RegCM Version 3.0 User's Guide
Contents
1. Hostetler S Giorgi F Bates G and Bartlein P 1994b Lake atmosphere feedbacks associated with paleolakes bonneville and lahontan Science Hsie E Anthes R and Keyser D 1984 Numerical simulation of frontogenisis in a moist atmosphere Journal of Atmospheric Sciences Kato H Hirakuchi H Nishizawa K and Giorgi F 1999 Performance of ncar regem in the simulation of june and january climate over eastern asia and the high resolution effect of the model Journal of Geophysical Research Kiehl J Hack J Bonan G Boville B Breigleb B Williamson D and Rasch P 1996 Description of the ncar community climate model ccm3 Technical report National Center for Atmospheric Research Kuo S L 1974 Further studies of the parameterization of the effect of cumulus con vection on large scale flow Journal of Atmospheric Sciences Leung L Ghan S Zhao Z C Luo Y Wang W C and Wei H L 1999 Inter comparison of regional climate simulations of the 1991 summer monsoon in eastern asia Journal of Geophysical Research Liu Y Q Avissar R and Giorgi F 1996 A simulation with the regional climate model regcm2 of extremely anomalous precipitation during the 1991 east asian flood An evaluation study Journal of Geophysical Research Liu Y Q Giorgi F and Washington W 1994 Simulation of summer monsoon climate over east asia with ncar regional climate model Monthly We2. The other primary changes are in the areas of cloud and precipitation processes The original explicit moisture scheme of Hsie et al 1984 has been substituted with a simpli fied version of it This is because the original scheme was computationally too expensive to be run in climate mode In the simplified scheme only a prognostic equation for cloud water is included which accounts for cloud water formation advection and mixing by tur bulence re evaporation in sub saturated conditions and conversion into rain via a bulk autoconversion term The main novelty of this scheme does not reside of course in the simplistic microphysics but in the fact that the prognosed cloud water variable is directly used in the cloud radiation calculations In the previous versions of the model cloud water variables for radiation calculations were diagnosed in terms of the local relative humidity This new feature adds a very important and far reaching element of interaction between the simulated hydrologic cycle and energy budget calculations Finally an important aspect of model development was the inclusion of a stretched grid model configuration by which the model horizontal resolution is relatively coarse in the lateral buffer zone and increases towards the interior of the domain Preliminary experiments using an adiabatic version of the model in stretched grid mode are presented by Qian et al 1999 However the stretched grid option is not available with
3. 0 5 2 0 0 02 100 1000 3000 0 08 0 28 11 0 35 0 0 0 10 0 0 150 0 5 2 0 0 02 100 1000 3000 0 17 0 34 12 0 00 0 4 0 01 0 0 200 2 0 0 02 100 1000 3000 0 80 0 60 13 0 80 0 0 0 03 0 0 0 5 2 0 0 02 100 1000 3000 0 06 0 18 14 0 00 0 0 0 0004 0 0 200 0 0 2 0 0 02 100 1000 3000 0 07 0 20 15 0 00 0 2 0 0004 0 0 200 0 0 2 0 0 02 100 1000 3000 0 07 0 20 16 0 80 0 3 0 10 0 0 0 02 100 1000 3000 0 05 0 23 17 0 80 0 2 0 10 0 0 0 02 100 1000 3000 0 08 0 28 18 0 80 0 4 0 80 0 0 0 06 100 2000 3000 0 06 0 24 19 0 80 0 4 0 3 0 0 120 0 5 2 0 0 02 100 2000 3000 0 06 0 18 20 0 80 0 3 0 0 120 0 5 2 0 0 02 100 2000 3000 0 06 0 18 The Terrain program horizontally interpolates the landuse and elevation data from a latitude longitude grid to the cartesian grid of the chosen domain RegCM currently uses the Global Land Cover Characterization GLCC datasets for the vegetation landuse data The GLCC dataset is derived from 1 km Advanced Very High Resolution Ra diometer AVHRR data spanning April 1992 through March 1993 and is based on the vegetation land cover types defined by BATS Biosphere Atmosphere Transfer Scheme The 18
4. Running the script will e Create soft links to the domain file and initial and boundary conditions files fort 10 Input DOMAIN fort 7x Input ICBCYYYYMMDDHH e Create the sub directory output where the model output files are written e Create the postproc in file which will be needed for postprocessing the output files This is discussed in the next section e Compile the source code and run the simulation Running the model generates the following output files Atmospheric model output see Table 8 ATM YYYYMMDDHH Land surface model output see Table 9 SRF YYYYMMDDHH Radiation model output see Table 10 RAD YYYYMMDDHH and a file used to restart the simulation SAVTMP YYYYMMDDHH 25 5 Post processing The postprocessor takes the model output files and generates new output files of averaged variables in commonly used formats such as NetCDF or GrADS You can modify the postproc in file to specify which of the following files to generate e Atmospheric Output see Table 8 for list of variables ATMYYMM NC all atmos variables and times ATMYYMMAVG NC all variables atmos averaged ATMYYMMDIUR NC diurnal cycle averages of atmos variables ATMYYMMCONT NC continual averages of atmos variables e Land Surface Model Output see Table 9 for list of variables SRFYYMM NC all surface model variables and times SRFYYMMAVG NC all variables surface model averaged SRFYYMMDIUR NC diurnal cycle averages of sur
5. amp 11 A Radiation Sch me seiso dg ea a Oh ghee ON A poke as a eA a 11 2 2 2 Land Surface Model 522 aan a 11 2 2 3 Planetary Boundary Layer Scheme 12 2 2 4 Convective Precipitation Schemes 12 2 2 5 Large Scale Precipitation Scheme 17 2 2 6 Ocean flux Parameterization 22 2 18 2 2 7 Pressure Gradient Scheme 4 0 n Se ee Se i ESE 18 2 2 8 Lake Model e is de do e a Oe ed At o AS gf 18 2 2 9 Tracer Model ems rs DAA DARA O d 19 3 Pre Processing 20 Bell E A 21 3 2 JOBE A OF A ah he cad Arta aoe ef oe a Rae aaa he ede ee tas ae 23 3 2 1 Sea surface temperature e 2 Las e e a 23 3 2 2 Initial and Boundary Conditions 23 4 RegCM 25 5 Post processing 26 6 Practice Run 27 6 1 Getting the model code and data adeudo ear te o 4a tre 27 6 2 Pre processihe A oa ee os ed a a ee ee a ee 27 6 2 1 Setting up the domain Gears ir woe ta ale oes 27 022 JOBO airia 4 p pas a p e G a hie a ae a a Gg 29 6 3 Running the Model aaa ia a a a e 29 6 3 1 Restarting the model or A Pass 30 List of Figures 1 Schematic representation of the vertical structure of the model This ex ample is for 14 vertical layers Dashed lines denote half sigma levels solid lines denote full sigma levels Adapted from the PSU NCAR Mesoscale Modeling System Tutorial Class Notes and User s Guide 6 Schematic representation showing the horizontal Arak
6. and rain water or snow mixing ratios 10 2 2 Physics 2 2 1 Radiation Scheme RegCM3 uses the radiation scheme of the NCAR CCM3 which is described in Kiehl et al 1996 Briefly the solar component which accounts for the effect of Oz H20 CO and O follows the 6 Eddington approximation of Kiehl et al 1996 It includes 18 spectral intervals from 0 2 to 5 um The cloud scattering and absorption parameterization follow that of Slingo 1989 whereby the optical properties of the cloud droplets extinction optical depth single scattering albedo and asymmetry parameter are expressed in terms of the cloud liquid water content and an effective droplet radius When cumulus clouds are formed the gridpoint fractional cloud cover is such that the total cover for the column extending from the model computed cloud base level to the cloud top level calculated assuming random overlap is a function of horizontal gridpoint spacing The thickness of the cloud layer is assumed to be equal to that of the model layer and a different cloud water content is specified for middle and low clouds 2 2 2 Land Surface Model The surface physics are performed using BATS1E Biosphere Atmosphere Transfer Scheme which is described in detail by Dickinson et al 1993 BATS is a state of the art surface package designed to describe the role of vegetation and interactive soil moisture in modify ing the surface atmosphere exchanges of momentum energy and
7. 21 GG Parameter Max fractional vegetation cover Difference between max fractional vegetation cover and cover at 269 K Roughness length m Displacement height m Min stomatal resistence s m Max Leaf Area Index Min Leaf Area Index Stem amp dead matter area index Inverse square root of eaf dimension m Light sensitivity factor m W Upper soil layer depth mm Root zone soil ayer depth mm Depth of total soil mm Soil texture type Soil color type Vegetation albedo for wavelengths lt 0 7 y m Vegetation albedo for wavelengths gt 0 7 u m 0 85 0 6 0 08 0 0 45 0 5 0 5 10 0 02 100 1000 3000 0 10 0 30 0 80 0 1 0 05 0 0 60 0 5 4 0 0 02 100 1000 3000 0 10 0 30 0 80 0 1 1 00 9 0 80 2 0 0 06 100 1500 3000 0 05 0 23 0 80 0 3 1 00 9 0 0 06 100 1500 3000 0 05 0 23 0 80 0 5 0 80 0 0 0 06 100 2000 3000 0 08 0 28 Table 2 BATS vegetation land cover 6 0 90 0 3 2 00 18 0 60 2 0 0 06 100 1500 3000 0 04 0 20 7 0 80 0 0 0 10 0 0 60 0 5 2 0 0 02 100 1000 3000 0 08 0 30 Land Cover Vegetation Type 8 0 00 0 2 0 05 0 0 200 0 0 0 5 0 02 100 1000 3000 0 20 0 40 9 0 60 0 6 0 04 0 0 80 6 0 5 0 5 0 02 100 1000 3000 0 10 0 30 10 0 80 0 1 0 06 0 0 45 6
