spatial resolution and rank as assessed by the Consistent Climate Scenarios Project Expert
73. Version 1 0 projections data provided in September 2010 included projections data from only eight models for which infilling of any trend per degree of global warming data has been reasonably straight forward By April 2011 projections data from nine additional GCMs had been progressively infilled enabling the 65 Department of Science Information Technology and Innovation production of the CF Version 1 1 data As at June 2012 the latest CF Version 1 2 and QM 2030 datasets are available from a total of nineteen GCMs Detail on the infilling methodology is discussed in Section 9 with the infilling status for each model summarised in Table 9 10 8 1 Selecting Global Climate Models The issue of Global Climate Model GCM selection is an important one and various views exist as to the best approach to take It is desirable that end users can relate their results based on their choice of GCMs to the results from another user who has chosen a different set of GCMs particularly if the results differ A DAFF sponsored workshop was held in June 2011 where participants were asked to consider how to ensure comparability between projects via use of GCMs linked to a common set of climate change scenarios The workshop agreed that a common approach to selecting model projections should be one based on the work of Watterson 2011 Watterson s paper describes how projected Australian 21t Century rainfall responses cluster for
74. across all SRES scenarios and to quantify the sensitivity of the warming response to a particular emissions scenario This method involves the use of a coupled gas climate model called MAGICC Model for the Assessment of Greenhouse gas Induced Climate Change Wigley 2000 which uses emissions scenarios for greenhouse gases reactive gases and sulphur dioxide to calculate projected global warming for the eight SRES emissions scenarios For each emissions scenario the IPCC provides low moderate and high warming responses equivalent to the 10 50 and 90 percentile projections representing the range of sensitivity of global temperature rise to each emissions scenario 4 3 Calculation of change factors The calculation of change factors for individual GCMs is relatively straight forward CSIRO employ a pattern scaling approach as used in OzClim described briefly below to calculate for each surface grid point of a GCM the projected change over the 21 Century in a given climate variable per degree global warming This is termed the pattern of change Patterns of change differ according to specific emissions scenarios climate warming sensitivities and GCM runs Whereas temperature change per degree global warming is expressed in absolute terms i e degrees Celsius change per degree global warming projected change in other climate variables e g rainfall evaporation solar radiation and vapour pressure is e
75. al Fossil 449 555 Represents the most population that peaks around 2050 and intensive extreme global warming risk declines thereafter Rapid introduction analysed to date of new and more efficient technologies f Observations suggest A1FI Satta a most closely represents the RCP8 5 541 ppm current trend in global CO2 ARS emissions Only a few runs have been made Not available through PCMDI Data obtained by pattern scaling 3 2 A1B This model has the most Balance 447 522 variables Submitted to IPCC across all E for PCMDI Data obtained sources 8 by pattern scaling iva A1T Not available through Emphasis on 435 496 PCMDI Data obtained by non fossil pattern scaling sources Same value as RCP4 5 A2 Preferred alternative to Self reliance and preservation of local 444 522 A1FI Similar to A1Fl for the identities Continuously increasing early 21 Century Submitted global population Economic to IPCC for PCMDI but not development regionally oriented Per as complete as A1B Data capita economic growth and obtained by pattern scaling technological change more fragmented and slower than for other storylines B1 Not available through A convergent world with the same Introduction of 432 482 PCMDI Data obtained by global population that peaks around clean and pattern scaling 2050 and declines thereafter Rapid resource changes in economic structures toward efficient a service and information economy with reductions in
76. anged from the source historical data Days without rain in the historical time series are projected to days without rain The SILO climate database http www longpaddock gld gov au silo can be used to assist in the selection and identification of climate stations 10 Consistent Climate Scenarios User Guide Version 2 2 e SILO daily climate are checked for quality and are constantly updated at least twice a year hence changes to some data may occur Any changes in the SILO base line climate data will affect the projected 2030 and 2050 data e The same disclaimers that apply to SILO historical data apply to the projections data e The daily dates presented in the CF projections files are the dates from which the 2030 or 2050 projections data are drawn from using change factor methodology These dates are essentially an ensemble of individual 2030 or 2050 years For example if 19600101 were to be used to represent the first instance of 20300101 then 19610101 would represent the second instance of 20300101 and 20100101would represent the 51st instance of 20300101 The use of historical dates creates the practical advantage of being able to run an analysis in a single pass using a single climate data input file e Inthe CF data at least 51 instances of data are available if using source data from 1960 2010 Source data prior to 1960 are less useful due to issues related to low climate station density and uncorrected c
77. apour pressure See Figure 9 5 PERCENT CHANGE IN VAPOUR PRESSURE AS A FUNCTION OF PERCENT CHANGE IN SPECIFIC HUMIDITY 60 directly computed Pressure 980 hPa linear estimate Pressure 980 hPa directly computed Pressure 1013 25 hPa linear estimate Pressure 1013 25 hPa 10 directly computed Pressure 1040 hPa linear estimate Pressure 1040 hPa y x 20 Relative Change in Vapour Pressure 40 Curves for all pressures are nearly indistinguishable 60 60 40 20 0 20 10 60 Relative Change in Specific Humidity Figure 9 5 Percentage changes in vapour pressure y axis as a function of changes in specific humidity x axis showing the small impact of changes in atmospheric surface pressure 77 Department of Science Information Technology and Innovation Method used in CF Version 1 1 and CF Version 1 2 projections In the Consistent Climate Scenarios project the climate variable screen specific humidity huss is an important input into calculation of vapour pressure deficit This climate variable was not available for a number of GCMs of importance in fact only four of the ten most preferred GCMs and hence it becomes necessary to estimate trends Butler 1998 gives the following relationship between precipitable water and water m Po vapour partial pressure h pikTy_ Where his precipitable water m H p k are constants Po is water vapour partial pressure at the
78. asets future CDFs for rainfall and temperature variables have been computed differently using daily data obtained from GCMs instead of extrapolating historical quantile trends As with CF data QM 2030 projections have been computed for 19 GCMs eight emissions scenarios three climate sensitivities and six climate variables However due to limited daily GCM data QM 2050 projections have only been prepared for a single GCM ECHAM 5 emissions scenario A1B and climate sensitivity median While the approach for 2050 rainfall and temperature projections uses daily GCM data QM 2050 projections for other climate variables use the QM 2030 method extended to 2050 discussed further in Section 5 2 5 1 Steps involved to calculate QM projections data for 2030 Step 1 Compute the quantile ranking of a climate variable for each historical day in its containing month based on a 1957 to 2010 training period Step 2 To construct a plausible future CDF to represent future variability project the historical trends in three or more pivot quantiles usually 0 1 0 5 and 0 9 out to the target projection year usually 2030 These historical trends Figure 5 1 are not used if their computed statistical significance exceeds a cut off threshold currently p statistic 0 1 Interpolation and use of conservative extrapolation methods has been used to estimate the future CDF at quantiles other than the pivot quantiles This process is carried out f
79. at change could be expected around the year 2030 assuming a worst case emissions scenario A1Fl and high climate sensitivity to global warming However they are only interested in the projections from a few five GCMs which are recommended as being relatively good performers They would also like to compare the differences based on the change factor CF and quantile matched QM methods The end user will first need to place a data order based on the above mentioned variables via the Long Paddock website s Climate Change Projections web portal http www longpaddock gld gov au climateprojections Once the projections data have been processed the end user will receive an email containing an ftp link for collection of the data In most cases ZIP archives containing the data will be ready for collection within 2 hours may take longer for large orders When projections data without diagnostics are requested there will be one ZIP archive per order In this example there would be two sets of ZIP archives one for CF data and one for QM data If the user were to open either the CF or QM ZIP archive the user will have access to many files For example in the CF ZIP archive the CF projections data files refer to the GCMs emissions scenarios target years climate sensitivity values and output formats that were selected during the order process In this case the user will have received for each of the five GCMs just those files
80. ate of change Change factor CF approach An OzClim based statistical approach whereby monthly climate change factors are used to scale historical climate data sets to produce daily climate projections for 2030 and 2050 The change factors have been derived for a range of GCMs emissions scenarios and climate sensitivities Climate Change Research Program CCRP The DAFF Climate Change Research Program funds research projects and on farm demonstrations to help prepare Australia s primary industries for climate change and build the resilience of our agricultural sector into the future Climate warming sensitivity A simple measure of the strength of the effect of CO2 concentrations on climate particularly global temperature In this project this term is expressed in terms of the uncertainty spread of the amount of global warming under a specified SRES emissions scenario at future point in time i e low 10 percentile median 50 percentile or high 90 percentile Commonwealth Scientific and Industrial Research Organisation CSIRO Australia s national science agency Consistent Climate Scenarios Project CCSP A DAFF funded program formed to develop a consistent set of synthetic climate projections data across Australia for use in biophysical models which maintain weather like properties and also account for uncertainties and biases in climate change projections as well as different methods
81. ate variables expressed as a percentage of the total distribution for that variable The CF based frequency distribution plots are not provided with the CF V1 2 datasets via the web Climate Change Projections web portal as not all users will request the A1B scenario with their data order However plots are for specific locations in Australia and can be made available on request for 17 GCMs the A1B emissions scenario high climate warming sensitivity and six climate variables The climate variables are e rainfall e maximum and minimum temperature e solar radiation e vapour pressure e pan evaporation The frequency distribution plots are gif files and are typically named ModelName_Scenario_ClimateSensitvity_Projections Year_LocationCode_Lat_Long_fdist_ v1 1 gif e g CSIRO Mk35_A1B_H_2050_ 051039 31 5495 147 1961 _fdist_v1 1 gif ModelName BCCR CCCMA 47 CCCMA 63 CNRM CSIRO MK30 CSIRO MK35 ECHAMS ECHO G GFDL 20 GFDL 21 GISS AOM HADCM3 HADGEM1 IAP FGOALS INMCM MIROC H MIROC M MRI GCM232 NCAR CCSM e Scenario emissions scenario A1B e Climate Warming sensitivity high H refers to the 90 percentile value e Projections year 2050 e LocationCode is a six digit number BoM station code if patched point i e 051039 or all zeros if drilled from interpolated surfaces i e 000000 e Latitude of the station or location in decimal degrees e Longitude as above e fdist frequency di
82. ation and specific humidity projections data and the 2050 solar radiation and specific humidity data and the projected standard deviation shown in the plot will differ from that of the historical baseline data e Where observed quantile trends are not significant the climate projections method defaults to the change factor approach for the above mentioned climate variables and the projected standard deviation shown in the plot will be the same as that of the historical baseline data 42 Consistent Climate Scenarios User Guide Version 2 2 3 11 Transient climate data test set for 1889 2100 A test set of transient climate change files and background documentation are located in a zip archive named transient_data_test_set zip at ftp climate mft derm gqid gov au Climate Scenarios Transient TestData The transient climate change files contained within the zip archive are named as follows e g 035149 24 8353_ 149 8003 NCAR CCSM_CO2 550_med_CF1 2EM09 wth e Station number also called Location Code 035194 e Latitude of the station in decimal degrees 24 8353 South e Longitude of the station 149 8003 East e GCM Model Name 17 are available i e NCAR_CCSM e SRES emissions scenario eight are available i e CO2 550 e Climate warming sensitivity i e low med high referring to the 10 50 and 90 percentile values respectively e Projections Method Code and Version CF 1 2 e Ensemble Code EM e En
83. by the residuals using a second qauntile matching pass Step 5 Lastly the entire projected data stream is renormalised to conform to the mean value indicated by the OzClim trend coefficient change factor associated with a specific GCM emissions scenario climate warming sensitivity The resulting QM output should honour the mean implied by the GCM and approximately the spread of values implied by the new target CDF 5 2 Variation of methodology for calculating 2050 QM projections data As discussed to acquire 2050 projections datasets future CDFs for rainfall and temperature variables have been computed using daily data obtained directly from GCMs rather than by extrapolating historical quantile trends Step 1 To start with pivot quantiles for 2050 for rainfall maximum and minimum temperature are computed and saved for each point in the GCM data grid This grid of GCM derived 2050 pivot quantiles is then interpolated using bilinear interpolation so that pivot quantiles can be obtained for each station location 52 Consistent Climate Scenarios User Guide Version 2 2 Step 2 The 2050 CDF is then computed from the station specific pivot quantiles and interpolation and conservative extrapolation methods are applied so that the 2050 CDF can be estimated at other quantiles besides the pivot quantiles This interpolation process is carried out for most of the climate variables but some of the climate variables wil
84. c ciesin columbia edu ddc ar5 scenario _process RCPs html The information contained in the decadal CO concentrations file is shown in Figure 3 3 and Figure 3 4 Notes e Projected 2030 CO concentrations for A1FI 449ppm are equivalent to RCP 8 5 A1T 435ppm are equivalent to RCP 4 5 e Projected 2050 CO concentrations for A1FI 555ppm slightly exceed RCP 8 5 541ppm e Users should note that AR3 and AR4 based COz concentration data is derived from the BERN model Some GCM models used CO from the ISAM model for their atmospheric forcing The difference between the ISAM and BERN models for the year 2050 for each SRES scenario is about 10ppm which is less than the difference between the high and low versions of each model Since more GCM models use CO derived from the BERN model we supply this estimate with the climate data e The preliminary IPCC AR3 COz estimates two older scenarios IS92A and IS92A SAR and the new RCP CO z data are for reference there are no corresponding CF or QM climate files 25 Department of Science Information Technology and Innovation Decadal CO concentrations ppm AR4 forcings atmospheric CO2 concentrations BERN Model medium source http www ipcc data org ancilliary tar bern txt Stabilisation scenarios CO02 450 and C02 550 Sigmoid function between 352ppm in 1990 and the target concentration in 2100 Stabilisation function shape based on data scanned from
85. cationCode_Scenario_Projections Year_Climate Warming sensitivity_ModelName_Latitude_Longitude_VersionNumber SILOformat e g 051039 A1FI_2030_M_CSIRO MK35_ 31 5495 147 1961_QMv3 0 met e LocationCode is a six digit number BoM station code if patched point i e 051039 or all Zeros if drilled from interpolated surfaces i e 000000 e Scenario emissions scenario i e A1B A1FI etc e Projections year 2030 e Climate warming sensitivity rate of global warming i e L M H L M and H refer to the 10 50 and 90 percentile values respectively e Model Name i e CSIRO MK35 HADGEM1 HI HP etc e Latitude and longitude of the station or location in decimal degrees e Version Number where QMv2 2 or QMv3 0 represents QM data e SILO format either met for APSIM or p51 for GRASP Data contained in the QM 2030 projections data files are formatted the same as those in the CF projections data files Figures 2 3 and 2 4 However metadata provided in QM projections data files includes an additional column containing reference codes for QM synthesis methods 14 Consistent Climate Scenarios User Guide Version 2 2 Notes where different from caveats in Section 2 1 associated with the CF data e The QM projections use a 1957 2010 training period to compute the perturbation rules that are applied to that historical baseline The tav annual average ambient temperature and amp annual ampl
86. cations 19 GCMs 3 emissions scenarios 3 climate sensitivities 322 MB 2470 files 1710 projections files 190 SILO files 190 multiplier files 190 CO matching files 190 log files CF 2030 or 2050 projections with diagnostic plots 0 7 MB 7 files Includes the same files as those above plus 1 comparison of model projections plot 1 time series plot 6 7 MB 70 files Includes the same files as those above plus 10 comparison of model projections plots 10 time series plots 324 MB 2490 files Includes the same files as those above plus 10 comparison of model projections plots 10 time series plots QM 2030 projections 0 5 MB 4 files 1 projections file 1 SILO file 1 multiplier file 1 CO matching file 5 3 MB 40 files 10 projections files 10 SILO files 10 multiplier files 10 CO matching files 369 MB 2280 files 1710 projections files 190 SILO files 190 multiplier files 190 CO matching files QM 2030 projections with diagnostic plots 5 9 MB 17 files Includes the same files as those above plus 1 comparison of model projections plot 1 time series plot 5 quantile trend plots 6 QM histogram plots 59 MB 170 files Includes the same files as those above plus 10 comparison of model projections plots 10 time series plots 50 quantile trend plots 60 QM histogram plots 443 MB 2410 files Includes the same files as those above
87. change change per degree global warming to global warming projections from MAGICC for each of the eight SRES emissions scenarios that part of the equation shown in brackets in Figure 4 2 Step 7 Apply change factors for specific locations to a suitable 20th Century baseline climatology to produce projections for a given climate variable Figure 4 2 In OzClim change factors are applied to a 1975 to 2004 baseline climate CSIRO 2010 Monthly climate change factors for individual locations have been calculated for seventeen individual GCMs see Section 6 using amounts of global warming at 2030 and 2050 for eight emissions scenarios A1Fl A1B A1T A2 B1 B2 C02 450 and C02 550 and three climate warming sensitivities low median and high The 47 Department of Science Information Technology and Innovation patterns of change and amounts of global warming from which the climate change factors can be calculated are contained in the monthly multiplier files described in Section 3 1 provided with the projections data Further details related to the method used to produce OzClim patterns of change are available at http www csiro au ozclim and documented by Mitchell et a 1999 Page and Jones 2001 Michell 2003 Whetton et a 2005 Ricketts 2009 and Ricketts and Page 2007 Global Temperature C Global Temperature C Repeat for each gridpoint 4 Point Temperatu
88. climate warming sensitivities The stabilisation of CO will result in reduced changes in rainfall temperature and soil moisture Dai Wigley Meehl and Washington 2001 Changes in average wind speed which may be an important for potential evaporation Heat stress on growing points especially close to ground surface The effect of fire and flood on insects pathogens plant competition e g weeds trees relative to grasses The responses for these may be unknown Insufficient data on the impacts of extremes to correctly parameterise models In some locations there may only be data for a single extreme event i e damaging frost We recommend coupling the OzClim change factor based projections data in the context of the broader range of current climate change risk assessment information 93 Department of Science Information Technology and Innovation 11 Differences between CF and QM projections data including versioning Users should note that there are differences between the CF and QM climate projections data The following discussion including Tables 11 1 and 11 2 outline these differences 11 1 GCMs emissions scenarios climate sensitivities and projections years Differences between CF and QM Versions related to available GCMs emissions scenarios climate sensitivities and projections years are listed in Table 11 1 The range of GCMs emissions scenarios and climate sensitivities are described in Sections 7
89. cribing these emissions scenarios in Section 7 1 In addition to the SRES scenarios the CO concentrations files include two stabilisation scenarios CO2 450 and CO2 550 based on the work of Wigley et al 1996 and the CCS web portal includes these as options when ordering data The stabilisation scenarios examine the implications of stabilising CO at 2100 at various concentrations The CO concentration files also contain the preliminary IPCC AR3 CO estimates headed with the subscript p and two older scenarios IS92A and IS92A SAR Users should be aware that projected futures CO concentration pathways are uncertain and may undershoot or overshoot the proposed trajectories 24 Consistent Climate Scenarios User Guide Version 2 2 Furthermore the CO concentration files also include CO data for four scenarios based on Representative Concentration Pathways RCP 3 PD RCP 4 5 RCP6 0 and RCP8 5 These RCPs that have been determined by projected radiative forcing and have been used for the development of information for the IPCC Fifth Assessment Report AR5 Further documentation about these RCPs is available from e the IPCC Expert Meeting Report on New Scenarios Noordwijkerhout report http www aimes ucar edu docs IPCC meetingreport final padf e the Representative Concentration Pathways RCPs Draft Handshake http www aimes ucar edu docs RCP_handshake paf e the IPCC website at http seda
90. cumented in e Ricketts J H 2011 Estimating trends in monthly maximum and minimum temperatures in GCMs for which these data are not archived MODSIM Congress Perth Australia 12 16 December 2011 75 Department of Science Information Technology and Innovation 9 2 Estimating vapour pressure Trends in vapour pressure per degree of global warming are not directly available from OzClim Although trends in relative humidity are available from OzClim it is difficult to estimate changes in vapour pressure due to corresponding changes in relative humidity without knowing the simultaneous 9am temperature Instead CSIRO has calculated trends per degree of global warming for surface specific humidity for this project Therefore changes in specific humidity have been used to compute trends in vapour pressure per degree of global warming Method used in CF Version 1 projections In CF Version 1 projections trends in vapour pressure had been estimated using standard WMO functions World Meteorological Organization Geneva 2008 relating vapour pressure to specific humidity and atmospheric pressure 7 SH xP 0 62198 x 1 SH SH where VP vapour pressure in hPa SH specific humidity in g water g moist air P surface atmospheric pressure in hPa The constant 0 62198 is the WMO value for the ratio of the molecular weight of water to the average molecular weight of dry air As Figure 9 4 shows over the range of specif
91. dal rather than short time sequences will allow users to better investigate impacts of year to year and decade to decade variability in rainfall and other climate elements e The period from 1960 to 2010 encompasses natural climate variability i e droughts and floods due to fluctuations in ENSO as well as opposite phases of the IPO For example IPO phase changes from Cool to Warm in 1978 and from Warm to Cool in 1999 strong El Nifio s in 1965 1972 73 1982 83 1991 92 1997 98 and 2009 10 and strong La Nifia s in 1973 74 1975 76 and 1988 89 based on the NOAA Oceanic Ni o Index NOAA 2011 e The mean and shape of the probability distribution will vary according to the base period used e We recommend the use of patched point data sets rather than interpolated gridded data where available Interpolated data although useful may have low accuracy in isolated station poor regions especially in the pre 1957 period due to the low density Australian climate station network until then e The post 1957 interpolated data is more accurate than pre 1957 interpolated data QCCCE 2009 in preparation Un detected bad data can have a large effect on the interpolated data especially in data sparse areas QCCCE 2008 in preparation e The use of the post 1957 data is consistent with the base period that is used in the QM approach e Users may wish to further subsample or randomise the 2030 205
92. delName_Latitude_Longitude_VersionNumber SILOformat CO e g 051039 A2 2030 L_CSIRO MK35_ 35 5495 147 1961_V1 2 met 444 00 e LocationCode is a six digit number BoM station code if patched point i e 051039 or all Zeros if drilled i e 000000 e Scenario listed for each emissions scenario i e A1B A1FI etc e Projections year i e 2030 or 2050 e Climate warming sensitivity rate of global warming i e L M H e L M and H refer to the 10 50 and 90 percentile values respectively e Model Name listed for each of 19 GCMs and four RFCs i e CSIRO MK35 HADGEM1 HI HP etc e Latitude and longitude of the station or location in decimal degrees e Version Number V1 2 represents CF data e SILO format either met for APSIM or p51 for GRASP e CO concentration for the listed climate projections year and climate warming sensitivity to three decimal places 23 Department of Science Information Technology and Innovation BM 056002_ 30 5167_151 6681_NamesList txt Notepad File Edit Format Yiew Help 056002_A1T_2030_H_BCCR_ 30 5167_151 6681_v1 2 apsim 437 857 056002_A1T_2050_H_BCCR_ 30 5167_151 6681_v1 2 apsim 498 864 056002_A1T_2030_M_BCCR_ 30 5167_151 6681_v1 2 apsim 437 857 056002_A1T_2050_M_BCCR_ 30 5167_151 6681_v1 2 apsim 498 864 056002_A1T_2030_L_BCCR_ 30 5167_151 6681_v1 2 apsim 437 857 056002_A1T_2050_L_BCCR_ 30 5167_151 6681_v1 2 apsim 498 864 056002_A2_203
93. distributions include quantile matched 2030 or 2050 projections data in comparison to an observed 1957 2010 baseline climate Months where no significant quantile trends are computed will be apparent in the histogram plots as a simple horizontal translation between the historical black shaded and quantile matched orange shaded data due to OzClim change factor based trends only The means u and sample standard deviations Sn of the historical and quantile matched data are shown on each plot Users should note that the plotting scales may differ between stations The histograms are png files and are typically named Climate Variable_LocationCode_PivotQuantiles_1957_2010_Projections Year_Histograms_ LocationCode_ModelName_Scenario_Climate Warming Sensitivity png e g T Min_040428 10 50 90 1957 2010 2030 Histograms 040428 HADCM3_A 1Fl_high png e ClimateVariable Radn solar radiation transformed by ground level proportion of extra terrestrial then logit transform used Rain rainfall not transformed RainCubeRoot rainfall transformed by cube root VP transformed to specific humidity then back to vapour pressure T Max maximum temperature no transform T Min minimum temperature no transform e LocationCode is a six digit number BoM station code if patched point i e 002012 or all Zeros if drilled from interpolated surfaces i e 000000 e PivotQuantiles 10 50 90 denot
94. e CLAMPED will be listed in a log warning file if the application of the change factor obtained from the pattern scaling produces more than a 90 decline in projected rainfall from the historical baseline climate In this case the rainfall projections data are clamped at 10 of the baseline climatology values to avoid occurrence of negative rainfall Information about limitations related to the capture of anomalous data in the Log Warning files is presented in Section 10 2 29 Department of Science Information Technology and Innovation Notes e Users should note that the precision of the values listed in the Log warning files is for calculation purposes only and will not occur in reality e nsome cases usually individual days anomalous values may occur These values may be derived from one of three sources which are 1 the raw data 2 interpolation or 3 the modification to climate changed data 3 5 Historical time series plots Historical time series plots for the six climate variables used in the CCS project have been provided as part of the user information framework The plots provide users with representations for specific locations showing the historical annual variability as well as a longer term trend i e rising mean annual temperature The plots are location specific and show the daily average for each year as an annual time series extracted from historical SILO climate data Although
95. e User Guide describes the methodology employed by CSIRO to prepare the OzClim climate change factors CSIRO 2010 The section also outlines how climate change factors are applied to historical data to produce climate estimates for the future The climate change factors used to calculate CCS data have been estimated using e Coupled Model Intercomparison Research Program 3 CMIP3 patterns of change data projected changes per degree of 21st Century global warming supplied by the CSIRO and the UK Met Office Hadley Centre and e data from SRES scenario temperature response curves projected amounts of global warming supplied by the CSIRO 4 1 Change factor definition The term climate change factor refers to the change in the climatological mean of a specific climate variable e g temperature between the current climate defined in terms of a suitable 20 Century base period and a projected time in the future for example the 30 years centred on 2050 Calculated on a monthly basis change factors may be used as scalars to transform historical daily climate time series to produce time series of projected future climate for use in biophysical models Climate change factors have been calculated using the OzClim methodology and applied to SILO historical baseline climate data to produce the Consistent Climate Scenarios CF Version 1 Version 1 1 and Version 1 2 projections datasets for 2030 and 2050
96. e based on output from individual GCM runs which have not been re scaled or normalised in any way and are not based on the Consistent Climate Scenarios projections data Snapshots of information contained in continental area weighted rainfall and temperature and global mean temperature files are presented in Figures 3 9 3 10 and 3 11 respectively On the Australian continental area weighted rainfall and mean temperature graphs the observed annual data Obs shown by the purple line and GCM simulated data 20C shown by the blue line cover the 20 Century The yellow and dark green regression lines represent the observed and simulated 20 Century trends respectively 21 century GCM projections 21C are shown by the red annual variability and light blue 100 year trend lines On the observed global mean temperature graphs the annual air temperature Obs GISS is shown by the red line and covers the time span of measured data to present The GCM simulated data 20C Model shown by the blue line cover the 20 century 21 Century GCM projections 21C Model A1B are shown by the green line 36 Consistent Climate Scenarios User Guide Version 2 2 Statistics presented on the graphs include e slope 20 and 21 Century positive for an increase negative for a decrease e Pr probability of rainfall or temperature trend being zero lt 0 05 is significant e mean 20 and 21 Century mean annual rainfall or te
97. e period data can make a difference in the statistics of the climate variables This is because climate fluctuates over time particularly due to the influence of annual and decadal climate drivers i e El Ni o Southern Oscillation ENSO and the Interdecadal Pacific Oscillation IPO and also due to longer term natural and anthropogenic climate influences The compilation of statistics using a relatively long base period is more likely to capture these trends The examples presented in Figure 10 1 and Table 10 1 show how the University of Queensland Gatton historical climate data and CF projections data can differ according to the selected historical base period July mean daily minimum air temperature University of Queensland Gatton Historical Historical 1899 2010 mean id ha h a ea TEE UE N AET a ana area i ie T A T T P LA E ow e e OLDO e MH OW HY emn OOD fe Be OM oo oo fF fe AN MOOmMO tT FH OOnKRe Ke DOD DWH SO FS naoanononenre mr ss FOF 97 DB FH OFDM FHF SS ee NAN Year Figure 10 1 Graph showing how the University of Queensland Gatton Location Code 040082 CF Version 1 1 projections data differ for 2050 depending on the historical base period that is selected The SILO historical data black are overlaid with Consistent Climate Scenarios CF Version 1 1 projections data green for July 2050 The projections data use MIROC H A1FI and high climate warming sensitivity Data will differ for other
98. e resulting fitted equations and the equation with the highest r were recorded Note that differing months yielded somewhat differing results and hence the following regression model was tested using multiple linear regression 79 Department of Science Information Technology and Innovation FUSS 6 Pm X PTW APC 4 Gm Pym X tas _trend cm Ps x dist P mX PY IPC Gm X dist Py m X PIW_IPC Gm X tas _trend cm 92500 P xhus_ave x tas _ trend x dist P X lt ave cmx tas _ p G m R mX PYW_tPC Gm clt _ trend Gm F m X tas _trend 6 m 92500 Pom X hus _trend Fy PYW_IPC Gm Pr tpe Fy xrsds_tpe p G m The CCCMA 47 CGCM3 1 T47 model was added to the above GCMs using estimated RSDS tpcs since rsds was not supplied for this model Huss trends x hus _ tren trend ee ger p Gim p G m associated with the MRI GCM232 GCM were totally non correlated with any of the predictands contrary to all other models so this model was omitted As in Section 9 1 multiple linear regression was then performed using the chosen model The regression statistics Table 9 7 indicated that parameters 3 4 6 and 11 were not different from zero so the final regression used was P l m P 5 m P 8 m Huss X PrW_IPC p Gm Py Xtas _ trend p Gwm P 7m p G m X PW _IPC Gm X tas _trend xclt __ave Gm Xtas _trend 500
99. ear trend for the selected base period in this case 1960 2010 is shown in blue Historical time series plots are not produced for periods of less than six years Notes e Pan evaporation data from 1970 onwards are from daily class A pan evaporation measurements Prior to this the data is synthetic pan e Linear trend lines showing the selected long term trend are not plotted for periods of 30 years or less e The historical time series plots are not produced for periods of less than six years 31 Department of Science Information Technology and Innovation 3 6 CF Frequency distribution plots CF v1 1 based frequency distribution plots based on the A1B emissions scenario for six climate variables were initially provided as test data as part of the user information framework for the Consistent Climate Scenarios project The plots provide users with a visual impression of historical and projected frequency distribution changes Initial inspection of the plots shows that the output is as expected i e increased temperatures These plots are not currently available through the CCS web portal see Section14 for contact details The plots are designed to inform users of the change in frequency distributions shown on the y axis for both the observed 1960 2009 baseline and the projected 2050 climate data The frequency distribution plots provide users with an analysis of the occurrence of discrete values of specified clim