8. and northward v velocity components are collocated at the corners The center points of grid squares will be referred to as cross points and the corner points are dot points Hence horizontal velocity is defined at dot points Data is input to the model the preprocessors do the necessary interpolation to assure consistency with the grid All the above variables are defined in the middle of each model vertical layer referred to as half levels and represented by the dashed lines in Figure 1 Vertical velocity is carried at the full levels solid lines In defining the sigma levels it is the full levels that are listed including levels at o 0 and 1 The number of model layers is therefore always one less than the number of full sigma levels The finite differencing in the model is of course crucially dependent upon the grid staggering wherever gradients or averaging are represent terms in the equation 1 3 Map Projections and Map Scale Factors The modeling system has a choice of four map projections Lambert Conformal is suit able for mid latitudes Polar Stereographic for high latitudes Normal Mercator for low latitudes and Rotated Mercator for extra choice The x and y directions in the model do not correspond to west east and north south except for the Normal Mercator pro jection and therefore the observed wind generally has to be rotated to the model grid and the model u and v components need to be rotated before comparison with obse
9. and Kennedy P 1993 Biosphere atmosphere trans fer scheme bats version le as coupled to the ncar community climate model Tech nical report National Center for Atmospheric Research Dickinson R Kennedy P Henderson Sellers A and Wilson M 1986 Biosphere atmosphere transfer scheme bats for the ncar community climate model Technical report National Center for Atmospheric Research Emanuel K A 1991 A scheme for representing cumulus convection in large scale models Quart J Roy Meteor Soc 48 2313 2335 Fritsch J M and Chappell C F 1980 Numerical prediction of convectively driven mesoscale pressure systems part i Convective parameterization Journal of Atmo spheric Sciences Giorgi F 1989 Two dimensional simulations of possible mesoscale effects of nuclear war fires Journal of Geophysical Research Giorgi F 1990a Sensitivity of simulated wintertime precipitation and soil hydrology simulation over the western united states to lower boundary specifications Atmos Ocean Giorgi F 1990b Simulation of regional climate using a limited area model nested in a general circulation model Journal of Climate Giorgi F 1991 Sensitivity of simulated summertime precipitation over the western united states to different physics parameterization Monthly Weather Review Giorgi F 1995 Perspectives for regional earth system modeling Global Planet Change 40 Giorgi F and Bates
10. file which contains information regarding domain and grid parameters go into the TERRAIN subdirectory cd RegCM PreProc Terrain edit the domain param file xemacs domain param STEP 3 Run the terrain x script This compiles code and creates an executable file called terrain that is used to generate the DOMAIN file and creates two symbolic links CAT CDF and ELEV CDF to the landuse and elevation datasets respectively copy the appropriate Makefile according to what kind of machine your working on cp Makefile_PGI Makefile execute the terrain script gt terrain This will generate two files in the RegCM Input sub directory DOMAIN INFO and DOMAIN CTL See Table 4 for a list of variables To view the file in GrADS go into the Input subdirectory gt cd RegCM Input open GrADS gt grads opens GrADS grads open DOMAIN CTL opens file in GrADS grads q file list variables in DOMAIN grads d ht displays elevation contours over domain 28 6 2 2 ICBC The second step is to interpolate the sea surface temperature and global analysis data that will be used for the initial and boundary conditions to the model grid This step is performed in the RegCM3 PreProc ICBC sub directory STEP 1 Go into the ICBC sub directory and execute the icbe script It is not necessary to modify any files in this directory Simply run the cbc x script and it will create and run the executables to gene
11. median cloud liquid water content equation according to the following qu CLARA ee 31 where T is temperature in degrees Celsius and Caes is the autoconversion scale factor Precipitation is assumed to fall instantaneously SUBEX also includes simple formulations for raindrop accretion and evaporation The formulation for the accretion of cloud droplets by falling rain droplets is based on the work of Beheng 1994 and is as follows Pgs ce Paim 32 where Pac is the amount of accreted cloud water Cace is the accretion rate coefficient and Psum is the accumulated precipitation from above falling through the cloud Precipitation evaporation is based on the work of Sundqvist et al 1989 and is as follows Leva Ceva 1 RH P um 33 p p where Pevap is the amount of evaporated precipitation and Cevap iS the rate coefficient For a more detailed description of SUBEX and a list of the parameter values refer to Pal et al 2000 17 2 2 6 Ocean flux Parameterization 1 BATS 2 Zeng Sensible heat SH latent heat LH and momentum 7 fluxes between the sea surface and lower atmosphere are calculated using the following bulk aerodynamic algorithms T patz Uy Uy 2 12 u 34 SH paCpatts 35 LH p L U Q 36 where us and u are mean wind components u is the frictional wind velocity 0 is the temperature scaling parameter q is the specific humidity scaling parameter pa is air density Cp
12. of Hsie et al 1984 A first major upgrade of the model physics and numerical schemes was documented by Giorgi et al 1993a Giorgi et al 1993b and resulted in a second generation RegCM hereafter referred to as RegCM2 The physics of RegCM2 was based on that of the NCAR CCM2 Hack et al 1993 and the mesoscale model MM5 Grell et al 1994a In particular the CCM2 radiative transfer package Briegleb 1992 was used for radiation calculations the non local boundary layer scheme of Holtslag et al 1990 replaced the older local scheme the mass flux cumulus cloud scheme of Grell 1993 was added as an option and the latest version of BATS1E Dickinson et al 1993 was included in the model In the last few years some new physics schemes have become available for use in the RegCM mostly based on physics schemes of the latest version of the CCM CCM3 Kiehl et al 1996 First the CCM2 radiative transfer package has been replaced by that of the CCM3 In the CCM2 package the effects of H20 Oz O2 CO2 and clouds were accounted for Solar radiative transfer was treated with a d Eddington approach and cloud radiation depended on three cloud parameters the cloud fractional cover the cloud liquid water content and the cloud effective droplet radius The CCM3 scheme retains the same structure as that of the CCM2 but it includes new features such as the effect of additional greenhouse gases NO2 CH4 CFCs atmospheric aerosols and cloud ice
13. the new RegCM3 version Other new features in the RegCM include improvements in the coupled lake model Small et al 1999 and the incorporation of a tracer model with capability of radiative interactions Qian et al 1999 There have been several improvements and additions to the newest version of the model RegCM3 which will be released at this workshop Changes in the model physics include a new large scale cloud and precipitation scheme which accounts for the subgrid scale variability of clouds Pal et al 2000 new parameterizations for ocean surface fluxes Zeng et al 1998 and a cumulus convection scheme Betts 1986 Also new in the model is a mosaic type parameterization of subgrid scale heterogeneity in topography and land use Giorgi et al 2003b however this scheme is not included in the version released for the workshop but will soon be available Other improvements in RegCM3 involve the input data The USGS Global Land Cover Characterization and Global 30 Arc Second Elevation datasets are now used to create the terrain files In addition NCEP and ECMWF global reanalysis datasets are used for the intial and boundary conditions Lastly improvements in the user friendliness of the model have been made New scripts have been included which make running the programs easier Also a new website has been developed where users can freely download the entire RegCM system as well as all of the input data necessary for a simulation The