100. ee yg HE 28 pau ae o T T T o 0 5 1 1 5 2 2 5 3 Tasmax Trend Figure 9 3 Comparison of maximum temperature Tasmax trends x axis to regression derived estimates of the trends y axis for all months for each of the six GCMs 74 Consistent Climate Scenarios User Guide Version 2 2 A comparison with the ratios method was performed to demonstrate improvements over the ratios methods Figure 9 2 and Figure 9 3 This comparison shows that a there was a significant reduction in scatter for the regression method compared to the ratios method and b the ratios method had a lower regression coefficient for maximum temperature The resulting gridded data were checked to ensure the computed tmax and tmin Tasmax Tasmin values correctly bracketed the original tmean tas values and range checked Tasmax Trends vs Pooled ratio estimates for Six GCMs y 0 TA 68s So gt o Miroc H Estimated Tas max Trend an X BCCR 1 E Z e S CSIRO Mk3 0 CSIRO Mk3 5 GISS 40M ia os 1 1 5 2 25 3 Tasmax Trend Figure 9 3 Comparison of maximum temperature Tasmax trends x axis to ratios method derived estimates of the trends for all months y axis for each of the six GCMs r 0 8987 Note the slope is lower than in Figure 9 2 and there is increased scatter Further information describing the infilling of trends per degree of global warming for missing climate variables has been do
101. elected climate variables only with QM orders Step 5 Confirm order Your order is summarised with an option to submit or revise it e Once submitted the option is available for a user to track the progress of the order An example of the web portal process for ordering CCS data is provided in the Appendix Further information is available in the Data Order Online Help page in the CCS web portal Department of Science Information Technology and Innovation FTP data collection site The user will be notified by email as soon as the data order has been processed The email will provide a link to the specific location of the data under CCCS_Web_Data_Outputs on the FTP data collection site Figure 2 2 Ancillary information hosted on the FTP data collection site includes the following e the User Guide e a52 station QM 2050 test set see Section 2 2 based on a single GCM one emissions scenario and one climate sensitivity e asingle station 1899 2100 CF based transient data set see Section 3 11 FTP data collection site ftp climate mft derm qid gov au Climate_ Scenarios CCCS_Web_Data_Outputs CCCS_Batch_Outputs Documentation QM_2050_TestData Transient TestData Web orders awaiting Non web based User Guide and 52 station dataset Single station 1889 2100 collection orders other ancillary CF based test set documents Brigalow Research Station Qld Figure 2 2 Information available at
102. enario These are discussed in the following section 7 2 Selecting emissions scenarios Although the 21 Century concentration pathways for the six emissions scenarios and two CO stabilisation scenarios described remain uncertain we strongly recommend the use of A1FI since recent observations show that A1Fl which carries the most extreme risk is the emissions scenario that most closely represents the present day situation A1Fl depicts a socio economic future of very rapid economic growth and rapid introduction of new and more efficient technologies The A2 emissions scenario is often preferred as an alternative to A1FI The low end B family of SRES emissions may be overly optimistic Inclusion of low end scenarios in modelling the range of uncertainties for climate change projections has the potential to bias final output and hence under estimate the effects of climate change Thus the use of mid to high end A family of SRES emissions scenarios which produce higher rates of global warming is recommended as being more realistic than using those from the low end low rate of global warming B family of emissions scenarios Table 5 1 Further reading about limitations in emissions scenarios is presented in Section 10 3 7 3 Climate warming sensitivity Climate warming sensitivity is a simple but useful measure of the strength of the effect of CO2 concentrations on climate particularly global temperature Rahms
103. ence Information Technology and Innovation The QM 2050 station location file and QM 2050 projections data can be downloaded via ftp climate mft derm qld gov au Climate_ Scenarios QM 2050 TestData QM 2050 ZIP archives QM 2050 climate projections data and diagnostic files for each climate site are contained in self extracting ZIP format archives at ftp climate mft derm qld gov au Climate_ Scenarios QM_ 2050 TestData as follows LocationCode_ProjectionsMethod_ProjectionsYear_DEMO Archivetype e g 002012 _QM_2050 DEMO zip exe e LocationCode is a 6 digit number BoM station code if patched point i e 051039 or all Zeros if drilled from interpolated surfaces i e 000000 e ProjectionsMethod QM represents quantile matched projections e ProjectionsYear 2050 e DEMO represents QMV2 2 10 2050 test dataset e Archivetype zip exe for Windows QM 2050 Projections files Once a ZIP archive is opened individual QM 2050 climate projections data files and ancillary files are then accessible The QM 2050 climate projections data files are named as follows LocationCode_Projections Year_ProjectionsMethod_VersionNumber SILOformat e g 002012 2050 QM2 2 met e LocationCode is a six digit number BoM station code if patched point i e 051039 or all Zeros if drilled i e 000000 e Projections year 2050 e Version Number where QM2 2 represents QM 2050 test data e SILO format either met for AP
104. eorology Class A evaporimeter which is a water filled circular pan of galvanized iron 121 cm in diameter and 25 cm deep mounted on an open wooden platform Manually observed Bureau of Meteorology Class A pan evaporation site data are used in the SILO historical data files i e APSIM and p51 formats from 1970 to current It is worth noting that bird cages were installed in the early 1970s to reduce errors e g evaporation readings were occasionally too high due to animals drinking or splashing water Prior to 1970 the SILO historical pan evaporation data have been interpolated from long term averages 7 i OzClim version 3 55 Department of Science Information Technology and Innovation Notes e The accuracy of the historical daily data depends on many factors including date location and how the climate variable has been measured e Although SILO data is supplied to one decimal place to maintain consistency in some cases it may not be accurate to that precision i e an observer measuring the temperature to the nearest 0 5 C e All SILO historical data are provided by the Australian Bureau of Meteorology and are collected and are prepared to their standards 6 2 Patched Point and drilled data The SILO historical datasets are available as either Patched Point or Drilled While both datasets provide continuous daily climate data suitable for use in simulation models there are subtle but importa
105. es 0 1 0 5 and 0 9 quantiles e 1957_2010 historical training period e Projections year 2030 or 2050 e Histograms denotes Histograms trend plot file e Model Name i e CSIRO MK35 e Scenario emissions scenario i e A1B A1FI etc e Climate warming sensitivity rate of global warming i e low med high referring to the 10 50 and 90 percentile values respectively A snapshot of a quantile based historgam is presented in Figure 3 13 41 Department of Science Information Technology and Innovation Augusts C obs ys 7 80 3 91 C boot u sy 8 68 3 92 Frequency T Min Figure 3 13 Minimum temperature information displayed in a histogram of quantile matched QM projections data for Augusts for a specific location showing differences between the observed 1957 to 2010 baseline climate and QM 2030 climate projections data This plot filename T Min_040428 10 50 90 _1957 2010_2030_Histograms_040428 HADCM3_A1FI_high png shows a positive shift in the QM frequency distribution The blue semi circles on the x axis represent projected 10 50 and 90 percentile values of the mean daily minimum temperature for 2030 using QM trends Observed 1957 2010 and projected 2030 means and corresponding standard deviations are presented in the top right panel Notes e Where observed quantile trends are significant p lt 0 05 the QM approach is used to produce the 2030 rainfall temperature solar radi
106. ess change in the annual mean temperature 35 Department of Science Information Technology and Innovation 3 8 Plots of simulated 20 and 21 Century climate Plots of GCM simulated 20 and 21 Century climate were initially made available for a selection of 23 AR4 A1B forced GCM runs see Section 8 Table 8 2 with CF Version 1 and Version 1 1 Consistent Climate Scenarios datasets The plots which also provide some indication of projected climate variability are now contained in a draft document entitled Comparisons of Australian continental area weighted mean rainfall and temperature and global mean temperature according to available AR4 A1B forced GCM runs Each plot included graphs showing time series of the observed and GCM simulated 20 Century and GCM simulated 21 Century for e Australian continental area weighted mean rainfall e Australian continental area weighted mean temperature and e global mean temperature The monthly mean precipitation and temperature data used to produce the Australian continental graphics were accessed from DSITIA copies of the IPCC CMIP3 data set For the comparisons of global mean temperature according to available GCM runs the measured series 1901 2009 initially provided as anomalies from the 1951 1980 base period were taken unmodified from GISS GLOBAL Land Ocean Temperature Index in 0 01 C sources being GHCN 1880 2010 SST and 1880 1 1 1981 HadSST1 These graphs ar
107. eviation minimum and maximum change in rainfall per degree of global warming expressed as a percentage of GCM base line 1975 2004 climate Errors marked in red exceed 100 percent These GCMs have been used in the CF Version 1 and 1 1 Consistent Climate Scenarios projection data Error flags for anomalous trends per degree of global warming GCM Climate Month Mean trend Standard Minimum Maximum Error flag Name variable in rain deviation CCCMA 63 Rainfall 6 6 43 11 99 7 53 101 46 extremes out of range in tpc Rainfall 9 19 49 22 01 99 33 131 09 extremes out of range in tpc CSIRO MK35 Rainfall 10 34 32 57 77 545 89 15 98 extremes out of range in tpc MRI GCM232 Rainfall 10 14 48 30 14 338 43 13 26 extremes out of range in tpc 10 6 Known issues related to the calculation of change factors The following matters need to be considered in regard to the pattern scaling technique and GCM ensemble members from which trends per degree of global warming also called patterns of change and hence change factors are calculated Pattern scaling e Pattern scaling the basis of the change factor approach is based on two main assumptions 1 Using GCMs all projected changes in the future climate follow a linear relationship between global warming and the variables of interest This assumption is applied even when the climate variables are known to have a non linear relationship between them 2
108. for each of the eight emissions scenarios and the three climate warming sensitivities are presented in Table 7 3 These data are also available in the monthly multiplier files supplied with the data Table 7 3 Projected amounts of global warming at 2030 and 2050 change from a 1990 baseline from MAGICC for the eight emissions scenarios and three climate warming sensitivities utilised in producing CF 2030 and CF 2050 data Emissions scenarios and projected amounts of global warming C at 2030 and 2050 from MAGICC 2030 2050 Climate warming sensitivity Climate warming sensitivity Change from 1990 baseline C Change from 1990 baseline C Low Median High Low Median High Scenario 10 percentile 50 percentile 90 percentile 10 percentile 50 percentile 90 percentile A1F1 0 63 0 87 1 13 1 31 1 81 2 35 A1B 0 66 0 90 1 17 1 12 1 53 1 99 A1T 0 73 1 00 1 30 1 24 1 71 2 22 A2 0 57 0 79 1 02 1 05 1 44 1 87 B1 0 54 0 74 0 96 0 82 1 13 1 47 B2 0 65 0 89 1 15 0 97 1 33 1 72 C02 450 0 52 0 71 0 89 0 83 1 15 1 43 C02 550 0 57 0 78 0 96 0 98 1 34 1 66 62 Consistent Climate Scenarios User Guide Version 2 2 We recommend the combined use of all three climate warming sensitivities low median and high as these sensitivity values will incorporate some of the uncertainty about the range of potential biosphere carbon feedbacks These biosphere carbon feedbacks could be as large as the difference in the
109. free sample of the data ZIP 1 3M last updated 03 00PM 14 May 2012 is available for testing without being registered To register for data access please fill out and submit the registration form Once we have returned your registration details you can start requesting data Only small orders for data can be accepted via this web site For example you can only download data for up to ten locations and up to three emission scenarios at any one time If you wish to order larger amounts of data please contact Grazing Land Systems REQUEST DATA The CCS login page Figure 2 1 will then appear Please Enter your Credentials to Access System Do not bookmark this page If you wish to create a bookmark wait until after you have logged in Email Address Password Figure 2 1 CCS login page Consistent Climate Scenarios User Guide Version 2 2 Following login the steps required to order data are Step 1 Select order type Choose the type of data Data can be for Weather stations or point locations by Latitude and Longitude with an option for Full data daily projections or Summary data i e plots showing comparisons of GCM model projections for 2030 based on the A1B emissions scenario Step 2 Select weather station locations Select the required weather stations by station Name or ID Code or point locations by latitude and longitude Step 3 Select data pa
110. g 97 12 GlOSSA Y issis aaa aa daaa aaa aaas 98 13 References iiia aaa aaa aaa aaa aaa aa Aaaa Eaa a aaa Ea a aaar anaana 101 14 AS OMT ACE CS CaaS eese oeeo Eo ESPES eE PEE EEEE SERE SEa Eepe eSEE EE EAEri Eee SEa 103 ND oo ao h A 104 DAFF Climate Change Research Program Projects 104 Consistent Climate Scenarios Web Portal 105 Consistent Climate Scenarios User Guide Version 2 2 1 Introduction Researchers conducting studies of climate change impacts on primary industries have previously not had access to a consistent set of climate change projections in a suitable format for use in biophysical models The aim of the Consistent Climate Scenarios Project CCSP has been to develop a consistent set of synthetic climate projections data across Australia for use in biophysical models which maintain weather like properties and also account for uncertainties and biases in climate change projections as well as different methods of downscaling Since July 2012 the CCSP has been delivering a consistent set of model ready AR4 based 2030 and 2050 Australia wide climate change projections data via the Long Paddock website s Climate Change Projections web portal http www longpaddock qld gov au climateprojections The first phase of the project used CSIRO s OzClimTM change factor CF approach described in Section 4 to transform historical climate data based on projections information from CSIRO s Oz
111. ght emissions scenarios matched QMV2 2 0 2030 Jul 2011 three climate sensitivities six climate QM data QMv2 2 in filename variables and one projections year 2030 Two extra GCMs HADCM3 and HADGEM1 added in June 2012 These Code adjusted to fix a trivial error in files have QMV3 0 2030 QOMV2 2 0 for which an insignificant amount of a QM 2030 amp QMv3 0 May 201 daily rainfall data had been affected and to method tag in filename negligible extent oe Test set 2050 only for a single GCM ECHAM filename 5 one emissions scenario A1B and one QMV2 2 10 climate sensitivity median Approach for 2050 2050 rainfall and temperature projections uses Nov 2011 i 2050 QM2 2 in daily GCM data Projections for other climate filename variables based on QM 2030 method extended to 2050 2050 QM2 2 in filename but internally known as QMV2 2 10 to reflect different data 97 Department of Science Information Technology and Innovation 12 Glossary Agricultural Production Systems Simulator APSIM An intensive grazing and crop modelling framework used primarily to investigate the management of climate variability at a farm scale Amount of global warming The estimated amount of global warming at a future point in time i e 2050 occurring with a specified SRES emissions scenario AR3 AR4 AR5 International Panel for Climate Change IPCC Third Fourth and Fifth Assessment Reports Australian Go
112. hange factors and climate projections data based on these e Inthe underlying historical climate data many stations may have a high degree of patching prior to 1957 in order to produce seamless data files back to 1889 The patching is based on spatial interpolation and other techniques Jeffrey et al 2001 The accuracy of patched data is lowest in isolated station poor regions and is not improved in the change factor based projections data e Inthe projections data there is no change in the probability distribution function PDF of temperature other than a uniform shift However inreality some amount of change is likely in the PDF which could lead to increased or decreased variability and hence a change in the frequency or high or low temperatures 92 Consistent Climate Scenarios User Guide Version 2 2 A uniformly applied increase in minimum temperature may completely eliminate the occurrence of frost at some locations as complete elimination of frost may not occur in reality and this outcome could be unreasonable since normal climate variability even after global warming has occurred is still likely to include a small probability of near sub zero temperatures that would result in frost In cases where a GCM shows a decreasing 21 Century trend in rainfall a uniformly applied percentage decrease in rainfall may generate negative rainfall at some point in the future particularly
113. he trends that have been used to estimate target cumulative distribution frequencies CDFs for 2030 for projected rainfall maximum temperature minimum temperature solar radiation and specific humidity or their transforms and target CDFs for 2050 for solar radiation and specific humidity The plots are arranged by month and show historical time series for quantiles 0 1 0 5 and 0 9 of daily data for each month including1957 to 2010 training period linear trends as well as forward extrapolation of these trends to 2030 2050 Computed statistical significance of each trend represented by p values over the training period is given as well for each climate variable Quantile trends have been classed as significant when the computed statistical significance called the p value statistic shown on each graphic is 0 1 or less i e when there is at most a 10 per cent probability that the observed data arose from an underlying stationary distribution If a computed p value for a quantile trend exceeds the 0 1 cut off the trend is not used in the QM projections and no trend is applied This represents a null model in which the future quantiles are the same as those over the training period In the QM processing corrections are made in cases where future trends would cross This does not apply to rainfall and temperature variables in the QMV2 2 10 datasets for 2050 which utilise a future CDF based on daily GCM data and not trend extrapolation but
114. ht both the Consistent Climate Scenarios Project Expert Review Panel and Suppiah s more favoured models GCMs Weighted Consistent Climate Demerit points based failure rate Scenario Project on Australian rainfall Name used in PCMDI CMIP3 Smith and Expert Review Panel temperature and MSLP Consistent GCM name Chiew 2009 recommendation Suppiah et al 2007 Climate Crimp et al 2010 Scenarios Project HADCM3 UKMO HadCM3 0 6 MIROC H MIROC3 2 hires 8 7 GFDL 21 GFDL CM2 1 13 2 GFDL 20 GFDL CM2 0 20 More likely to produce 2 MIROC M MIROC3 2 medres 25 credible projections 7 ECHO G ECHO G 33 4 HADGEM1 UKMO HadGEM1 33 2 ECHAM5 ECHAM5 MPI OM 38 1 MRI GCM232 MRI CGCM2 3 2 40 3 NCAR CCSM CCSM3 44 2 CCCMA 63 CGCM3 1 T63 50 Likely to be less reliable 10 GISS AOM GISS AOM 58 8 INMCM INM CM3 0 59 7 CCCMA 47 CGCMs3 1 T47 63 8 IAP FGOALS G10 FGOALS g1 0 63 Likely to be less reliable 2 CSIRO MK30 CSIRO Mk3 0 73 Consistent under 7 CNRM CNRM CM3 75 ae 4 4 Not used in project IPSL CM4 75 pe 14 BCCR BCCR BCM2 0 88 5 Not used in project GISS ER 88 Not recommended 8 Not used in project PCM 89 11 Not used in project GISS EH 100 14 CSIRO MK35 CSIRO MK3 5 Not assessed but expected to be better than CSIRO Mk3 0 Note e The INMCM GCM previously rated likely to be less reliable is no longer recommended for use due to unstable drift in the model The CF
115. huss plus the nominated predictands 78 Consistent Climate Scenarios User Guide Version 2 2 Table 9 6 GCMs selected for the availability of specific humidity huss GCM name used in Consistent Climate Expert Review Panel recommendation Scenarios Project CSIRO MK35 Not assessed but expected to be better than CSIRO Mk3 0 MIROC H ae More likely to produce credible projections CCCMA 47 added later CCCMA 63 liabl GIS AOM Less likely to be reliable IAP FGOALS G10 MRI GCM232 omitted later CNRM Consistently underperformed BCCR Not recommended INMCM Euriqua was used to select candidate variables and simple combinations of variables for each month from 13 candidate sites within Australia The final runs of Euriqua attempted to fit the following relation huss_toc f hus_92500_trend hus_92500_ave pr_tpc prw_tpc prw_ave tas_trend clt_trend tas_prw tas_pr pr_prw rsds_tpc dist where huss screen specific humidity hus_92500 air column specific humidity at 925 HP pr precipitation prw precipitable water tas mean temperature at surface clt cloud cover rsds solar radiation toc percentage trend per degree of global warming ave model average of present 1975 2004 trend trend per degree of global warming dist distance from coast For each month Euriqua ran the above model for 10 minutes and then the partial functions from th
116. ic humidities and pressures typically encountered in Australia this function is very nearly linear A linear approximation is used to compute changes in vapour pressure as a function of changes in specific humidity 0 62198 x VP x ASH SH x P AVP AP where VP SH and P are as above AVP ASH and AP are all relative percentage changes 76 Consistent Climate Scenarios User Guide Version 2 2 VAPOUR PRESSURE AS A FUNCTION OF SPECIFIC HUMIDITY AND ATMOSPHERIC PRESSURE Pressure 980 hPa Pressure 1013 25 hPa Pressure 1040 hPa 90 80 70 60 50 10 Vapour Pressure hPa RH 100 T 1 C P 1013 25 RH 100 T 40 C P 1013 25 30 20 10 9 60 0 01 0 02 0 03 0 04 0 05 0 06 Specific Humidity g vapour g moist air Figure 9 4 Vapour pressure y axis as a function of specific humidity x axis and atmospheric surface pressure over ranges of values typically encountered in Australia While trends in specific humidity were provided by CSIRO we have assumed zero change in average monthly atmospheric surface pressure for this calculation We have assumed a fixed baseline atmospheric pressure of P 1013 25 hPa with no alteration due to climate change AP 0 Mean monthly surface pressure changes in the GCM models tend to be small in comparison with changes in specific humidity about 0 3 c f 5 and will have a relatively small impact on the change in v
117. ile is shown in Figure 3 5 Warning Potential Extreme Percent Change Rate INMCM Precipitation 2050 high ALFI Sep 51 93 Warning Potential Extreme Percent Change Rate INMCM Precipitation 2050 high ALFI Nov 52 11 ECHAMS_ALB_med_2050 Number of projected Tmin greater than projected Tmax 0 ECHAMS_ALB_med_2050 Number of projected synthetic evaporation less than 0 0 ECHAMS_ALB_med_2050 Number of projected vp greater than vp Saturated 0 ECHAMS_ALB_med_2050 Number of clamped projected radiation greater than ET radiation 0 Warning Potential Extreme Percent Change Rate CCCMA 47 Precipitation 2050 high ALT oct 52 01 Warning Potential Extreme Percent Change Rate CCCMA 47 Precipitation 2050 nigh ALFI oct 55 05 Warning Potential Extreme Percent Change Rate CSIRO Mk30 Precipitation 2050 high ALT Jun 59 71 Warning Potential Extreme Percent Change Rate CSIRO Mk30 Precipitation 2050 high AlT Auq 52 10 Figure 3 5 Snapshot of information contained in a log warning ancillary data file filename 056002_ 30 5167_151 6681_V1 2 log For example Potential Extreme Percent Change Rate will be listed in a log warning file if the application of the change factor produces more than a 50 decline in projected rainfall from the historical baseline climate In this case no adjustment is made to the rainfall projections data but users need to be cautious if applying that data in any modelling study An Implausible Extreme Percent Change Rat
118. ily climate projections data for 2030 and 2050 can be ordered for each of the four RFCs HI HP WI and WP through the CCS web portal see Appendix Data Order Step 3 In addition for each RFC the multiplier files Section 3 1 that are provided list monthly and annual projected rates of change per degree of 21 Century global warming for a range of climate variables Notes e Before running composite projections it may help to become familiar with the Plots of simulated 20 and 21 Century climate provided in this project These plots have been provided for individual runs of all 19 GCMs used in the CCS project based on the A1B emissions scenario for Australian continental area weighted rainfall and mean temperature and are discussed in Section 3 8 of the User Guide e Users should note that the partitioning approach described by Watterson 2011 is focused on annual climate change over the whole of the 21st Century nominally for 2100 relative to 2000 using 23 CMIP3 GCMs forced by the SRES A1B emissions scenario for which the best estimate for global warming was an increase of 2 8 C The Consistent Climate Scenarios projections data do not extend to 2100 They are for 2030 and 2050 and are not restricted to the A1B emissions scenario e We caution that composite temperature projections based on GCMs grouped into each of the four RFC partitions may show unexpected temperature results for Australian locations as the amo
119. ions scenario at 2030 and 2050 e projected rates of change per degree of 21st Century global warming for a range of climate variables for each GCM The multipliers are listed for each of the 19 GCMs Section 8 Table 8 2 four GCM composites based on the Representative Future Climate partitions WP WI HP and HI described in Section 8 Table 8 3 eight emissions scenarios Table 7 1 and three climate sensitivities Table 7 3 These multipliers are the ones that have been used to calculate the climate change factors that have been applied to the SILO historical daily climate data to produce the 2030 and 2050 CF climate projections data The multiplier files include both monthly and annual values The change factors are calculated for specific climate variables by multiplying amounts of global warming by rates of change per degree of 21 Century global warming Further detail related to the application of the data contained in the 20 Consistent Climate Scenarios User Guide Version 2 2 monthly multiplier files is contained in Section 4 Change factor CF methodology and Section 7 3 Climate warming sensitivity The multiplier files are typically named LocationCode_Latitude_Longitude_VersionNo mutltiplier e g 051039_ 31 5495_147 1961_V1 2 multiplier e LocationCode is a six digit number BoM station code if patched point i e 051039 or all Zeros if drilled from interpolated surfaces i e 00
120. itude 151 6681 DECIMAL DEGREES tav 13 89 oC Annual average ambient temperature amp 13 37 oC Annual amplitude in mean monthly temperature Metadata t _ pats Extracted from Silo on 20120429 for APSIM tt Some early data in this file is only possible because of the Climarc project Thanks Climarc IAs evaporation is read at 9am it has been shifted to day before lie The evaporation measured on 20 April is in row for 19 April lFor further information see the documentation on the datadrill http Wwww longpaddock qld gov au silo year day radn maxt mint rain evap wp 0 O MJ m oC oC mm mm hPa 1980 1 28 0 28 6 133 00 56 168 1980 2 31 0 339 13 3 00 56 18 1 1960 3 28 0 33 3 147 0 0 58 209 o Year 960 4 24 0 31 7 194 00 58 16 1 1960 5 12 0 19 4 139 11 7 6 0 15 2 oO OO 1960 6 25 0 250 106 00 6 0 128 x Solar radiation 1986 4290 26 4 106 0 0 58 135 2 Maximum air temperature 5 t Minimum air temperature 4960 49 59 294 439 00 56 163 I Rainfall _ 211 280 283 133 00 54 161 1960 8 29 0 27 2 111 00 58 145 1960 9 22 0 27 2 139 0 0 58 163 1960 12 29 0 30 0 11 7 00 54 150 Pan evaporation 1980 13 260 287 133 00 5 4 158 1960 14 27 0 27 6 122 00 52 150 Vapour pressure err Ere 115 0057 150 Figure 2 6 A snapshot of information observed data presented in a historical baseline climate data file suitable for use in APSIM filename 056
121. itude in mean monthly temperature parameters shown in the QM metadata in the APSIM files are calculated based on the 1957 2010 training period e The daily dates presented in the QM projections files indicate only the historical date that was the source of the associated perturbed data before the QM methodology was applied The month day and Julian day fields in the projection files are correct e QMv3 0 represents a code upgrade from QMv2 2 see section 11 10 Availability of QM 2050 Projections files Due to a lack of raw daily GCM data QM 2050 projections data are only available for a single GCM ECHAMS the A1B emissions scenario and median climate warming sensitivity Each QM 2050 file contains projections data for rainfall maximum and minimum temperature projections are based on the QM 2050 methodology see Section 5 2 while and vapour pressure evaporation and solar radiation projections are based on the QM 2030 methodology see Section 5 1 Since July 1 2012 QM 2050 projections data have been limited to a test set of 52 locations within Australia Figure 2 5 A file testSites csv lists those locations including station names location codes latitudes and longitudes Access to more locations is expected via the Climate Change Projections web portal at some stage in the future Figure 2 5 Map showing 52 locations for which quantile matched projections data for 2050 are available Department of Sci
122. l have been pre transformed to limit any non physical projections such as negative rainfall The pre transformation is also done to normalise the data i e make the regression residuals more nearly normally distributed The QM procedure is applied to the transformed rainfall and temperature variables An inverse transform is applied to the resulting projected data to generate the projected climate variables After this steps 3 to 5 are applied as per the 2030 QM method Caution e The QM 2050 projections data should be used with care as these projections are only available for a single GCM ECHAM 5 emissions scenario A1B and climate sensitivity median e QM 2050 projections are limited to the ECHAM 5 GCM Of the 17 GCMs used in this project i e CF and QM 2030 projections ECHAM 5 is the only GCM that is both noted by the Expert Panel as being more likely to produce credible projections and having a complete set of raw daily rainfall maximum and minimum temperature data available from 1900 to 2100 particularly 15 years either side of 2050 e Users should note that QM 2050 projections are currently limited to the A1B emissions scenario which while being recommended for use does not represent the most extreme risk e Furthermore while QM 2050 projections for daily rainfall maximum and minimum temperature projections are sourced from daily GCM data other daily climate variables were not available in GCMs due to limited holding
123. l projections file VersionNumber V1 2 represents CF data FileType gif for v1 1 or png for V1 2 A snapshot of a comparison of model projections plot based on the climate change factor methodology is presented in Figure 3 8 Model Projections for A1B Emission Scenarios 13 8345_131 1872 Change in median for a specific model GDFL 21 N Average change in median for aggregate of all models H H N Change in annual mean temperature for the 90th percentile climate warming sensitivity applied to a specific model 0 8 Change in annual mean temperature for the 10 percentile climate warming sensitivity applied to a specific model Change in Annual Mean Temperature C o io 0235 10 5 0 5 10 Change in Annual Mean Rainfall m 10th L and 90th H e Model M 2012 05 25 Base Period 1960_2010 O Average Figure 3 8 Comparison of model projections plot for 2030 for Douglas River LocationCode 014901 in Northern Territory Plot generated using Version 1 2 projections data climate change factors for GCMs forced by the A1B emissions scenario filename 014901_A1B_ M_2030_ 13 8345 131 1872_mdlperf_V1 2 png Note e GCMs with large changes in annual mean temperature tend to have a wide range of uncertainty between the 10 and 90 percentile values for both temperature and rainfall The uncertainty range between the 10 and 90 percentile values is smaller where GCMs have l
124. les 2030 amp QMv3 0 y rainfall data had been affected and to negligible extent have a QM in filename method tag in the filename QMV2 2 10 2050 only Limited to a 52 station test set for a single GCM ECHAM 5 one emissions 2050 a scenario A1B and one climate sensitivity median Approach for 2050 rainfall and Nov 2011 temperature projections uses daily GCM data Projections for other climate variables based on QM 2030 method extended to 2050 2050_QM2 2 in filename but internally known as 2050 amp QM2 2 in filename QMV2 2 10 to reflect different data available via ftp climate mft derm qld gov au Climate_Scenarios QM 2050 TestData V1 0 Nov 2010 Development version supporting change factor V1 data for eight GCMs V1 1 May 2011 Development version supporting CF V1 1 data for 17 GCMs V2 Oct 2011 Information added to support QM 2030 test set Includes CF and QM 2030 methodologies Information added to support HADCM3 and HADGEM1 GCMs CF V1 2 and QM 2050 data User Guide V21 Aug 2012 sets updated filenames descriptions for QM trend plots and QM 2050 methodology Brief 9 notes about transient data sets the Climate Change Projections data interactive web portal and the ftp site Additional information on the availability of CCS datasets through the web portal updated V2 2 May 2015 file naming conventions Section 2 and projections data for four Representative Future Climate partitions Section 8 http longpaddock qld gov
125. les were surveyed using Eureqa Schmidt amp Lipson 2009 Eureqa from Cornell Creative Machines Lab is a symbolic regression genetic programming package used to determine a set of candidate predictive relationships The climate variables and combinations that had the best predictive power were found to be cloud cover cit precipitable water jprw mean temperature at surface tas and the two combinations tasxcit and tasxprw Data were tested on a set of 13 locations around Australia representing a range of conditions to derive the shape of the function Multiple linear regression was performed for each month using all continental native model grid points fitted to the following equation Although OzClim output are available for the INMCM GCM these were omitted i e not used in the equation after cross validation diagnostics revealed a severe skewing effect on the regression Tt Pin x LAS p G m P 4 m P 3 m P m X Prw x clt p G m P 5 m p G m x las p G m x clt p G m x taS Gm x PYW p G m Equation 1 General regression model fitted Where Tis one of Tasmax and Tasmin pis an individual grid point G is a GCM from the set with known Tasmax and Tasmin p denotes individual points selected at native GCM resolution and G denotes all GCMs m is a month and P is a set of parameters estimated by multiple linear regression Note that for each month a single set of regression parameters is comp
126. limate trends e The date formatting in the APSIM and p51 files differs as follows APSIM YYYY DayNumber 1 365 6 p51 YYYYMMDD Examples of information presented in CF climate projections data files are presented in Figures 2 3 and 2 4 Figure 2 3 is a screenshot of projections data for 2050 based on output from the HADCM3 GCM forced by the A1Fl emissions scenario with high climate warming sensitivity formatted for use in APSIM Figure 2 3 provides an example of information presented in a climate projections data file in the p51 format suitable for use in the GRASP pasture model APSIM format CF V1 2 Annual average ambient temperature mean monthly texperature 311 for APSIM is only possible because of Metadata has been shifted to day before fie The evapor O April is in row for 19 April Date projecte For further infor tion see the documentation on the datadrill 8 http www longpaddock qid gov au silo year day radn maxt mint rain evap y MJ m 2 oC oC za zm hPa Solar radiation 1960 25 3 32 7 o A 0 7s 22 2 2 Maximum air temperature 960 avg J3 Aa S e 1960 0 0 7 5 21 2 5 Minimum air temperature 560 0 0 7 5 23 3 is 0 0 7 5 26 5 0 5 7 5 23 3 0 0 7 5 21 2 0 0 7 20 1 Vapour pressure 5 Sn 25 4 Figure 2 3 A snapshot of information projections for 2030 presented in a climate projections data file suitable for use in APSIM filename 040428_A1FI_2030_H_HADCM
127. m temperature Tasmax and Tasmin GCM name used in Consistent Climate Expert Review Panel recommendation Scenarios Project CSIRO MK35 Not assessed but expected to be better than CSIRO Mk3 0 MIROC H ar MIROC M More likely to produce credible projections GIS AOM Less likely to be reliable CSIRO MK30 Consistently underperformed BCCR INMCM Not recommended Table 9 3 GCMs for which trends per degree of global warming daily maximum and minimum temperature Tasmax and Tasmin have been estimated GCM name used in Consistent Climate Trend estimation method Scenarios Project CCCMA 47 CCCMA 63 CNRM ECHAM5 ECHO G Regressed GDFL 20 GDFL 21 IAP FGOALS G10 MRI GCM232 NCAR CCSM While two methods were initially available for use in calculating Tasmax and Tasmin the mean ratios method and the regression method the regression method was the preferred method of use as it used more information i e precipitable water and cloud cover from the individual GCM runs 71 Department of Science Information Technology and Innovation Regression method A regression method developed by Ricketts 2011 has been used to produce a single multiple regression model which would apply to trends in both maximum and minimum temperature As a first step towards estimating these trends the relationship between the trends in maximum and minimum temperatures and other climate variab