14. to be equal to 0 For the calculation of the eddy diffusivity and countergradient terms the PBL height is diagnostically computed from bh Rir u h v hy 15 9 0s 10 h 8s where u h v h and 0 are the wind components and the virtual potential temperature at the PBL height g is gravity Ri r is the critical bulk Richardson number and 0 is an appropriate temperature of are near the surface Refer to Holtslag et al 1990 and Holtslag and Boville 1993 for a more detailed description 2 2 4 Convective Precipitation Schemes Convective precipitation is computed using one of three schemes 1 Grell scheme Grell 1993 2 Modified Kuo scheme Anthes 1977 and 3 Betts Miller Scheme In addi tion the Grell parameterization is implemented using one of two closure assumptions 1 the Arakawa and Schubert closure Grell et al 1994a and 2 the Fritsch and Chappell closure Fritsch and Chappell 1980 hereafter refered to as AS74 and FC80 respectively 12 1 Grell Scheme The Grell scheme Grell 1993 similar to the AS74 parameterization considers clouds as two steady state circulations an updraft and a downdraft No direct mixing occurs between the cloudy air and the environmental air except at the top and bottom of the circulations The mass flux is constant with height and no entrainment or detrainment occurs along the cloud edges The originating levels of the updraft and down draft are given by the levels of m
15. vegetation land cover types and associated parameters are presented in Table 2 Each grid cell of the model is assigned one of the eighteen categories More information regarding GLCC datasets can be found at http edcdaac usgs gov glcc glec html The elevation data used is from the United States Geological Survey USGS Both the landuse and elevation data files are available at 30 and 10 minute resolutions and can be downloaded from the ICTP PWC website at the following URL http www ictp trieste it pubregem RegCM3 DATA SURFACE Parameters such as domain size input data and length of simulation are defined in the RegCM PreProc Terrain domain param file Table 3 After editing this file running the terrain x script will compile and execute the terrain program This will generate the output file DOMAIN INFO containing elevation landuse type and other variables Table 4 in the RegCM Input sub directory A GrADS descriptor file DOMAIN CTL is also created 3 2 ICBC The ICBC program interpolates sea surface temperature SST and global re analysis data to the model grid These files are used for the initial and boundary conditions during the simulation Eventually interfaces will exist to allow GCM output to be used for initial and boundary conditions 3 2 1 Sea surface temperature In the RegCM PreProc Terrain domain param file there are two options for SST data One is the Global Sea Surface Temperature GISST one degree m
16. whole released RegCM modeling system is composed by four components Ter rain ICBC RegCM and Postprocessor Terrain and ICBC are the two components of RegCM preprocessor Terrestrial variables include elevation landuse and sea surface temperature and three dimensional isobaric meteorological data are horizontally inter polated from a latitude longitude mesh to a high resolution domain on either a Rotated and Normal Mercator Lambert Conformal or Polar Stereographic projection Vertical interpolation from pressure levels to the o coordinate system of RegCM is also performed c surfaces near the ground closely follow the terrain and the higher level surfaces tend to approximate isobaric surfaces Since the vertical and horizontal resolution and domain size can vary the modeling package programs employ parameterized dimensions requiring a variable amount of core memory and the requisite hard disk storage amount is varied accordingly ie ho N Nile 4 0 3 45 5 0 4 eal 55 6 0 5 0 6 7 8 0 7 9 0 78 10 0 84 11 0 89 12 0 93 13 0 06 16 1 00 Po amp O Figure 1 Schematic representation of the vertical structure of the model This example is for 14 vertical layers Dashed lines denote half sigma levels solid lines denote full sigma levels Adapted from the PSU NCAR Mesoscale Modeling System Tutorial Class Notes and User s Guide 1 2 The RegCM Model Horizontal and Vertical Grid It is useful to firs
17. G 1989 The climatological skill of a regional model over complex terrain Monthly Weather Review Giorgi F Bates G and Nieman S 1992a Simulation of the arid climate of the southern great basin using a regional climate model Bulletin of the American Me teorological Society Giorgi F Bates G and Nieman S 1993a The multi year surface climatology of a regional atmospheric model over the western united states Journal of Climate Giorgi F Bi X and Qian Y 2002 Radiative forcing and regional climatic effects of anthropogenic aerosols over east asia A regional coupled climate chemistry aerosols model study Journal of Geophysical Research Giorgi F Bi X and Qian Y 2003a Indirect vs direct effects of anthropogenic sulfate on the climate of east asia as simulated with a regional coupled climate chemistry aerosol model Climate Change Giorgi F Francisco R and Pal J 2003b Effects of a subgrid scale topography and land use scheme on the simulation of surface climate and hydrology part 1 Effects of temperature and water vapor disaggregation Journal of Hydrometeorology Giorgi F Hostetler S and Brodeur C S 1994a Analysis of the surface hydrology in a regional climate model Quarterly Journal of the Royal Meteorological Society Giorgi F Huang Y Nishizawa K and Fu C B 1999 A seasonal cycle simula tion over eastern asia and its sensitivity to radiative transfer an
18. K Total precipitation Geopotential height gpm Total soil water mm rb Base flow mm day T L 1 mm Table 9 List of output variables from surface model mm 3 m m Snow water equivalent Sensible heat W m Solar incident W m j m Surface runoff mm day J 35 Table 10 List of output variables from radiation model fc Cloud fraction fraction clwp Cld liquid H20 path g m i K s qi fw clrst Clear sky col abs sol W m gt clrss Clear sky surf abs sol W m qrs Solar heating rate wW Wim cirit chr Solin sabtp firtp Table 11 List of output variables from tracer model i rur 2 2 Wet deposition convective kg m sdrdp_tr Wet deposition large scale kg m emiss_tr Surface dry deposition kg m 36 Table 12 List of variables to be modified in domain param file Fas ooo grid point separation inn IDATEI DATE SSTTYP DATTYP Table 13 List of variables to be modified in regem in file time interval to save output hrs for restart Capia 6 to interval to save atmospheric ouput radisp 6 time interval to save radiation opui Wus 37 Table 14 List of variables to be modified in regem in file iboudy 5 lateral boundary conditions 0 fixed 1 relaxation linear 2 time dependent 3 time and inflow outflow dependent 4 sponge 5 relaxation exponential cumu
19. RegCM Version 3 0 User s Guide Physics of Weather and Climate Group International Centre for Theoretical Physics MIRAMARE TRIESTE February 2004 Written by Nellie Elguindi Xungiang Bi Filippo Giorgi Badrinath Nagarajan Jeremy Pal and Fabien Solmon Abstract As one of the main aims of the ICTP is to foster the growth of advanced studies and research in developing countries the main purpose of this Regional Climate Model RegCM Tutorial Class Notes is to give model users a guide to learn the whole RegCM Model System The RegCM Tutorial Class is offered as a part of extended hands on lab sessions during a series of Workshops organized by the Physics of Weather and Climate PWC group at the Abdus Salam International Centre for Theoretical Physics ICTP RegCM was originally developed at the National Center for Atmospheric Re search NCAR and has been mostly applied to studies of regional climate and seasonal predictability around the world The workshop participants are welcome to use RegCM for regional climate simulation over different areas of interest The RegCM is available on the World Wide Web at http www ictp trieste it pubregcem RegCM3 Contents 1 Introduction 4 Ele Str Les rs a Ec A e Me dd SS ded amp 4 1 2 The RegCM Model Horizontal and Vertical Grid 6 1 3 Map Projections and Map Scale Factors oaoa a 8 2 Model Description 9 Bl Dynamics ps LAI A o e a 9 2D ES Dl A REA RAI AAA a wk oe