128. material intensity Technologies Emphasis on global solutions to a economic social and environmental g sustainability including improved equity but without additional climate 53 initiatives D z B2 Submitted to IPCC for Emphasis on local solutions to 425 473 S PCMDI but not as complete economic social and environmental 2 as A1B Data obtained by sustainability Continuously increasing D pattern scaling global population rate lower than A2 Intermediate levels of economic 9 development and less rapid and more 4 diverse technological change than B1 and A1 storylines Oriented toward environmental protection and social equity but focused on local and regional levels 58 Consistent Climate Scenarios User Guide Version 2 2 Further reading and details regarding the full range of greenhouse gas emissions scenarios is available at http www ipcc ch pdf special reports spm sres en pdf Table 7 2 CO stabilisation scenarios used in the Consistent Climate Scenarios Project adapted from IPCC 2000 CO stabilisation Storyline scenario Remarks CO2 450 CO emissions increase and then stabilise by 2100 a very low Data for this emissions scenario project obtained by pattern CO2 550 CO emissions increase and then stabilise by 2150 similar to scaling B1 1000 900 800 700 600 CO2 concentration ppm 500 400 300 O D hones oO Oo ao oOo
129. me used in OzClim and Consistent Climate MRI GCM232 NCAR CCSM CCCMA 63 GISS AOM CCCMA 47 IAP FGOALS G10 CSIRO Mk30 CHRM ISPL CM4 not used in project GISS ER not used in project NCAR PCM1 not used in project NOT ASSESSED CSIRO MK35 Not assessed Model recommended by Suppiah Model not in Suppiah s set Surface Surface downwelling Daily max amp min Wet area Expert Review Total cloud i Precipitable short wave Mean surface air surface air potential Panel s fraction idity Precipitation water radiation temperature temperature evaporation reccomendation completed city huss d prw sS tas tasmax tasmi wap More likely to produce credible projections Less likely to be reliable Not recommended Less likely to be reliable p Consistently under performed Not recommended cit i PT s oT NAA ipr iprw Se ee Rea See ee a SS rae ere a S S a S S a E a E A SE T il Sas i Trend per degree global warming available from OzClim Version 1 projections data Sep 2010 Version 1 1 Apr 2011 Version 1 2 May 2012 ew Ne o Trends in total cloud fraction preciptiable water and mean surface air temperature used to calculate trends in daily maximum and minimum air temperature Version 1 Trends in surface specific humicty used to calculate trends in vapour pressure Version 1 1 amp 1 2 Trends in total cloud fraction preciptiable water do
130. mean 21 49 cv 1 38 26 F 8 B 24 w v gt a w 22 H a E g 20 18 20C slope 0 005 r 0 31 21C slope 0 028 r 0 82 21C intercept 10 09 Pr 0 00148284 intercept 36 57 Pr o Obs mse 0 44 mean 19 89 cv 2 23 _ mse 0 56 mean 21 44 cv 2 62 1900 1950 2000 2050 2100 Year Figure 3 10 Time series presented in a continental area weighted temperature file filename CSIRO Mk30 tas run1 png for CSIRO Mk3 0 with the A1B emissions scenario This run shows a marked increase in Australian continental weighted mean temperature in the 21 Century csiro_mk3_0 20c run1 gw txt 16 0 T T T T T 20C Model 21C Model A1B 15 5 Obs GISS y W 15 0 o 145 nna D v 3 a E 14 0 2 3 Q 13 5 13 0 12 5 129 i i i f 50 1900 1950 2000 2050 2100 2150 2200 year Figure 3 11 Time series presented in a global mean temperature file filename CSIRO Mk30 gw run1 png for CSIRO Mk3 0 with the A1B emissions scenario This run shows a marked increase in global mean temperature in the 21 Century 38 Consistent Climate Scenarios User Guide Version 2 2 3 9 Quantile trend plots If diagnostics are selected when ordering through the web portal the QM 2030 datasets will include plots of location specific quantile trends QM 2050 quantile trend plots can be downloaded via ftp climate mft derm qld gov au Climate Scenarios QM 2050 TestData These are t
131. mperature e mse mean standard error of annual rainfall or temperature e cv coefficient of variation of the mean annual rainfall or temperature e r regression coefficient e intercept intercept on the y axis where year 0 The yellow shading indicates periods where GCMs have both daily and monthly data Many GCMS do not have daily data prior to 1945 GCM precipitation vs observed Continental Australia csiro_mk3_0 run1 20c csiro_mk3_0 run1 21c 1400 20C slope 0 79 r 0 15 intercept 952 95 Pr 0 14 mse 150 56 mean 578 41 cv 26 1200 1000 S E E 800 IM INN x 2 2 i il I tH y Mii 400 j 200 OBS slope 0 73 r 0 26 21C slope 1 4 r 0 24 intercept 984 93 Pr 0 008 intercept 3509 85 Pr 0 015 mse 79 48 mean 446 30 cv 18 mse 163 74 mean 573 16 cv 29 Boo 1950 2000 2050 2100 Year Figure 3 9 Time series presented in a continental area weighted mean rainfall file filename CSIRO Mk30 pr run1 png for CSIRO Mk3 0 with the A1B emissions scenario This run shows a decrease in Australian continental weighted rainfall in the 21 Century Yellow bands represent periods for which DSITI holds daily information for that GCM 37 Department of Science Information Technology and Innovation GCM mean temperature vs observed Continental Australia csiro_mk3_0 run1 20c csiro_mk3_0 run1 21 OBS slope 0 0029 r 0 27 intercept 15 93 Pr 0 0059777 28 mse 0 30
132. n 105 Department of Science Information Technology and Innovation Data Order Step 2 Select weather stations Search by station number or name Within boundary box Station number or name bris Min Latitude Min Longitude Within state QLD Max Latitude Max Longitude Search results Search displays up to 30 results Station ID Station Name Location Select 40140 MT BRISBANE 27 14920 152 57810 Add to list 40214 BRISBANE REGIONAL OFFICE 27 47780 153 03060 Add to list 40215 BRISBANE BOTANICAL GARDENS 27 48330 153 03330 Add to list 40216 BRISBANE SHOW GROUNDS 27 45060 153 03330 Add to list 40223 BRISBANE AERO 27 41780 153 11420 Add to list Selected stations Station ID Station Name Remove 40223 BRISBANE AERO Remove This online form only allows for up to ten stations For larger requests please use our feedback form to request the information Data Order Step 3 Select data parameters Select a climate perturbation method Change factor Quantile matching Select historical baseline year start and end greater than 40 year duration strongly recommended Start year 1960 End year 2010 Select a projection year 2030 2050 Select up to three emission scenarios SRES Marker Scenario A1FI SRES Marker Scenario A2 SRES Marker Scenario A1B SRES Marker Scenario B2 E SRES Marker Scenario A1T E SRES Marker Scenario B1 E 550ppm stabilisation by 2150 450ppm stabilisation by 2100
133. n Opportunities and Challenges in Monsoon Prediction in a Changing Climate OQCHAMP 2012 Pune India 21 25 February 2012 Additional information summarising the Consistent Climate Scenarios Project is documented in Ricketts J H Kokic P N and Carter J O 2013 Consistent Climate Scenarios projecting representative future daily climate from global climate models based on historical data Department of Science Information Technology Innovation and the Arts Queensland Government CSIRO Mathematics Informatics and Statistics Australia 20 International Congress on Modelling and Simulation MODSIM Adelaide Australia 1 6 December 2013 http www mssanz org au modsim2013 L11 ricketts pdf Department of Science Information Technology and Innovation Contents 1 MroduChoM a a a 2 Accessing and interpreting data s sssssss s 5ss515515551555155555555555555555551255555552555555551512551255155555551533 2 1 Change factor CF data 2 2 Quantile matched QM data 2 3 Historical baseline climate data files 2 4 An end user example 3 Files with additional information sisicecssssicccicsctcsciesctencsseteccieectecsevecteccevectuecieectuecnesctueceeestunetevesneteees 3 1 Multiplier files 3 2 CO matching files 3 3 CQO concentrations files 3 4 Log warning files 3 5 Historical time series plots 3 6 CF Frequency distribution plots 3 7 Comparison of model projections plots 3 8 Plots of simulated 20 and 21
134. n evaporation mm for the 24 hours from 9am on the date listed While rainfall is captured in standard 203 mm eight inch diameter gauges at the majority of Australian recording sites some sites have automated gauges A standard rain gauge can hold 300 mm of water before overflowing Accumulated rainfall totals i e for periods of more than a day are not included in the SILO patched point datasets unless they have already been apportioned Maximum and minimum air temperatures are measured using thermometers or electronic temperature sensors The instruments which are standardised at 1 3 metres above level ground are located inside a Stevenson screen or other standard enclosure to protect them from exposure to direct sunlight As with rainfall SILO does not include accumulated maximum or minimum temperatures in its patched point data sets unless they have been apportioned Solar radiation is radiant energy emitted by the sun The measured data is the combined total of both direct and diffuse solar radiation per day received ona horizontal surface Vapour pressure is the partial pressure exerted by the water in the atmosphere The vapour pressure is dependant on the dry bulb air temperature and the coincident relative humidity Pan evaporation which integrates the evaporative effects of temperature humidity solar radiation and wind is normally highest when it is hot windy and dry Pan evaporation is measured using a Bureau of Met
135. n for the selected baseline period mm e CO2Conc Projected CO concentration ppm Values listed in the SILORain column are not projections but are the observed means for the historical climate baseline as selected by the user which is displayed in the second line of the data file In some cases nan may be displayed where a numerical value has been expected indicating that the computed values were out of range An example of information presented in a CF V1 2 multiplier file is presented in Figure 3 1 Department of Science Information Technology and Innovation Multipliers CF V1 2 Projected amount Historical Projected CO of global warming C rainfall concentration Station latitude and longitude from MAGICC MOTE Potential evaporation a cos version SILO na 1 baseline from 1950 to 2010 derived from changes in specific humidity and solar radiation month Year Sensitivity x Tassin Precipitation huss tpc sILORain co2zcanc s x x x pos Soca alt 1 Jan 2030 high 1 30 0 91629 379242 7 082777 0 2 Q 0 97 7 111 996078 437 9 SccR ALT 2 Feb 2030 high 1 30 0 ei 2 67 74003 2 a 0 97 84 376471 437 9 BccR alt 3 Mar 2030 high 1 30 0 3 7 7 034734 0 le 58 817647 437 9 Sock ALT 4 ior 2030 high 1 30 0 345180 7 0 Q 0 9 42 356863 437 9 Sor alr 5 May 2030 high 1 30 0 7 224912 le 0 gt 49 076471 437 9 BocR alt 6 mn 2030 high 1 30 0 6 02 437830 1 6 0 0 856519 40 07
136. n reality 10 3 Emissions and CO stabilisation scenarios Climate projections data using six SRES emissions scenarios and two CO stabilisation scenarios are available for use in the Consistent Climate Scenarios project Users should be aware that e We strongly recommend using the A1Fl emissions scenario as it most closely represents the current trend in global CO emissions A1FI also represents the most extreme global warming risk analysed to date e A1B is the scenario for which amounts of global warming and associated climate warming sensitivities were available for the greatest number of GCM runs through PCMDI Whilst runs for A2 and B2 were also submitted to PCMDI these were not as complete Where there was any missing data for any climate variable in A1B runs A2 runs were used for that GCM e Pattern scaling has been used to derive amounts of global warming and associated climate warming sensitivities for all emissions scenarios normalising them to the responses computed using the MAGICC simple climate model This includes the A1B and A2 scenarios e The low end less extreme B family of SRES emissions scenarios are looking increasingly unlikely The inclusion of these low end scenarios in modelling the range of uncertainties for climate change projections has the potential to bias outputs and hence under estimate the effects of climate change As the 21 Century concentration pathways for the six emissions scenarios and
137. ns scenario and historical time series plots for each selected climate site are contained in a ZIP format archive as follows User name_JobNumber_FileLabel_Plots_Archivetype e g john smith_its138_plots_zip exe Username derived from your email address JobNumber i e its138 its139 its140 etc FileLabel 1 8 character label specified by the user i e cf2030 testdata etc Plots shows that plot files are included Archivetype i e zip exe for Windows or zip for Unix Linux File types packaged in CF ZIP archives containing diagnostics plots including filename examples are o Comparison of model projections plots 051039_A1B_M_2030_31 5495_147 1961_mdlperf_V1 2 png described in Section 3 7 o Historical time series plots 051039_ 31 5495 147 1961 _V1 2 png described in Section 3 5 CF Projections files Once the ZIP archive is opened individual CF climate projections data files and ancillary files are then accessible The CF climate projections data files are named as follows LocationCode_Scenario_Projections Year_Climate Warming Sensitivity_ModelName_Latitu de_Longitude_VersionNumber SILOformat e g 051039 _A1FI_2030_M_CSIRO MK35_ 31 5495 147 1961 _V1 2 met e LocationCode is a six digit number BoM station code if patched point i e 051039 or all zeros if drilled from interpolated surfaces i e 000000 e Scenario emissions scenario i e A1B A1FI etc e Projections year i e 2
138. nt differences between the two sets Patched Point data uses original Bureau of Meteorology measurements for a particular meteorological station but with interpolated data used to fill patch any gaps in the observation record Drilled data is based on grids of data derived by interpolating the Bureau of Meteorology s station records Interpolations are calculated by splining and kriging techniques The drilled data are all synthetic there are no original meteorological station data left in the calculated grid fields However the drilled data does have the advantage of being available for any set of coordinates in Australia The Patched Point data would typically be used when an analysis or simulation is needed quite close to a meteorological station However if an analysis is required for a location which has no meteorological station nearby then the drilled data is the more relevant product Drilled data sets can be processed when a specific latitude and longitude is provided 56 Consistent Climate Scenarios User Guide Version 2 2 7 Emissions scenarios and climate warming sensitivity Available CF data includes 2030 and 2050 climate projections data based on nineteen GCMs for approximately 4700 Australian patched point climate stations To better inform users about the projections datasets and assist in dataset selection this section provides users with some background information about emissions scenarios and associa
139. ntained in a CF V1 2 multiplier file For the selected location multipliers include for each climate variable the projected change per degree of 21 Century global warming and projected amount of global warming at 2030 and 2050 used to construct CF daily climate projections data filename 056002_30 5167_151 6681_V1 2 multiplier Multipliers are listed for 19 individual GCMs and the four GCM composites WP WI HP and HI based on Representative Future Climate partitions Other information historical rainfall and projected CO concentration are also included 22 Consistent Climate Scenarios User Guide Version 2 2 3 2 CO matching files CF Version 1 2 data includes look up tables called CO matching files that have been provided for each location These files list the CO concentrations associated with each projections file that has been provided The CO matching files are named as follows LocationCode_Latitude_Longitude_Names List txt e g 051039_ 31 5495_147 1961_NamesList txt e LocationCode is a six digit number BoM station code if patched point i e 051039 or all zeros if drilled from interpolated surfaces i e 000000 e Latitude and longitude of the station or location in decimal degrees e NamesList txt CO2 matching file An example of information presented in a CO matching file is presented in Figure 3 2 with metadata as follows LocationCode_Scenario_Projections Year_Climate Warming sensitivity_Mo
140. ntly 1960 to 2010 If 1970 to 2000 is selected the calculation will be based on 1970 to 2000 Annual amplitude in mean monthly temperature amp C The difference between the long term mean of the warmest month of the year and the long term mean of the coolest month of the year Statistical period the same as tav The date that the projections data were computed Projections year 2030 or 2050 Emissions scenario Eight are available see Section 7 1 GCM model or RFC composite 19 Global Climate Models see Section 8 Table 8 2 and four Representative Future Climate composites see Section 8 Table 8 3 are available The CCS project s file naming convention for AR4 GCMs documented in Section 8 Table 8 2 uses abbreviations of the formal model names used by the Program for Climate Model Diagnosis and Intercomparison PCMDI Coupled Model Intercomparison Project phase 3 CMIP3 Model sensitivity low median or high refer to climate warming sensitivity Section 7 Ambient CO2 in year of projection ppm see Section 3 3 Notes Each file contains projections data for six climate variables based on the CF methodology see Section 4 The projections data are synthetic and do not represent a forecast The projections data have been developed for use as input to agricultural simulation models In the projections data the intensity of the rain on rain days is perturbed but the sequence of rain no rain days remains unch
141. o QM projections for 2030 051039_A1F1_2030_M_CSIRO MK35_ 31 5495_147 1961_QMv3 0 met described in this Section o Historical baseline climate data 051039_SILO_ 31 5495_147 1961_QMv3 0 met described in Section 2 3 o Multiplier files 051039_ 31 5495_147 1961_V1 2 multiplier described in Section 3 1 o CO2 matching files 051039_ 31 5495_147 1961_NamesList txt described in Section 3 2 QM 2030 ZIP archives containing diagnostic plots QM 2030 diagnostic plots which include comparison of model projections plots for 2030 based on the A1B emissions scenario historical time series plots QM histograms plots and quantile trend plots for each selected climate site are contained in a ZIP format archive as follows User name_JobNumber_FileLabel_Plots_Archivetype e g john smith_its138_plots_zip exe Username derived from your email address JobNumber i e its138 its139 its140 etc FileLabel 1 8 character label specified by the user i e qm2030 testdata etc Plots shows that plot files are included Archivetype i e zip exe for Windows or zip for Unix Linux Department of Science Information Technology and Innovation File types packaged in QM 2030 ZIP archives containing diagnostics plots including filename examples are o Comparison of model projections plots 051039_A1B M_2030_31 5495 147 1961_mdlperf_V1 2 png described in Section 3 7 o Historical time series plots