20. a is specific heat of air and Le is the latent heat of vaporization For further details on the calculation of these parameters refer to Zeng et al 1998 2 2 7 Pressure Gradient Scheme Two options are available for calculating the pressure gradient force The normal way uses the full fields The other way is the hydrostatic deduction scheme which makes use of a perturbation temperature In this scheme extra smoothing on the top is done in order to reduce errors related to the PGF calculation 2 2 8 Lake Model The lake model developed by Hostetler et al 1993 can be interactively coupled to the atmospheric model In the lake model fluxes of heat moisture and momentum are calculated based on meteorological inputs and the lake surface temperature and albedo Heat is transferred vertically between lake model layers by eddy and convective mixing Ice and snow may cover part or all of the lake surface In the lake model the prognostic equation for temperature is oT OT a Bet Bm ar a where T is the temperature of the lake layer and ke and km are the eddy and molecular diffusivities respectively The parameterization of Henderson Sellers 1986 is used to calculate ke and km is set to a constant value of 39 x 1077 m s7 except under ice and at the deepest points in the lake Sensible and latent heat fluxes from the lake are calculated using the BATS param eterizations Dickinson et al 1993 The bulk aerodynamic formulations for l
21. al scales The effects of shal low convection is viewed as a mixing process between the surface layer air and the free atmosphere Observational basis for deep convection The thermodynamics of the BM scheme is based on the saturation point formulation as reported in Betts 1982 The saturation point sp is defined as the temperature and pressure T p at the lifting condensation level LCL The subsaturation parameter P is the difference between air parcel saturation level pressure and the actual pressure level i e P p p Betts 1986 observed that temperature profiles below the freezing level in deep convection is parallel to the gs5y isopleth where gsy is defined as a constant virtual equivalent potential temperature This led to the proposal that the reference lapse rate in the lower troposphere is moist virtual adiabatic rather than the widely accepted moist adiabat Since the slope of the Oxsv isopleth is 0 9 times that of the moist adiabat the air parcel buoyancy reduction due to cloud water content is accounted for This reference structure in the presence of deep convection is universal as reported by Betts 1986 for the cases of hurricanes 14 GATE slow and fast moving squall lines and Venezuela convective episodes Thus the reference structure below 600 hPa or the freezing level is the 0Ogsy with 0gs constrained to a minimum at 600 hPa Above 600 hPa the observed thermal profile increases to cloud top gs value Considerab
22. ata set and P Pr Ors p 0gs M Ozs T On5 M 2 aa forpr lt p lt pm 23 where 0gs T is the environmental saturated equivalent potential temperature at cloud top and gs M is the minimum saturated equivalent potential temperature at the freez ing level The moisture profile is found by specifying the P p p parameter at three levels i e at the cloud base Pg 3875 Pa at the freezing level Pu 5875 Pa and at the cloud top Pr 1875 Pa with linear gradients in between While the temperature profiles show greater universality the observed moisture profiles exhibit considerable variability The above pre specified profile defined by Pg Pr and Pr corresponds to a mean profile observed in the tropics In the present simulation the average moisture structure is representative of the GATE region squall lines The first guess temperature and moisture profiles are corrected to satisfy the total enthalpy constraint tE dp 0 24 where k CpI p Lar and Tr qn are the first guess reference temperature and specific humidity respectively k C T Lg T q are the grid mean temperature and specific humidity before the onset of deep convection Reference profiles for shallow convection The first guess profiles for temperature and specific humidity are constructed from properties of the air at the cloud base pressure 15 pp and air above the cloud top pressure pp Equal quantities of air f
23. atent heat flux F and sensible heat flux F are as follows Fa PaCD Valds da 38 F paCpCpVa T Ta 39 18 where the subscripts s and a refer to surface and air respectively pa is the density of air Va is the wind speed Cr q is specific humidity and T is temperature The momentum drag coefficient Cp depends on roughness length and the surface bulk Richardson number Under ice free conditions the lake surface albedo is calculated as a function of solar zenith angle Henderson Sellers 1986 Longwave radiation emitted from the lake is calculated according to the Stefan Boltzmann law The lake model uses the partial ice cover scheme of Patterson and Hamblin 1988 to represent the different heat and moisture exchanges between open water and ice surfaces and the atmosphere and to calculate the surface energy of lake ice and overlying snow For further details refer to Hostetler et al 1993 and Small and Sloan 1999 2 2 9 Tracer Model Tracer is a generic term used to design quantities other than classical meteorological variables transported by and interacting with the physics of RegCM Due to their climatic impact aerosols and gases are the first tracers of interest that we consider here For these quantities the model solves a prognostic equation of the form Qian et al 2001 A V Vx Fy Fy Toum Sy Ruts Rw cum Daep Y Qp Q 40 where the first term on the right hand side repr
24. ather Review Marinucci M and Giorgi F 1992 A 2xco2 climate change scenario over europe gener ated using a limited area model nested in a general circulation model i Present day climate simulation Journal of Geophysical Research Mearns L Bogardi I Giorgi F Matyasovszky I and Palecki M 1999a Comparison of climate change scenarios generated from regional climate model experiments and statistical downscaling Journal of Geophysical Research Mearns L Giorgi F McDaniel L and Brodeur C S 1995 Analysis of daily variabil ity and diurnal range of temperature in a nested regional climate models Comparison with observations and doubled co2 results Climate Dynamics 43 Mearns L Mavromatis T and Tsvetsinskaya E 1999b Comparative responses of epic and ceres crop models to high and low spatial resolution climate change scenarios Journal of Geophysical Research Pal J Small E and Eltahir E 2000 Simulation of regional scale water and energy budgets Representation of subgrid cloud and precipitation processes within regcm Journal of Geophysical Research Atmospheres Pan Z T Takle E Segal M and Arritt R 1999 Simulation of potential impacts of man made land use changes on u s summer climate under various synoptic regimes Journal of Geophysical Research Patterson J and Hamblin P 1988 Thermal simulation of a lake with winter ice cover Limn Oceanography Perk
25. awa B grid staggering of the dot and cross grid points le poke E amp ei ea Bees T List of Tables OoOnN nor wonder Ra Ne A 0 Land Cover Vegetation classes 2 2 2 eee 21 BATS vegetation land cover ia 4 See e aS amp ae A 22 List of variables defined in domain param file 31 List of output variables from Terrain DOMAIN 32 List of variables in ICBCYYYYMMDDHH files 32 List of restart timestep and output parameters defined in regem in file 33 List of physic options in regem in file sia rata Ae as ta AS 34 List of output variables from atmosphere aaa o 35 List of output variables from surface model a o a aa o 35 List of output variables from radiation model 36 List of output variables from tracer model oaoa aaa o 36 List of variables to be modified in domain param file 37 List of variables to be modified in regem in file 37 List of variables to be modified in regem in file 38 1 Introduction 1 1 History The idea that limited area models LAMs could be used for regional studies was originally proposed by Dickinson et al 1989 and Giorgi 1990b This idea was based on the con cept of one way nesting in which large scale meteorological fields from general circulation model GCM runs provide initial and time dependent meteorological lateral boundary condit