142. of highest quality data SILO grew from a need for climate data in various formats suitable for systems modelling The SILO datasets are based on historical data provided by the Bureau of Meteorology BoM which DSITI has enhanced by error checking interpolating across Australia on a 5km grid and on this basis infilling missing data at each station Both CF and QM historical data files are named as follows LocationCode_SILO_Latitude_Longitude_VersionNumber SILOformat e g 056002_SILO_ 30 5167_151 6681_V1 2 met e LocationCode is a six digit number BoM station code if patched point i e 051039 or all Zeros if drilled from interpolated surfaces i e 000000 e Obs indicates observed shistorical data e SILO indicates SILO baseline data e Latitude of your station in decimal degrees e Longitude as above e Version Number i e V1 2 as used in CF data QMv2 2 or QMv3 0 as used in QM data e SILO format either met for APSIM or p51 for GRASP Note the different date formats in the APSIM and p51 files e APSIM YYYY DayNumber 1 365 6 e p51 YYYYMMDD An example of information presented in an historical baseline climate data file is presented in Figure 2 6 This format is suitable for use in APSIM Department of Science Information Technology and Innovation APSIM format V1 2 weather met weather Istation number 056002 Istation name None atitude 30 5167 DECIMAL DEGREES long
143. olation of historical quantile trends to 2050 with renormalisation to the GCM mean by using change factors with median sensitivity to global warming refer to Section 5 2 This means that there will be changes in both the mean and the standard deviation of the projected datum 11 8 Non uniformity of perturbations In the CF projections perturbations are applied uniformly across the time series within each month family However in the QM projections the applied perturbations come from three sources being 1 the initial projection onto the target CDF 2 the transport of historical residuals from the projection trends to the projection 3 the final debiasing of the projected sequence to honour the OzClim trends Sources 1 and 3 are uniform but source 2 is not For example a month that is abnormally low relative to its historical quantile trends may end up with negative perturbations after projecting The likelihood of downward perturbations lessens as the projection year or global warming is increased but the data will always be non uniformly perturbed 11 9 Differences between CF and QM projections A comparison between CF and QM projections in summarised in Table1 1 Table 11 1 Comparison between change factor CF V 1 2 and quantile matched QM projections data Property Projections method Change factor Quantile matched Applies shifts to the mean Yes Yes Alte
144. on of rainfall normalisation procedures This update involved recalculation of normalisation parameters and an update to the station dictionary affecting all variables within SILO Furthermore in CF Version 1 1 and 1 2 more accurate daily synthetic pan estimates have replaced the previous average daily pan evaporation values in APSIM files prior to 1970 11 5 Quality control measures Additional quality control measures were applied to each climate variable in the CF Version 1 1 and 1 2 data over and above those applied in CF Version 1 so that application of change patterns 94 Consistent Climate Scenarios User Guide Version 2 2 will not produce climate projections values outside the bounds of what may be reasonably expected The additional quality control measures include clamping to calculated thresholds if e the projected vapour pressure is greater than the saturated vapour pressure at maximum temperature reset to this value and e the projected solar radiation is greater than the maximum clear sky radiation for the day reset to this value 11 6 Calculation of pan evaporation SILO can deliver a number of different estimates of potential evaporation The historical data supplied in SILO files for APSIM met and GRASP p51 consists of two estimates merged in time Class A pan evaporation estimates either actual or in filled by interpolated Class A pan is supplied for the period 1970 to current Prior to widespread use
145. or most of the climate variables but some of the climate variables will have been pre transformed to limit any non physical projections such Pan evaporation is not computed by Quantile matching but uses the Change factor methodology based on synthetic pan evaporation functions Rayner 2005 using Change factors for vapour pressure and solar radiation 50 Consistent Climate Scenarios User Guide Version 2 2 as negative rainfall or solar radiation in excess of extraterrestrial incoming radiation The pre transformation is also done to normalise the data i e make the regression residuals more nearly normally distributed The QM procedure is applied to the transformed variables An inverse transform is applied to the resulting projected data to generate the projected climate variables Statistically Septembers TMax significant 40 trend p 0 051 35 oO D 2030 pivot 30 quantiles for x o Septembers E SJ 9 26 66 _ 0 5 21 90 20 MARAE PI 0 051 Ae F Z 0 1 16 46 PENATAN AENT 13 These become q 0 1 p 0 414 points on the 2030 1960 1970 1980 1990 2000 2010 2020 z030 September CDF Year Figure 5 1 Quantile trend plot for September daily maximum temperature showing how quantiles 0 1 0 5 and 0 9 are computed for 2030 Step 3 Each historical value is replaced by a value from the future CDF with the same
146. ormation Technology and Innovation Model realisation A single run of a Global Climate Model GCM OzClim A system containing patterns of regional climate change from a selection of Global Climate Models run by CSIRO and other research centres Patched point data The term used where data that are either missing or suspect have been patched with interpolated data Pattern of change The projected mean annual rate of change for a particular climate variable per degree of 21 Century global warming Percentile Percentiles are points taken at regular intervals from the cumulative distribution function CDF of a random variable expressed as a percentage between 0 and 100 The sample 50 percentile is equivalent to the sample 0 5 quantile Perturb To modify i e scale Program for Climate Model Diagnosis and Intercomparison PCMDI A research organisation in the USA who s mission is to develop improved methods and tools for the diagnosis and intercomparison of GCMs that simulate global climate The PCMDI archives the WCRP CMIP3 and CMIP5 Multi Model Datasets Probability distribution function PDF Like the Cumulative distribution function the PDF gives the probability that an observation X is less than or equal to a given value x and gives the percentile or quantile rank of an observation Quantile Quantiles are points taken at regular intervals from the cumulative distribution function CDF of a random variable expres
147. ortal A variation in the QM method to incorporate daily GCM data has been used to calculate QM based projections data for 2050 Version QM2 2 10 2050 released mid November 2011 The QM 2050 projections which were limited to a single GCM due to the availability of raw daily GCM data for 2050 are available via ftp climate mft derm qid gov au Climate Scenarios QM_ 2050 TestData Users are encouraged to consider the limitations of both the CF and QM approaches when interpreting model output based on these climate change projections data Department of Science Information Technology and Innovation Figure 1 1 indicates the relationship between the CCSP partners and their roles in delivering the project outputs Table 1 1 presents the CCSP versioning Figure 1 1 Relationships between Consistent Climate Scenarios Project partners and project outputs In the AR4 CCS projections data 2030 or 2050 represents a period e g 30 years centred on that year The length of the period will be the same as the user s selected SILO historical base period For applications model evaluation we recommend usage of a 1960 2010 base period the quality of post 1960 historical climate data is higher than that of earlier data The period from 1960 to 2010 also encompasses a wide range of natural climate variability i e droughts and floods due to fluctuations in the El Ni o Southern Oscillation phenomenon ENSO as well as opposite
148. projections data If you need to access this document in a language other than English please call the Translating and Interpreting Service TIS National on 131 450 and ask them to telephone Library Services on 61 7 3170 5725 Citation Content from this document should be attributed as The State of Queensland Department of Science Information Technology and Innovation Consistent Climate Scenarios User Guide Version 2 2 2015 Acknowledgements The Consistent Climate Scenarios Project CCSP was initially undertaken by the former Queensland Climate Change Centre of Excellence QCCCE now under the Queensland Government s Department of Science Information Technology and Innovation DSITI DSITI acknowledges that the development of the Consistent Climate Scenarios CCS User Guide and the daily climate projections data referred to herein was supported by funding from the Australian Government Department of Agriculture Forestry and Fisheries DAFF under the Australia s Farming Future Climate Change Research Program CCRP DSITI also acknowledges guidance provided by Dr Stephen McMaugh DAFF and the project s Expert Panel chaired by Mr Steven Crimp CSIRO The authors also thank Dr lan Smith formerly CSIRO for his assistance in reviewing this document Furthermore this User Guide has benefitted from feedback and advice of data users in particular from the CCRP project teams who have applied CCS projections data in their projects
149. r additional orders CF ZIP archives containing daily projections CF climate projections data files as well as historical baseline climate data files monthly multiplier files CO2 matching files and log warnings files for each selected climate site are contained in a ZIP format archive named as follows User name_JobNumber_FileLabel_Archivetype e g john smith_its138_FileLabel_zip exe Username derived from your email address JobNumber i e its138 its139 its140 etc FileLabel 1 8 character label specified by the user i e cf2030 testdata etc Archivetype i e zip exe for Windows or zip for Unix Linux File types packaged in CF ZIP archives containing daily projections data including filename examples are o CF projections 051039_A1F1_2030_M_CSIRO MK35_ 31 5495_147 1961_V1 2 met details in this Section o Historical baseline climate data 051039_SILO_ 31 5495_147 1961_V1 2 met described in Section 2 3 o Multiplier files 051039_ 31 5495_ 147 1961_V1 2 multiplier described in Section 3 1 o CO2 matching files 051039_ 31 5495_147 1961_NamesList txt described in Section 3 2 o Log warning files 051039_ 31 5495_147 1961_V1 2 log described in Section 3 4 Consistent Climate Scenarios User Guide Version 2 2 CF ZIP archives containing diagnostic plots CF 2030 diagnostic charts which include comparison of model projections plots for 2030 based on the A1B emissio
150. rameters Choose Change factor CF or Quantile matching QM projections the historical baseline period start and end years the projections year 2030 or 2050 and choice of up to three emissions scenarios up to three global warming sensitivities and up to 23 climate models e Projections data are always packaged with the following ancillary information which can be used with or independently of the projections data _historical baseline climate data files multiplier files containing the information used to calculate change factors comparison of model projections plots showing change in rainfall and temperature at 2030 for 19 GCMs based on change factors a CO matching file look up table log warning files only with CF orders Step 4 Select delivery details Enter a unique label to assist in identification of the ZIP archive that will be created for your projections data files Then select the projections data file format APSIM or p51 option to include diagnostic charts or not and finally how you wish to receive the data FTP is preferred e If diagnostic charts is selected the user will receive the following which can be used with or independently of the projections datasets historical time series plots of histograms showing historical and quantile matched distributions for each climate variable only with QM orders monthly quantile trend plots for s
151. re C Point Temperature C D aes Reproduced by permission of CSIRO Australia CSIRO Figure 4 1 Steps to producing OzClim patterns of change for an individual month and single Global Climate Model GCM Panel A is the annual global temperature for the GCM Panel B is the monthly temperature time series for an individual grid point Panel C is the combination of Panel A and Panel B with a regression line from which the annual rate of change per degree of global warming slope for that month is calculated for a specific grid point Panel D displays the slope for each grid point across Australia following interpolation to a finer scale across Australia after CSIRO 2010 2 W Av Baseline Regional pattern Global Result 2060 Climatology from GCM warming Reproduced by permission of CSIRO Australia CSIRO Figure 4 2 Climate scenario generation for annual temperature in 2060 The regional pattern from a GCM is multiplied by the global warming for 2060 and added to the observed baseline climatology i e OzClim has used 1975 2004 In this example the GCM is CSIRO Mk3 0 the emissions scenario is SRES A1FI and the rate of global warming is high after CSIRO 2010 Further information about issues related in the calculation of change factors including downscaling from Global Climate Models and the calculation of trends per degree of global warming is discussed in Section 9 and 10
152. red Cumulative frequency distribution No Yes where p value is significant Applies changes to the standard deviation No Yes where p value is significant Carries historical trends Yes Only if p value not significant Historical base line climate 1960 to 2010 1957 to 2010 Available projections years 2030 and 2050 2030 and 2050 11 9 Changes between QMV2 2 0 and QMV3 0 In 2014 an insignificant error was found in QM 2030 daily rainfall projections data produced by the QMV2 2 0 code when developing the code for the QM with bootstrapping Other climate variables were not affected In testing the new QMV3 0 code it was found that a small proportion of QMV2 2 0 generated data were affected and only to an extremely minor extent 80 individual daily rainfall instances out of 912 500 days Of those data the average error in daily rainfall was 0 1 mm with the largest error being 0 3 mm Therefore projections data based on both codes are almost identical In fact improvements to raw SILO climate data over time would more likely generate much larger changes Output based on the QMV3 0 code represents a slight improvement from that of QMV2 2 0 as it integrates the original QM methodology outlined in Section 5 of this User Guide with a modified QM method called Q5 that incorporates bootstrapping Q5 is currently as at May 2015 being developed for AR4 and then AR5 projections data The Q5 methodology is to be addressed in a separate CCS User
153. rms applied Some climate variables are transformed before applying the quantile matched projection The currently used transforms and their inverses are provisional only and may be later changed Maximum and minimum temperature No transforms Evaporation Evaporation is computed using synthetic pan evaporation formulae described in Rayner 2005 and no transforms have been applied However clamping is applied after having quantile matched the projection to ensure that evaporation values are not negative Rainfall Only positive rainfall data are projected Positive rainfall data are transformed by taking the cube root The immediate goal is to reduce the leverage of the high rainfall events in the regressions Post projected data is then back transformed to rainfall by cubing This is followed by a procedure which is consistent with the CF based projections clamping algorithm to ensure that all rainfalls are positive Vapour Pressure Vapour pressure VP has been transformed to specific humidity SH using WMO 2008 formulae and then transformed back to VP post projection The immediate goal is to reduce the spread of data to decrease leverage of any outliers A final clamping post projection occurs to ensure that vapour pressures are strictly positive The lower cut off is arbitrarily chosen as the saturation VP at the lowest recorded temperature in the Australian observed record 23 0 C 29 June 1994 Charlotte Pa
154. rse to trends in cloudiness as expected It is also consistent to find that the trends are inversely proportional to the top of atmosphere radiation Lastly an analysis of multiple regression residues showed that the winter months are not as well estimated as other months 9 5 Summary of infilling In summary of the 19 GCMs currently available in the Consistent Climate Scenarios project no infilling of trends per degree of 21 Century global warming was required in order to compute change factors for the associated rainfall projections data However several GCMs did not have trends per degree of 21 Century global warming for specific climate elements which include solar radiation humidity and maximum and minimum air temperature so the DSITI has adopted multiple regression techniques to formulate these missing trends Of those GCMs that had missing trends per degree of 21 Century global warming 10 GCMs required infilling for maximum and minimum temperature and five GCMs required infilling for humidity Less infilling was required for other climate variables Further information related to the infilling for individual GCMs is presented in Table 9 10 84 Consistent Climate Scenarios User Guide Version 2 2 Table 9 10 State of infilling for missing trends per degree of global warming for specific climate variables at 30 May 2012 GLOBAL CLIMATE MODELS AND INFILLING as at 30 May 2012 Expert Review NOTES GCM na
155. s of daily data around 2050 e Placeholder columns in the QM 2050 met files for the missing vapour pressure pan evaporation and solar radiation have been infilled with projections data for 2050 based on the QM 2030 methodology using extrapolation of historical quantile trends to 2050 with renormalisation to the GCM mean by using change factors with median sensitivity to global warming The QM 2050 pan evaporation has been computed using Rayner 2005 synthetic pan evaporation functions using vapour pressure as an input 5 3 Post projection clamping The projection method can result in occasional non physical values i e values that will not occur in reality such as negative rainfalls evaporations or vapour pressures or radiation data that exceed incoming solar radiation The occurrence of such non physical values is reduced by pre transforming the data but some non physical values may still occur Where such data occurs it is corrected by a brute force clamping back to the physical range In the QM datasets currently supplied such values are rare In the prototype data most of the 312 variables stations showed no 53 Department of Science Information Technology and Innovation clamping Of the 35 variables stations that required clamping the number of clamped values was less than 87 in 19723 days lt 0 4 apart from four cases where the counts were 115 0 58 255 1 3 146 0 74 and 2 6 5 4 Transfo