26. aximum and minimum moist static energy respectively The Grell scheme is activated when a lifted parcel attains moist convection Condensation in the updraft is calculated by lifting a saturated parcel The downdraft mass flux mo depends on the updraft mass flux m according to the following relation Mo Mob 16 where J is the normalized updraft condensation Jy is the normalized downdraft evapo ration and 3 is the fraction of updraft condensation that re evaporates in the downdraft 8 depends on the wind shear and typically varies between 0 3 and 0 5 Rainfall is given by P hm 8 17 Heating and moistening in the Grell scheme are determined both by the mass fluxes and the detrainment at the cloud top and bottom In addition the cooling effect of moist downdrafts is included Due to the simplistic nature of the Grell scheme several closure assumptions can be adopted RegCM3 s default version directly implements the quasi equilibrium assumption of AS74 It assumes that convective clouds stabilize the environment as fast as non convective processes destabilize it as follows _ ABE ABE where ABE is the buoyant energy available for convection ABE is the amount of buoy ant energy available for convection in addition to the buoyant energy generated by some of the non convective processes during the time interval At and NA is the rate of change of ABE per unit my The difference ABE ABE can be thought of as t
27. d surface processes Journal of Geophysical Research Giorgi F and Marinucci M 1991 Validation of a regional atmospheric model over europe Sensitivity of wintertime and summertime simulations to selected physics parameterizations and lower boundary conditions Quarterly Journal of the Royal Meteorological Society Giorgi F and Marinucci M 1996 An investigation of the sensitivity of simulated precipitation to the model resolution and its implications for climate studies Monthly Weather Review Giorgi F Marinucci M and Bates G 1993b Development of a second generation regional climate model regem2 i Boundary layer and radiative transfer processes Monthly Weather Review Giorgi F Marinucci M Bates G and DeCanio G 1993c Development of a second generation regional climate model regcm2 ii Convective processes and assimilation of lateral boundary conditions Monthly Weather Review Giorgi F Marinucci M and Visconti G 1990 Use of a limited area model nested in a general circulation model for regional climate simulation over europe Journal of Geophysical Research 41 Giorgi F Marinucci R and Visconti G 1992b 2xco2 climate change scenario over europe generated using a limited area model nested in a general circulation model 2 Climate Change Scenario Giorgi F and Mearns L 1991 Approaches to regional climate change simulation A review Review of Geophysics Giorg
28. e B Briegleb B Kiehl J Rasch P and Williamson D 1993 De scription of the ncar community climate model ccm2 Technical report National Center for Atmospheric Research Henderson Sellers B 1986 Calculating the surface energy balance for lake and reservoir modeling A review Rev Geophys High resolution models of the planetary boundary layer in Advances in Environmental Science and Engineering 1979 Gordon and Breach Newark N J Hirakucci H and Giorgi F 1995 Multiyear present day and 2xco2 simulations of monsoon dominated climate over eastern asia and japan with a regional climate model nested in a general circulation model Journal of Geophysical Research Holtslag A and Boville B 1989 Local versus nonlocal boundary layer diffusion in a global climate model Journal of Climate 42 Holtslag A and Boville B 1993 Local versus nonlocal boundary layer diffusion in a global climate model Journal of Climate Holtslag A de Bruijn E and Pan H L 1990 A high resolution air mass transforma tion model for short range weather forecasting Monthly Weather Review Hostetler S Bates G and Giorgi F 1993 Interactive nesting of a lake thermal model within a regional climate model for climate change studies Geophysical Research Hostetler S Giorgi F and Bates G 1994a Role of lake atmosphere feedbacks in sustaining lakes bonneville and lahontan 18 000 years ago Science
29. e gravitational 16 acceleration ppandpag are the respective pressure at the lowest model level and at the top of the ABL and Aq is the change of q along the downdraft descent path Simply put the temperature and moisture profiles for the downdraft are parallel to the moist adiabat at constant subsaturation The downdraft air is injected into the 3 lowest model levels in the ABL 2 2 5 Large Scale Precipitation Scheme Subgrid Explicit Moisture Scheme SUBEX is used to handle nonconvective clouds and precipitation resolved by the model This is one of the new components of the model SUBEX accounts for the subgrid variability in clouds by linking the average grid cell relative humidity to the cloud fraction and cloud water following the work of Sundqvist et al 1989 The fraction of the grid cell covered by clouds FC is determined by RH RA nin See E ee I 2 si tH max 1 tH min l where RH in is the relative humidity threshold at which clouds begin to form and RH mas is the relative humidity where FC reaches unity FC is assumed to be zero when RH is less than RHmin and unity when RH is greater than RA maz Precipitation P forms when the cloud water content exceeds the autoconversion thresh old Q according to the following relation P Oppe Qe FC Qe FC 30 where 1 Cppi can be considered the characteristic time for which cloud droplets are con verted to raindrops The threshold is obtained by scaling the
30. egional climate model to eastern africa 1 simulation of the short rains of 1988 Journal of Geophysical Research Sun L Q Semazzi F Giorgi F and Ogallo L 1999b Application of the ncar regional climate model to eastern africa 2 simulation of interannual variability of short rains Journal of Geophysical Research 44 Sundqvist H Berge E and Kristjansson J 1989 The effects of domain choice on summer precipitation simulation and sensitivity in a regional climate model Journal of Climate Trenberth K 1992 Global analyses from ecmwf and atlas of 1000 to 10mb circulation statistics Technical report National Center for Atmospheric Research Zeng X Zhao M and Dickinson R E 1998 Intercomparison of bulk aerodynamic algoriths for the computation of sea surface fluxes using toga coare and tao data Journal of Climate Zhang D L Hsie E Y and Moncrieff M 1988 A comparison of explicit and implicit predictions of convective and stratiform precipitating weather systems with a meso P scale numerical model 114 31 60 Zhang G J and McFarlane N 1995 Sensitivity of climate simulations to parame terization of cumulus convection in the canadian climate centre general circulation model Atmos Ocean 45
31. erwise 21 Note that the moisture convergence term includes only the advective tendencies for water vapor However evapotranspiration from the previous time step is indirectly included in M since it tends to moisten the lower atmosphere Hence as the evapotranspiration increases more and more of it is converted into rainfall assuming the column is unstable The latent heating resulting from condensation is distributed between the cloud top and bottom by a function that allocates the maximum heating to the upper portion of the cloud layer To eliminate numerical point storms a horizontal diffusion term and a time release constant are included so that the redistributions of moisture and the latent heat release are not performed instantaneously Giorgi and Bates 1989 Giorgi and Marinucci 1991 3 Betts Miller Scheme In the Betts Miller scheme the sub grid scale effects of convective clouds are represented by adjusting temperature and moisture profiles to the observed quasi equilibrium structures for deep convection and to a mixing line structure for shallow convection Quasi equilibrium between the cloud field and the large scale forc ing forms the basis of representing deep convection in the BM scheme Quasi equilibrium means that the convective cloud field constrains the thermal and moisture structure of the atmosphere against the destabilizing influence of the large scale flow The concept has been found to be valid on large spatial and tempor
32. esents the horizontal and vertical advection which are solved using a relaxed upstream scheme Fy and Fy are horizontal and vertical turbulent diffusion treated in a same way as for water vapor Toy is the subgrid transport linked to convection using well mixing approximation Qian et al 1999 or mass flux transport S is the source term prescribed from external data R ls and Ru cum are wet removal by large scale and cumulus clouds Giorgi 1989 Da is the dry deposition on surface depending on the tracer nature its concentration at lower levels and surface type Qp and Q represent the production and loss terms respectively linked to physico chemical transformations depending on the nature of tracer for example the chemistry of sulfate aerosol Qian et al 2001 The prognostic concentrations together with specific optical properties of aerosols are then used to calculate the atmospheric column optical properties required by the radiation scheme and ensuing climatic impact Giorgi et al 2002 A parameterization of the aerosol indirect effect modification of cloud properties has also been introduced e g Qian et al 1999 19 3 Pre Processing Before performing a regional climate simulation there are two pre processing steps that need to be completed The first step involves defining the domain and grid interval and interpolating the landuse and elevation data to the model grid This task is performed in the RegCM PreProc Te
33. ey D and Kreitzberg C 1976 A time dependent lateral boundary scheme for limited area primitive equation models Monthly Weather Review Qian J H Giorgi F and Fox Rabinovitz M 1999 Regional stretched grid generation and its application to the ncar regcm Journal of Geophysical Research Qian Y Giorgi F and Huang Y 2001 Regional simulation of anthropogenic sulfur over east asia and its sensitivity to model parameters Tellus Seth A and Giorgi F 1998 The effect of domain choice on summer precipitation simulation and sensitivity in a regional climate model Journal of Climate Shea D Trenberth K and Reynolds R 1992 A global monthly sea surface temper ature climatology Journal of Climate Slingo J 1980 A cloud parameterization scheme derived from gate data for use with a numerical model Quarterly Journal of the Royal Meteorological Society Slingo J 1989 A gcm parameterization for the shortwave radiative properties of water clouds Journal of Atmospheric Sciences Small E Giorgi F and Sloan L 1999 Regional climate model simulation of pre cipitation in central asia Mean and interannual variability Journal of Geophysical Research Small E and Sloan L 1999 Simulating the water balance of the aral sea with a coupled regional climate lake model Journal of Geophysical Research Sun L Q Semazzi F Giorgi F and Ogallo L 1999a Application of the ncar r
34. face model variables SRFYYMMCONT NC continual averages of surface model variables e Radiation Model Output Table 10 for list of variables RADYYMM NC all radiation variables and times RADYYMMAVG NC all variables radiation averaged RADYYMMDIUR NC diurnal cycle averages of radiation variables RADYYMMCONT NC continual averages of radiation variables 26 6 Practice Run The purpose of this section is to help new users become familiar with setting up and running RegCM by going through a practice run A step by step tutorial is presented for performing one month simulation over a south Asian domain for July 1991 To demon strate how to use restart option first a 5 day simulation is run in June then the model is restarted and run for an additional 31 days in July In this practice run the 10 minute resolution GLCC and GTOPO datasets are used to create the terrain file and ECMWE global reanalysis datasets are used for the initial and boundary conditions You will create links from your working directory to the directories where these data are using the RegCM PreProc DATA datalinker x script 6 1 Getting the model code and data STEP 1 Create a working directory for yourself mkdir yourname cd yourname STEP 2 Download regcm tar gz to your account from the RegCM3 website at http www ictp trieste it pubregcm RegCM3 STEP 3 Uncompress and untar regcm tar gz gzip d regem tar gz gt tar xvf regcm tar Un
35. feedback in climate models 92 8483 8485 Betts A K and Miller M 1993 The Betts Miller scheme Chapter 9 The rep resentation of cumulus convection in numerical models of the atmosphere Eds K A Emanuel and D J Raymond Meteo Mon American Meteorological Society pp 107 121 Blackadar A 1979 Boundary layer wind maxima and their significance for the growth of nocturnal inversions Bulletin of the American Meteorological Society 39 Briegleb B 1992 Delta eddington approximation for solar radiation in the ncar com munity climate model Journal of Geophysical Research Climate Processes and Climate Sensitivity 1984 American Geophysical Union chapter Modeling evapotranspiration processes for three dimensional global climate models Dai A Giorgi F and Trenberth K E 1999 Observed and model simulated diur nal cycles of precipitation over the contiguous united states Journal of Geophysical Research Davies H and Turner R 1977 Updating prediction models by dynamical relaxation An examination of the technique Quarterly Journal of the Royal Meteorological So ciety Deardoff J 1978 Efficient prediction of ground surface temperature and moisture with inclusion of a layer of vegetation Journal of Geophysical Research Dickinson R Errico R Giorgi F and Bates G 1989 A regional climate model for the western united states Climate Change Dickinson R Henderson Sellers A
36. he interaction of deep convection with the ABL processes cannot be captured by the original deep scheme Furthermore convective downdraft cooling and drying at low levels can have an impact on convective development downstream through advective effects of the cold and drier air over a period of several hours even under weak surface wind regimes such as in the WPR where maximum ABL winds are lt 7 ms In fact the motivation to include downdrafts in the original scheme was to mitigate the problem of grid scale instabilities Zhang et al 1988 arising downstream of the island of New Guinea much later in the simulation The instability arose due to the absence of low level cold and dry air advection emanating from deep convection present over the island between 0000 0600 UTC 15 December 1992 The original scheme is modified with the inclusion of downdrafts suggested by Betts and Miller 1993 The downdrafts are parameterized by defining a simple unsaturated downdraft thermodynamic path constant 0 and constant subsaturation originating at 850 hPa level A different adjustment time for this process is employed and it is a function of evaporation in downdrafts and the deep convective PR precipitation The adjustment time TAg is given by 1 aPR Ee ee EE 28 TaBe dpe Ago E where the constant a is set to 0 10 in this study and is a measure of the precipitation efficiency of the cumulus clouds dp is the vertical pressure interval g is th
37. he parent domain Not recommended for coarse mesh where only one outer row and column would be specified e Linear relaxation Outer row and column is specified by time dependent value next four points are relaxed towards the boundary values with a relaxation constant that decreases linearly away from the boundary e Sponge Perkey and Kreitzberg 1976 e Exponential relaxation Davies and Turner 1977 default Note The type of boundary conditions used in the simulation is selected in the RegCM PreProc Terrain domain param file It is not necessary to modify any files in the RegCM PreProc ICBC sub directory The SST_1DEG f and ICBC f programs interpolate the SST and global analysis data to the model grid Running the cbc x script will compile and execute these programs The following files will be generated RegCM Input ICBC YYYYMMDDHH see Table 11 for list of variables RegCM Input ICBC YY YYMMDDHH CTL 24 4 RegCM All of the source code for the model is in the RegCM Main sub directory The RegCM Commons sub directory contains two files necessary for running a new simu lation Physics options discussed in Section 2 2 are selected in the regcm in file Table 7 Restart timestep and output frequency parameters are also defined in regcm in Ta ble 13 The regem z script will compile and execute the model It is recommended to create a new directory for specific projects and to copy these files into this new project directory
38. he rate of destabilization over time At ABE is computed from the current fields plus the fu ture tendencies resulting from the advection of heat and moisture and the dry adiabatic adjustment Another stability based closure assumption that is commonly implemented in GCMs and RCMs is the FC80 type closure assumption In this closure it is assumed that convection removes the ABE over a given time scale as follows _ ABE NAT Mp 19 where 7 is the ABE removal time scale The fundamental difference between the two assumptions is that the AS74 closure assumption relates the convective fluxes and rainfall to the tendencies in the state of the atmosphere while the FC80 closure assumption relates the convective fluxes to the degree of instability in the atmosphere Both schemes achieve a statistical equilibrium 13 between convection and the large scale processes However this subtle distinction in the implementation of the closure will prove to be an important difference 2 Kuo Scheme Convective activity in the Kuo scheme is initiated when the moisture convergence M in a column exceeds a given threshold and the vertical sounding is con vectively unstable A fraction of the moisture convergence 3 moistens the column and the rest is converted into rainfall POV according to the following relation P M 1 8 20 B is a function of the average relative humidity RH of the sounding as follows 2 0 RH RH gt 05 ps 1 0 oth