156. s of a given GCM forced by a specific emissions scenario Step 1 Extract global annual averages of 21 Century surface temperature projections Figure 4 1a Step 2 For an individual GCM model s grid point typically 150 250km extract monthly projections of a given climate variable e g temperature or rainfall Figure 4 1b Step 3 Using a simple linear regression approach compute the linear trend slope between the global annual average surface temperature Step 1 and the average monthly values of a given climate variable for that grid point Step 2 The calculated slope for each grid point represents the mean rate of change of that climate variable per degree global warming Figure 4 1c Step 4 Repeat the above steps for each grid point across Australia to produce the projected pattern of change per degree of 21 Century global warming across Australia for that climate variable Figure 4 1d Step 5 As GCMs vary in their spatial resolution interpolate the pattern of change obtained in Step 4 to a finer regional scale It should be noted that representing GCM output on acommon 25km x 25km base provides utility in terms of application but does not imply increased accuracy over the native resolution of the GCM outputs Patterns of change for specific locations are available in the monthly multiplier ancillary files supplied with CF data Step 6 Calculate change factors by applying the above pattern of
157. sed as a fraction between 0 and 1 The sample 0 5 quantile is equivalent to the sample 50 percentile Quantile matching QM QM maps historical values to projected values with the same quantile in each class Quantile matching QM approach QM is more sophisticated than the CF approach in that it considers not only change factors but also the projected changes in the cumulative distribution function of the climate projections Queensland Climate Change Centre of Excellence QCCCE QCCCE existed from April 2009 to April 2012 as an Australian state based climate science research centre within the Office of Climate Change under DERM DERM undertook targeted research to deliver specialised information to inform Queensland s response to climate change climate variability and climate extremes QCCCE provided information and science on climate change impacts including the application of international research and science to the Queensland context In May 2012 QCCCE transferred to DSITIA Scenario A plausible description but not a predication of how the future may develop based on a coherent and internally consistent set of assumptions about key relationships and driving forces i e technological and economic change SILO A climate database hosted by the Department of Science Information Technology and Innovation DSITI containing Australian climate data from 1889 current to yesterday in a number of ready to use formats suitable for
158. semble number e g 01 starting point 1962 02 1963 03 1968 04 1972 05 1975 06 1985 07 1986 08 1996 09 2001 10 2009 e File type transient climate data wth The transient climate change files have been constructed for a single location Brigalow Research Station Central Queensland and formatted for use in the CENTURY model as proof of concept prior to further systems development The transient climate change files include historical data back to 1889 and CF based projections data to 2100 The test data are for the same range of seventeen GCMs eight emissions scenarios and three climate warming sensitivities as the CF Version 1 1 2030 and 2050 projections datasets Ten ensemble members with starting years between 1962 and 2009 are provided The data files contain the projections year and for each month precipitation total mean daily minimum and mean daily maximum temperatures and at the end of each data row the source year the index 1 10 to the analogue start year and the expected degrees of global warming C These extra data fields are not read by the CENTURY model and are for diagnostic purposes in this data set 43 Department of Science Information Technology and Innovation An example of information presented in a transient climate change data file is presented in Figure 3 14 prec 2098 7 44 5 72
159. sites and projections scenarios 86 Consistent Climate Scenarios User Guide Version 2 2 Table 10 1 Analysis of SILO historical and projected 2050 July climate data for the University of Queensland Gatton Location Code 040082 showing changes in climate statistics for several climate variables due to the use of different historical base periods Statistics will differ for other sites and projections scenarios July Base period Mean Mean daily Mean daily Mean days with monthly maximum minimum frost per month Rainfall temperature temperature screen minimum less than 2 0 C SILO historical 1889 2010 122 yrs 45 6mm 20 5 C 5 9 C 3 8 data 1961 2010 50 yrs 45 9mm 20 8 C 6 2 C 1 5 1980 2010 30 yrs 36 0mm 21 1 C 6 6 C 0 7 2050 projections 559 0010 122 yrs 39 1mm_ 22 6 C 8 42C 0 5 MIROC H GCM A1F1 emissions 1961 2010 50 yrs 39 3mm 22 9 C 8 7 C 0 3 scenario and high climate warming 1980 2010 30 yrs 30 9mm 23 2 C 9 1 C 0 1 sensitivity Notes e Inthe CF Version 1 1 data climate change factors were applied to a measured historical sequence from 1889 2010 However in the web based CF Version 1 2 the historical baseline is restricted to data from 1960 onwards e For applications model evaluation we recommend usage of a 1960 2010 base period as the quality of post 1960 historical climate data is higher e The use of long i e multi deca
160. ss NSW Solar radiation Solar radiation is first scaled by the extra terrestrial incoming radiation to produce a ground level proportion of extra terrestrial radiation between 0 and 1 inclusive This proportion is transformed by the logit transform to allow values to range over negative infinity to positive infinity The post projected data is back transformed by the inverse of the logit function the logistic function The goal of these transforms is to confine the projected proportional data to the 0 1 interval with the main effect at the high and low end of the range The proportion data is then restored to unscaled solar radiation data No clamping is needed as the transform projection back transform sequence is guaranteed to return values in the physical range 54 Consistent Climate Scenarios User Guide Version 2 2 6 Description of daily climate variables 6 1 SILO data The SILO historical data sets to which the OzClim scaling approach has been applied to produce the 2030 and 2050 projections data include the following six daily climate variables useful for biological modelling Rainfall mm for the 24 hours to 9am on the date listed Maximum air temperature C for the 24 hours from 9am on the date listed Minimum air temperature C for the 24 hours to 9am on the date listed Solar radiation megajoules per square metre on the date listed Vapour pressure hectopascals at 9am on the date listed Pa
161. st the saturated vapour pressure using WMO 2008 recommended functions without pressure corrections e Number of clamped projected radiation greater than ET radiation Instances where the projected radiation are clamped to 0 81 of the calculated extra terrestrial maximum solar radiation on a horizontal surface and clear sky radiation e Warning Potential or Implausible Extreme Percent Change Rate Month affected Climate element affected i e rain Change from the historical baseline climate as a percentage change for rain radiation relative humidity and pan evaporation but an absolute change for temperature e g for rain a change of 75 means a 75 percent decline The thresholds for which log warnings are assessed are presented in Table 3 1 28 Consistent Climate Scenarios User Guide Version 2 2 Table 3 1 Thresholds for which log warnings are assessed Element Warning Values Warning High value Warning Low value Implausible High value Implausible Low value gt 50 lt 50 gt 90 lt 90 RESET to 90 gt 50 lt 90 RESET to 90 lt 90 RESET to 90 lt 3 C Radiation Relative gt 50 Humidity Maximum gt 7 C Temperature Minimum gt 7 C lt 3 C Temperature Pot gt 50 Evaporation lt 90 RESET to 90 A snapshot of information contained in a log warning ancillary data f
162. stribution file e VersionNumber v1 1 represents CF data 32 Consistent Climate Scenarios User Guide Version 2 2 Change in frequency is greatest where there are large differences between the observed coloured and projected black points on the plots A snapshot of a frequency distribution plot is presented in Figure 3 7 Model CSIRO MK35 SRES A1B SENSITIVITY H YEAR 2050 5 4 z g 3 c E w w 3 m Cc co gt a a 1 0 Minimum Temperature C Maximum Temperature C CCCS PROCESSING VERSION v1 1 OBSERVATIONS BASE PERIOD 1960 2009 5 8 e Observed coloured 4 Projected 2050 6 Dois S S 4 3 3 o gt cr a a 2 1 o o 8 10 20 30 40 50 Rainfall mm day Solar Radiation MJ day Location 051039 At 31 5495_147 1961 1S g 10 g g c A w wv 3 3 o o w w 5 te 2011 11 22 Vapour Pressure Hpa Pan Evaporation mrm day Figure 3 7 Frequency distribution plots observed with 2050 projections for Nyngan Airport LocationCode 051039 in South Australia filename CSIRO MK35_A1B_H_2050_051039_ 31 5495 _147 1961_fdist_v1 1 gif from a run of the CSIRO Mk3 5 GCM forced by the A1B emissions scenario with high climate warming sensitivity Observed values are coloured projections are shown in black 33 Department of Science Information Technology and Innovation Spikes in pan evaporation frequency distribution plots
163. t changes in extremes daily mean or variance e Trends per degree of global warming are spatially interpolated to higher resolution using bilinear interpolation Topography is not taken into account neither are any other factors Thus as no further information is added small scale differences on small spatial scales may be deceptive 10 7 Issues important to biological modelling The initial scaling approach using change factors has several known limitations which need to be considered when applying the CF data to biological modelling These limitations are important since the application of a simple scaling approach can lead to unintended consequences in some biophysical models especially those which are threshold dependant Users should note that uncertainty and errors in climate projections data can propagate from many sources including e the underlying climate data e GCM models e spatial downscaling from a coarse to fine grid and e temporal down scaling i e from monthly to daily Furthermore additional errors can occur when the computed climate projections data are then applied to biological models due to e biology with inadequate functionality parameterisation and e systems responses management adaptation natural selection adaptation The following caveats related to unrealistic outcomes that can occur with the OzClim change factor approach need to be considered in regard to the usage of c
164. ted climate warming sensitivity including brief recommendations 7 1 Emissions scenarios background information Economic development demographic and technological change play critical roles in the outcome of future greenhouse gas emissions and potentially climate change As such the IPCC documented a range of greenhouse gas emissions scenarios for the future used in both AR3 and AR4 modelling in their Special Report on Emissions Scenarios SRES IPCC 2000 The emissions scenarios are grouped into four categories termed families A1 A2 B1 and B2 each family having a storyline based on specific socio economic and environmental characteristics The six IPCC AR4 emissions scenarios considered in the Consistent Climate Scenarios Project are described in Table 7 1 Additional to the six A and B family SRES emissions scenarios this project considers two CO stabilisation scenarios based on the work of Wigley et al 1996 listed in Table 7 2 Projections to 2100 for all eight scenarios are presented in Figure 7 1 57 Department of Science Information Technology and Innovation Table 7 1 Project adapted from IPCC 2000 IPCC AR4 SRES emissions scenarios used in the Consistent Climate Scenarios SRES Remarks Storyline Energy use Median of projected Emissions CO concentrations Scenario ppm 2030 2050 A1Fl Most recommended Very rapid economic growth Glob
165. the corresponding ancillary files that are contained in each archive This can provide a check that the correct data has been accessed for the Department of Science Information Technology and Innovation specific purpose In some cases the ancillary files may contain the exact information that the end user is interested in 3 Files with additional information Many ancillary files are supplied in the ZIP archives to supplement the climate projections data The information contained in these ancillary files can be used independently from the 2030 50 projections data files Change factor CF based ancillary files include e multiplier files e CO2 matching files e log warning files e historical time series plots e comparison of model projections plots Additional to these quantile matched QM based ancillary files include e monthly quantile trend plots e histograms of projected frequency distributions In addition the following ancillary files not available via the CCS web portal can be made available if requested e an historical and projected CO2 concentrations file e CF based frequency distribution plots e plots of simulated 20th and 21st Century climate according to available GCM runs e asingle station CF based transient climate data test set for 1889 2100 3 1 Multiplier files The data or multipliers contained in these files include e projected amounts of global warming for each emiss
166. the range of CMIP3 GCMs according to global warming sensitivity and East Indian verses West Pacific Ocean temperature responses The rainfall responses which can be split into four Representative Future Climate RFC partitions Figure 8 1 are based on e HI A high level of global warming where the Eastern Indian Ocean warms faster than the Western Pacific Ocean e HP A high level of global warming where the Western Pacific Ocean warms faster than Eastern Indian Ocean e WI A low level of global warming where the Eastern Indian Ocean warms fasters than Western Pacific Ocean e WP A low level of global warming where the Western Pacific Ocean warms fasters than Eastern Indian Ocean Considering the above a mean 21 Century climate response for rainfall and other climate variables has been derived for the CCS project by compositing modelled data based on selected GCMs within each of Watterson s four RFC partitions Figures 8 1 8 2 and 8 3 The groups of GCMs that are being used to produce CCS Composite climate projections data for each RFC are listed in Table 8 3 While the CCS Expert Panel has recommended use of the complete set of 19 GCMs that are available in CCS for climate projections analysis the use of data based on the mean RFC related climate response patterns HI HP WI and WP has been offered as a secondary option where this is not practicable In cases where single GCMs are selected the selection of at
167. thin this document Any decisions made by other parties based on this document are solely the responsibility of those parties Information contained in this document is from a number of sources and as such does not necessarily represent government or departmental policy Some of the pages in this document contain links to pages and or sites which are not under the control of the State of Queensland No representation or warranty is made by the State of Queensland regarding the content of any such pages or sites Merely because links are made to third party sites does not mean that the State of Queensland through the Department of Science Information Technology and Innovation promotes or endorses any of those sites It is possible that adverse consequences including viruses or loss of privacy may result from use of third party sites Furthermore in regard to material or information provided by the CSIRO the CSIRO does not guarantee that the material or information it has provided is complete or accurate or without flaw of any kind or is wholly appropriate for your particular purposes and therefore disclaims all liability for any error loss or other consequence which may arise directly or indirectly from you relying on any information or material it has provided in part or in whole Any reliance on the information or material CSIRO has provided is made at the reader s own risk The same disclaimers that apply to SILO historical data apply to the CCS
168. torf 2008 noted that a doubling of CO from the pre industrial 280 ppm to the 560 ppm estimate for the mid to late 21 Century the classic IPCC range would lead to between 1 5 C and 4 5 C of global warming but added that this range could now have narrowed to 2 C to 4 C The US National Academies 2010 confirm a similar range 2 1 C to 4 4 C with a best estimate of 3 2 C The pre industrial value refers to the period circa 1750 IPCC 2007 Ramsdorf 2008 also states that global 60 Consistent Climate Scenarios User Guide Version 2 2 warming at a particular time depends on the time history of past CO and other forcing changes not just anthropogenic climate change As such climate warming sensitivity may not be constant over time and is usually grouped into three ranges low median and high Climate warming sensitivities used for the projections data The CF projections data provided in the Consistent Climate Scenarios Project utilise climate warming sensitivities low median and high calculated from estimates of global warming due to the effect of global CO2 concentrations at both 2030 and 2050 These climate warming sensitivities are provided for each of the six SRES emissions and two CO stabilisation scenarios listed in Tables 7 1 and 7 2 Note e These climate warming sensitivity options should not be confused with those provided on the OzClim website which refer to the emissions scenarios per se That
169. trend per degree of global warming on a grid cell by grid cell basis over all GCM models using a filter of 50 for a nominal degree of global warming revealed some anomalous large magnitude change factors that would become much larger when scaled by one to four degrees of global warming The anomalies discovered were mainly in rainfall and evaporation trend files The most common effects are e projections of extreme drying trends e unreasonably large increases in wet area evaporation at some locations and e significant errors not only of scale of trend but also potentially of sign of trend leading to a reduction in confidence in projections As an example one GCM model had an implausibly low base line 30 year monthly average rainfall of 7 9 x 10 mm Fig 10 2a The resulting trend per degree of global warming was a 2 6 x 10 percent change from base per degree of global warming Fig 10 2b This corresponds to an original trend of 2 1 mm per degree of global warming which is almost but not quite plausible given an observed monthly average of 2 mm Fig 10 2c Note that wet area evaporation is not used in the final calculation in which changes in pan evaporation are estimated as a function of changes in vapour pressure deficit and changes in solar radiation Model Mean Rainfall monthly total pr r_monthly 7 9x10 mm pr_monthiy am Observed Mean Rainfall r monthly mm a 2 MM w s Trend in Precipitation
170. two COs stabilisation scenarios are uncertain it is possible that they may under or over shoot the proposed trajectories 88 Consistent Climate Scenarios User Guide Version 2 2 10 4 Downscaling from Global Climate Models A typical coarse grained Global Climate Model GCM is computed at a resolution of about 2 giving about 16 200 grid points over the whole globe of which some 350 grid points are likely to occur within the Australian bounding rectangle 10S to 45S 115E to 155E and less over the Australian continental land area Therefore in attempting to produce analyses of future trends over the Australian continent we are constrained to using less than two percent of the available information from GCMs As a consequence this creates increased sample variance and e the range of available GCMs produce widely differing estimates of both simulated 20th Century and projected 21st Century trends in rainfall over Australia and similarly e GCMs also give a diversity of 20th Century Australian continental mean temperature trends We assume the spatial variation in climate variable response is e more related to a GCM itself than individual runs or the forcing scenario and e is approximately linear in response to global warming in the model Therefore we can scale the local climate response according to global warming In this project information obtained from GCM grid points has been interpolated over Australia on a
171. under a high emissions scenario but this outcome is non physical Ininitial Brian Pastures CF Version 0 test data we found that the CSIRO Mk3 5 GCM generated future projections early in the 21 Century containing negative rainfall under a high emissions scenario due to a rapid decline in rainfall per degree of global warming An analysis of other GCMs showed that this problem was widespread across Australia especially in the drier months of the year To avoid this problem in CF Version 1 we have truncated rainfall to a minimum 90 decline so it cannot ever be negative The same constraint is applied to other climate variables that cannot be negative by definition The application of these rules appears in the log warning ancillary files While changes in daily rainfall intensity occur the approach also assumes no change in the occurrence of the number of dry days which would be important to germination harvest and irrigation Simultaneously applying simple changes to daily climate variables may or may not consistent For example wet days tend to be associated with lower than average maximum temperatures and dry days with higher than average maximum temperatures and auniform increase in both daily rainfall and temperature may not fully capture this interdependence Other climate related factors affecting biological response worth consideration are The variability in the CO responses of plants relative to
172. unt of warming as shown on the Y axis of Figure 8 1 is for the globe and will vary on local and regional scales For example the projected change in the mean temperature in Brisbane over the 21 Century will differ from mean global warming Projected climates will also differ due to choices of emissions scenario and sensitivity to global warming 69 Department of Science Information Technology and Innovation 9 Infilling of trends per degree of global warming To produce the CF projections data trends per degree of 21 Century global warming are required for each climate variable of interest for each GCM of interest These trends per degree of global warming are then applied to historical observations to give future projections However in many cases OzClim trends per degree of global warming were not available for specific climate variables Therefore trends for those climate variables were estimated by the Queensland Climate Change Centre of Excellence and then infilled Of the 19 GCMs selected for use in the CCS project only five have a complete set of trends per degree of global warming also called patterns of change from OzClim for all the required climate variables In addition since June 1 2012 trends per degree of global warming have been obtained for two extra highly ranked Hadley Centre GCMs HADCM3 HADGEM1 The seven GCMs for which no infilling was required are listed in Table 9 1 Table 9 1 GC
173. users can select any period from 1960 onwards 1960 to 2010 is the recommended historical baseline Periods of less than 30 years are insufficient for climate change trend analysis The historical time series plots are png files and are typically named LocationCode_Lat_Long_V1 2 png e g 040112 _ 26 5544 151 8456_V1 2 png e LocationCode is a six digit number BoM station code if patched point i e 040112 or all Zeros if drilled from interpolated surfaces i e 000000 e Latitude of the station or location in decimal degrees e Longitude as above e VersionNumber V1 2 represents CF data A snapshot of a historical time series plot is presented in Figure 3 6 30 Consistent Climate Scenarios User Guide Version 2 2 040140 27 1492 152 5781 V1 2 png 97 paily Average Maximum Temperature 5 Daily Average Minimum Temperature 27 0 15 0 o 26 5 o 14 5 v 26 0 o 14 0 H PH 25 13 24 5 7 12 5 24 0 12 Q 1960 1970 1980 1990 2000 2010 1960 1970 1980 1990 2000 2010 Year Year 5 Raily Average Pan Evaporation Per Year 4 8 4 6 E 4 4 4 2 4 0 i 60 1970 1980 1990 2000 2010 Year Year 19 pally Average Solar Radiation Per Year ely Average Vapour Pressure Deficit 60 1970 1980 1990 2000 2010 Eo 1970 1980 1990 2000 2010 Year Year Figure 3 6 Historical time series plot for Mt Brisbane Queensland LocationCode 040140 filename 040140_ 27 1492_152 5781_V1 2 png Annual variability is shown is black The lin
174. uted and this regression model is fitted for the same month to all regression models over all grid points Tables 9 4 and 9 5 present Tasmax and Tasmin regression constants computed by month and applied to Equation 1 Finally the parameters of the function were then applied to all pixels over all available Australian grid points 72 Consistent Climate Scenarios User Guide Version 2 2 Table 9 4 Tasmax regression constants P computed by month and applied to Equation 1 Climate variables are cloud cover clt precipitable water orw and mean temperature at surface tas Month tas prw clt tas clt tas prw P1 P2 P3 P4 P5 Jan 1 0750 0 0001 0 0845 0 0379 0 0238 Feb 1 0444 0 0128 0 0514 0 0155 0 0017 Mar 1 0844 0 0096 0 0686 0 0258 0 0221 Apr 1 0171 0 0245 0 0465 0 0032 0 0004 May 0 9986 0 0311 0 0460 0 0091 0 0208 Jun 0 9903 0 0545 0 0378 0 0205 0 0395 Jul 1 0049 0 1167 0 0302 0 0293 0 0818 Aug 1 0535 0 0225 0 0096 0 0327 0 0188 Sep 1 0077 0 0218 0 0462 0 0036 0 0199 Oct 1 0778 0 0037 0 0485 0 0157 0 0305 Nov 1 0786 0 0044 0 0818 0 0323 0 0242 Dec 1 0925 0 0110 0 0641 0 0213 0 0200 Table 9 5 Tasmin regression constants P computed by month and applied to Equation 1 Climate variables are cloud cover clt precipitable water orw and mean temperature at surface tas Month tas prw clt tas clt tas prw P1 P2 P3 P4 P5 Jan 0 9325 0 0018 0 0455 0 0075 0 0261 Feb 0
175. uted from the following equation rsds_toc clt_ave P1 P2 clt_trend P3 pr_tpc P4 clit_trend P5 pr_tpc P6 prw_tpc Ra where clt total cloud fraction pr precipitation prw precipitable water Ra reciprocal top of atmosphere radiation _tpc percentage trend per degree of global warming ave model average of present 1975 2004 trend trend per degree of global warming Multiple regression The next step taken to estimate solar radiation was drawn from the GCMs at model native resolution All data points over Australia for each of the climate variables from each Global Climate Model were extracted The data was separated into months and the above regression equation was fitted to all of the data points for that month pooling all the available points from all of the GCMs The result was a set of regression coefficients for each month Table 9 9 Additionally the multiple linear regression was performed on a per GCM basis to allow us to test the strength of the regression model predictions Figure 9 8 Table 9 9 Regression parameters found for each of the parts of the regression model Month Clt_ave Clit_ave clt clt_ave pr clt Ra Pr Ra Prw Ra Residue Jan 0 0003 0 0032 0 0007 14 4229 0 1190 0 9620 543 1265 Feb 0 0004 0 0017 0 0019 12 0832 0 7943 1 2286 948 7009 Mar 0 0097 0 0006 0 0008 10 7869 0 2588 2 1887 555 5375 Apr 0 0033 0 0000 0 0001 17 0826 0 5170 1 5816 549 2540 May 0 0052 0
176. vernment Department of Agriculture Fisheries and Forestry DAFF An agency of the Australian Government responsible for developing and implementing policies and programs that ensure Australia s agricultural fisheries food and forestry industries remain competitive profitable and sustainable Australian Grassland and Rangeland Assessment by Spatial Simulation AussieGRASS A leading Australian climate and biological modelling system run by the Department of Science Information Technology and Innovation DSITI using advanced spatial simulation techniques and super computing facilities http www longpaddock qld gov au about researchprojects aussiegrass Bureau of Meteorology BoM An agency of the Australian Government responsible for providing weather services to Australia and surrounding areas Carbon dioxide CO2 A naturally occurring gas which is also a by product of burning fossil fuels and biomass as well as land use changes and other industrial processes Changes in CO2 concentrations have been linked to changes in the earth s temperature Change factor The change in the climatological mean of a specific climate variable e g temperature between the current climate defined in terms of a suitable 20th Century base period and a projected time in the future for example the 30 years centred on 2050 Change factors are based on the amount of global warming at a future point in time and the 21st Century pattern or r
177. warm and 1999 to cool Intergovernmental Panel on Climate Change IPCC The leading international body for the assessment of climate change established by the United Nations Environment Programme UNEP and the World Meteorological Organization WMO to provide the world with a clear scientific view on the current state of knowledge in climate change and its potential environmental and socio economic impacts La Ni a La Ni a represents the cool phase of the ENSO cycle the opposite of an El Ni o and is sometimes referred to as a Pacific cold episode The large scale periodic cooling of the central and east central tropical Pacific results in changes in the atmosphere that affect weather patterns across much of the Pacific Basin including Australia During La Ni a episodes the SOI is positive due to higher than average air pressure at Tahiti and lower than average pressure at Darwin Logit transformation Used with logit regression where the predicted values for the dependent or response variable will never be less than or equal to 0 or greater than or equal to 7 regardless of the values of the independent variables Long Paddock An award winning website providing climate information and outlooks to improve understanding and management of climate and variability in Queensland Model ready Data in a ready to use format i e can be implemented into a modelling system without further adjustment 99 Department of Science Inf
178. wind effects and does have areas of biases but is much better than using long term average data for the pre 1970 period In the future it is hoped to produce an improved synthetic pan estimate by incorporating wind run from GCM re analysis data fields as an additional predictor variable 11 7 Projected CF and QM means and standard deviations To calculate the CF 2030 and 2050 projections change factors are applied uniformly to each of the historical daily datum This means that while the means of the CF projected datum i e for 2030 or 2050 will differ from the historical datum the standard deviations of the CF projected datum will remain the same as that of the historical baseline data In calculating the QM 2030 projections the method takes into account the variance implied by the target 2030 CDF refer to Section 5 1 This means that we should expect changes in both the QM projected means and standard deviations In calculating the QM 2050 projections for rainfall and temperature the method uses raw daily GCM data refer to Section 5 2 This means that there will be changes in both the QM 2050 projected means and standard deviations QM 2050 projections are limited to the ECHAM 5 GCM A1B emissions scenario and median climate sensitivity QM 2050 projections for vapour pressure 95 Department of Science Information Technology and Innovation pan evaporation and solar radiation are based on the QM 2030 method using extrap
179. wnwelling short wave radiation and mean surface air temperature used to calculate trends in vapour pressure Version 1 Trends in Morton s wet area potential evaporation used to calculate trends in pan evaporation Version 1 1 amp 1 2 Trends in downvwelling short wave radiation and vapour pressure used to calculate trends in pan evaporation Version 1 1 amp 1 2 Trends in total cloud fraction reciprocal top of atmosphere radiation precipitation and precipitable water used to calculate trends in solar radiation Trends in surface relative humidity wind speed and ocean variables sea surface air temperature and ocean salinity are avaialable in OzClim but have not been used in the Consistent Climate Scenarios project to date The INMCM GCM previously rated likely to be less reliable is no longer recommended for use due to unstable drift in the model The rank for this model will be adjusted 85 Department of Science Information Technology and Innovation 10 Known limitations of CF projections data This section of the User Guide provides users with information about known limitations of the CF data projections data accompanied by some caveats and guidelines It remains the user s responsibility to fully evaluate inputs to and outputs from biophysical models 10 1 Base period selection The CF projections data have not been de trended in any way and users of the data need to be aware that their choice of historical bas
180. ximum and minimum temperatures in GCMs for which these data are not archived Queensland Climate Change Centre of Excellence Queensland Government CSIRO Mathematics Informatics and Statistics 19 International Congress on Modelling and Simulation MODSIM Perth Australia 12 16 December 2011 http www mssanz org au modsim2011 F5 ricketts pdf Ricketts J H Kokic P N and Carter J O 2013 Consistent Climate Scenarios projecting representative future daily climate from global climate models based on historical data 7Department of Science Information Technology Innovation and the Arts Queensland Government CSIRO Mathematics Informatics and Statistics Australia 20 International Congress on Modelling and Simulation MODSIM Adelaide Australia 1 6 December 2013 http www mssanz org au modsim2013 L1 1 ricketts pdf Ricketts J and Page C M 2007 A web based version of OzClim for exploring climate change impacts and risks in the Australian region In L Oxley and D Kulasiri eds MODSIM 2007 International Congress on Modelling and Simulation November 2007 Modelling and Simulation Society of Australia and New Zealand http www mssanz org au MODSIM07 papers 10_s61 AWebBasedVersion s61_Ricketts pdf Schmidt M Lipson H 2009 Distilling Free Form Natural Laws from Experimental Data Science Vol 324 no 5923 pp 81 85 Smith and Chiew F 2009 Document and assess methods for generating
181. xpressed in relative terms i e as percentage change per degree global warming To produce a change factor the pattern of change i e projected change per degree global warming for a given climate variable is then multiplied by the warming curve from each of the eight SRES scenarios and climate warming sensitivities established using MAGICC This change factor can then be applied to a suitable 20 Century baseline climatology to produce projections of a given climate element The patterns of change used in the Consistent Climate Scenarios project have been obtained from OzClim patterns of change files and raw GCM files provided by CSIRO Marine and Atmospheric Research CMAR The source files for the patterns of change computations are monthly averages of GCM derived model output for individual climate variables and for specific runs Further detail on the use of patterns of change to calculate estimates for specified climate variables at a particular point in the future is described in detail in several papers Mitchell et al 1999 Mitchell 2003 Whetton et al 2005 and Ricketts 2009 The pattern scaling approach employed by CSIRO 2010 and adopted for use in this project can be summarised as follows gt A GCM grid point which is typically 150 250km is the essentially the same as a grid cell 46 Consistent Climate Scenarios User Guide Version 2 2 Using output from individual run
182. zing atmospheric COz on global climate in the next two centuries Geophysical Research Letters Vol 28 No 23 pp 4511 4514 Department of Science Information Technology Innovation and the Arts Queensland Government SILO Enhanced Meteorological Datasets from http www longpaddock gld gov au Ssilo Intergovernmental Panel on Climate Change 2000 Special Report on Emissions Scenarios retrieved from http www ipcc ch pdf special reports spm sres en pdf Intergovernmental Panel on Climate Change 2007 IPCC Fourth Assessment Report retrieved May 23 2011 from http www ipcc ch publications and_data ar4 wg1 en somsspm human and html Jeffrey S J Carter J O Moodie K M and Beswick A R 2001 Using spatial interpolation to construct a comprehensive archive of Australian climate data Environmental Modelling and Software Vol 16 4 pp 309 330 Kokic P Jin H and Crimp S 2012 Statistical Forecasts of Observational Climate Data Extended Abstract International conference on Opportunities and Challenges in Monsoon Prediction in a Changing Climate OCHAMP 2012 Pune India 21 25 February 2012 Mitchell J F B T C Johns M Eagles W J Ingram and R A Davis 1999 Towards the Construction of Climate Change Scenarios Climatic Change 41 3 547 581 Mitchell T D 2003 Pattern Scaling An Examination of the Accuracy of the Technique for Describing Future Climates Climatic Change 60 3 217 242

Consistent Climate Scenarios User Guide- Version 2.2

Contents

Download Pdf Manuals

Related Search

Related Contents