39. i F and Mearns L O 1999 Introduction to special section Regional climate modeling revisited Journal of Geophysical Research Giorgi F Mearns L Shields C and L Mayer 1996 A regional model study of the importance of local versus remote controls of the 1988 drought and the 1993 flood over the central united states Journal of Climate Giorgi F Mearns L Shields C and McDaniel L 1998 Regional nested model simulations of present day and 2xc02 climate over the central great plains of the united states Climate Change Giorgi F Shields C Brodeur and Bates G 1994b Regional climate change scenarios over the united states produced with a nested regional climate model Spatial and seasonal characteristics Journal of Climate Grell G 1993 Prognostic evaluation of assumptions used by cumulus parameteriza tions Monthly Weather Review Grell G A Dudhia J and Stauffer D R 1994a A description of the fifth generation penn state ncar mesoscale model mm5 Technical report National Center for At mospheric Research Grell G Dudhia J and Stauffer D 1994b A description of the fifth generation Penn State NCAR mesoscale model MM5 Tech Note TN 398 IA Technical report National Center for Atmospheric Research Grell G Kuo Y H and Pasch R J 1991 Semi prognostic tests of cumulus parame terization schemes in the middle latitudes Monthly Weather Review Hack J Bovill
40. ions LBC for high resolution regional climate model RCM simulations with no feedback from the RCM to the driving GCM The first generation NCAR RegCM was built upon the National Center for Atmo spheric Research Pennsylvania State University NCAR PSU Mesoscale Model version MM4 in the late 1980s Dickinson et al 1989 Giorgi 1989 The dynamical component of the model originated from that of the MM4 which is a compressible finite difference model with hydrostatic balance and vertical v coordinates Later the use of a split explicit time integration scheme was added along with an algorithm for reducing horizontal dif fusion in the presence of steep topographical gradients Giorgi et al 1993a Giorgi et al 1993b As a result the dynamical core of the RegCM is similar to that of the hydrostatic version of MM5 Grell et al 1994a For application of the MM4 to climate studies a number of physics parameterizations were replaced mostly in the areas of radiative transfer and land surface physics which led to the first generation RegCM Dickinson et al 1989 Giorgi 1990b The first gen eration RegCM included the Biosphere Atmosphere Transfer Scheme BATS Dickinson et al 1986 for surface process representation the radiative transfer scheme of the NCAR Community Climate Model CCM version CCM1 a medium resolution local planetary boundary layer scheme the Kuo type cumulus convection scheme of Anthes 1977 and the explicit moisture scheme
41. le variability is associated with the observed moisture structure Despite this a reference moisture profile is specified in the scheme Observational basis for shallow convection A shallow cumulus cloud field is regarded as a mixing process between the surface layer air and the free atmosphere This mixing is characterized by a mixing line Thus when two air parcels mix in the vertical the sp of every possible mixture lies on the mixing line joining the sps of the two parcels This mixing line structure for shallow cumulus convection was highlighted by Betts 1982 by plotting sps between 900 700 hPa and the sps were found to lie close to the line joining the sps This evidence was presented for the undisturbed trade wind region and tropical land stations Reference profiles for deep convection The scheme involves a lagged adjustment of the resolvable scale T and q profile towards a reference quasi equilibrium structure in the presence of grid scale radiative and advective processes The reference T q profiles for deep convection are based on the observation that 0gs at cloud base decreases with height upto the freezing level while being parallel to the 0gsy isopleth and increasing to Omg of the environment near the cloud top The first guess temperature profiles are constructed using On5 p 0gs B aV p pg forpg gt p gt pm 22 where V is the vertical lapse rate of Ogsy a is the weighting factor set to 1 5 based on the GATE d
42. lus parameterization scheme 1 Anthes Kuo 2 Grell 3 Betts Miller igcc 2 Grell Scheme Convective Closure Scheme ats ocean flux parameterization scheme 1 BATS 2 Zeng ipf o pressure gradient scheme 0 normal way 1 hydrostatic deduction time freq for solar rad calculation min time freq for surface model min EE o ao E T ibayt boundary condition iera Gow maa 5 of bottom model eves without dous 3 gulland fraction of Gultepe eqn qcth when precip occurs land fraction of Gultepe eqn qcth when precip occurs ocean rh0oce 0 90 RH threshold for ocean rhOland RH threshold for land raindrop evap rate coef kg m s cacr 60 raindrop accretion rate m kg s7 edtmax maximum precip efficiency 0 18 0 3 minim precip ffiieney 0 Fedtmine 0 0 minimum precip efficiency Ga SSCS Fritsch amp Chappell 1980 ABE Removal Timescale min qckloce autoconversion rate for ocean 4 4 0 0 0 4 38 References Anthes R 1977 A cumulus parameterization scheme utilizing a one dimensional cloud model Monthly Weather Review Anthes R A Kuo Y H Low Nam S Hsie E Y and Bettge T M 1989 Estimation of episodic and climatological skill and uncertainty in regional numerical models Quarterly Journal of the Royal Meteorological Society Anthes R Hsie E and Kuo Y H 1987 Description of the penn state ncar mesoscale model version 4 mm4
43. n cp Main Makefile_PGI Main Makefile STEP 3 Run the regem z script This will compile the source code and start the simulation gt regem z After the simulation is completed you will have the following monthly files of model out put in the RegCM PracticeRun output sub directory ATM 1991062500 output from the atmospheric model RAD 1991062500 output from the radiation model SFC 1991062500 output from the land surface model SAV 1991062500 restart file 6 3 1 Restarting the model To restart the model you only need to modify a few parameters in the regcm in file STEP 1 Edit the the following restart parameters in the regcm in file e ifrest true indicates this is a restart simulations e idate0 1994070100 start date of first simulation e idatel 1994070600 start date for restart simulation e idate2 1994080100 end date for restart simulation STEP 3 Run the regcm z script to restart the simulation gt regem z After the simulation is complete you will have the following monthly files of model output in the RegCM PracticeRun output sub directory ATM 1991070100 output from the atmospheric model RAD 1991070100 output from the radiation model SFC 1991070100 output from the land surface model SAV 1991070100 restart file 30 Table 3 List of variables defined in domain param file as ard point separation dla clon central longitude of model domain in degrees ntypec resol
44. ncar tech note ncar tn 282 Technical report National Center for Atmospheric Research Anthes R Seaman N and Warner T 1980 Comparisons of numerical simulations of the planetary boundary layer by a mixed layer and multi level model Monthly Weather Review Anthes R T W 1978 Development of hydrodynamic models suitable for air pollution and other mesometeorological studies Monthly Weather Review Bates G Giorgi F and Hostetler S 1993 Towards the simulation of the effects of the great lakes on regional climate Monthly Weather Review Bates G Hostetler S and Giorgi F 1995 Two year simulation of the great lakes region with a coupled modeling system Monthly Weather Review Bates G T 1990 A case study of the effects of topography on cyclone development in the western united states Monthly Weather Review Beheng K 1994 A parameterization of warm cloud microphysical conversion processes Atmospheric Research Betts A 1982 Cloud thermodynamic models in saturation point coordinates 39 2182 2191 Betts A 1986 A new convective adjustment scheme Part I Observational and theo retical basis 112 667 691 Betts A 1997 The parameterization of deep convection in R Smith ed The physics and parameterization of moist atmospheric convection Kluwer Academic Publishers pp 255 279 Betts A and Harshavardhan 1987 Thermodynamic constraint on the cloud liquid water
45. on exponential cumulus scheme 1 Anthes Kuo 2 Grell 3 Betts Miller Grell Scheme Convective Closure Scheme 1 Arakawa amp Schubert 2 Fritsch amp Chappell pressure gradient scheme 0 normal way 1 hydrostatic deduction Tracer Chemistry model 0 no 1 yes Description number of bottom model levels without clouds maximum cloud cover fraction autoconversion rate for land ocean fraction of Gultepe eqn qcth when precip occurs over land ocean RH at which FCC 1 0 RH threshold for ocean land raindrop evap rate coef kg m s raindrop accretion rate m kg s7 Grell parameter shrmin shrmax edtmin edtmax edtmino edtmaxo edtminx edtmaxx pbcmax mincld htmin htmax skbmax dtauc Chemistry parameters ichremlsc ichremcvc ichdrdepo ichcumtra idirect minimum maximum shear effect on precip eff minimum maximum precip efficiency minimum maximum precip efficiency o var minimum maximum precip efficiency x var max depth mb of stable layer between LCL amp LFC minimum cloud depth mb minimum maximum convective heating maximum cloud base height in sigma coords Fritsch amp Chappell 1980 ABE Removal Timescale min Description aerosol wet removal from large scale clouds aerosol wet removal from convective clouds aerosol dry deposition on surface convective transport of aerosols direct radiative effect of aerosols 34 Table 8 List of output variables from atmosphere
46. onthly gridded data 1871 2002 available from the Hadley Centre Met Office http badc nerc ac uk data gisst note special permission is needed from the Hadley Center Met Office to use the GISST datasets Also available is the Optimum Interpolation Sea Surface Temperature OISST one degree weekly analysis 1981 2002 available from the National Ocean and Atmosphere Administration http www cdc noaa gov 3 2 2 Initial and Boundary Conditions In the RegCM PreProc Terrain domain param file there are three options to choose from for the global analysis datasets to use for the initial and boundary conditions e ECMWE The European Centre for Medium Range Weather Forecasts Reanalysis datasets T42 L15 from 1993 1997 23 e NNRP1 The National Center for Environmental Prediction NCEP Reanalysis datasets 2 5 degree grid L17 from 1948 2001 e NNRP2 The National Center for Environmental Prediction NCEP Reanalysis datasets 2 5 degree grid L17 from 1979 2001 The numerical treatment of the lateral boundaries is a difficult but very important aspect of the regional climate model There are five types of boundary conditions that can be used in the model e Fixed This will not allow time variation at lateral boundaries Not recommended for real data applications e Time dependent Outer two rows and columns have specified values of all pre dicted fields Recommended for nests where time dependent values are supplied by t
47. rate the files for initial and boundary conditions copy the appropriate Makefile according to what kind of machine your working on gt cp Makefile_PGI Makefile gt cd RegCM PreProc ICBC icbc x This will generate two files in the RegCM Input sub directory CBC1991062500 and ICBC1994062500 CTL These files are used to for the initial and boundary conditions during the simulation 6 3 Running the Model It is convenient to create a new directory for your simulation where the executable file and model output files will be written STEP 1 Create a sub directory called RegCM PracticeRun and copy the regcm in and regcm z in the RegCM Commons subdirectory to it make a second level subdirectory called PracticeRun mkdir PracticeRun go into the new subdirectory PracticeRun cd PracticeRun copy the two files regem in and regcm x from the Commons subdirectory gt cp Commons regcm in gt cp Commons regcm z STEP 2 Before running the simulation you only need to modify the the regcm in file This file contains parameters regarding the use of restart files and physics options Edit the file according to the parameters defined in Table 13 and Table 14 First a 5 day simulation from 25 June 1991 00 UTC through 30 June 1991 00 UTC will be performed edit the regcm in file xemacs regcm in copy the appropriate Makefile in the Main subdirectory according to what kind of ma 29 chine you are working o
48. rom pg and pz are mixed and the corresponding sp i e level 1 determined The slope of the mixing line is found as _ 051 05 B Ps 1 PsL B where B corresponds to the cloud base and Pgz is the pressure at saturation levels i e for parcels from cloud base B and the air mixture between pg and pz The temperature profile reference is parallel to the mixing line and is given by 25 Ons p Pns B M p pp 26 Ors p is inverted to yield T and p which with the subsaturation parameter at level 1 gives sp and qv The first guess reference T and q profiles are corrected to satisfy the following energy constraints Op Ea T av f Llar ap 0 27 B Adjustment time 7 The adjustment time 7 in the scheme is set such that the atmosphere nearly saturates on the grid scale in the presence of a convective disturbance According to Betts and Miller 1993 in the T106 resolution 1 125 model 7 for deep convection and for shallow convection is 2 h Betts 1997 expressed 7 as a function of horizontal scales He found 7 to lie between 40 80 min for the T106 resolution model For Ax 60 km 7 lies between 20 40 min In the present simulation we employ 7 55 min for deep and shallow convection at Ax 60 km resolution Downdrafts in the BM scheme The BM scheme Betts 1986 was designed primarily for tropical convection One of the weakness in the original deep convective scheme is the absence of downdrafts T
49. rrain sub directory The second step is to generate the files used for the initial and boundary conditions during the simulation This step is performed in the RegCM PreProc ICBC sub directory All of input data necessary to run the model can be downloaded from the PWC website at the following URL http www ictp trieste it pubregem RegCM3 Input data used by the Terrain and ICBC programs are stored in the RegCM PreProc DATA sub directory A script called datalinker xis provided in this directory in case the data exists elsewhere It can be modified and run to create soft links between the RegCM PreProc DATA sub directory and some other directory The present version of RegCM3 support multi planform of UNIX or LINUX op eration system such as IBM SGI SUN DEC and PC LINUX with PGI FORTRAN compiler or Intel IFC FORTRAN compiler You must make your choices of Makefile un der PreProc Terrain PreProc ICBC and Main directories by copying the appropriate Makefile 20 3 1 Terrain Table 1 Land Cover Vegetation classes 1 11 12 13 14 15 16 17 18 19 20 OO SE DS Crop mixed farming Short grass Evergreen needleleaf tree Deciduous needleleaf tree Deciduous broadleaf tree Evergreen broadleaf tree Tall grass Desert Tundra Irrigated Crop Semi desert Ice cap glacier Bog or marsh Inland water Ocean Evergreen shrub Deciduous shrub Mixed Woodland Forest Field mosaic Water and Land mixture
50. rva tions These transformations are accounted for in the model pre processors that provide data on the model grid and in the post processors The map scale factor m is defined by m distance on grid actual distance on earth and its value is usually close to one varying with latitude The projections in the model preserve the shape of small areas so that dx dy everywhere but the grid length varies across the domain to allow a representation of a spherical surface on a plane surface Map scale factors need to be accounted for in the model equations wherever horizontal gradients are used 2 Model Description 2 1 Dynamics The model dynamic equations and numerical discretization are described by Grell et al 1994a Horizontal Momentum Equations Optu an Op uu m Op vu m Op ua Ot Ox Oy Oo RT 2 mp CESAS p p o Ox Ox t fp v Pou Fyu 2 p v 22 Op uv m a Op vu m 3 Op va Ot Ox Oy Oo RT Op z i ae aaa 4 HEF Fyv 3 K ena Og Oye ee 05 where u and v are the eastward and northward components of velocity T is virtual tem perature is geopotential height f is the coriolis parameter R is the gas constant for dry air m is the map scale factor for either the Polar Stereographic Lambert Conformal do or Mercator map projections 9 4 and Fy and Fy represent the effects of horizontal and vertical diffusion and p Ps Pt Continuity and Sigmado
51. t 9 Equations Op Optu m Op v m Op a me Pe t y TE 4 The vertical integral of Equation 4 is used to compute the temporal variation of the surface pressure in the model Op 1 Optu m 0ptu m Ot 0 Ox Oy After calculation of the surface pressure tendency opr the vertical velocity in sigma coor dinates 9 is computed at each level in the model from the vertical integral of Equation 4 de NO gt Optu m Optu m A KETA A rae add 6 where o is a dummy variable of integration and 0 0 Thermodynamic Equation and Equation for Omega w The thermodynamic equation is pT _ OptuT m OptvT m Op T oa oF ae RT w PQ a AL 8 Colo F P Past Cpm ss l 7 where Cpm is the specific heat for moist air at constant pressure Q is the diabatic heating FyT represents the effect of horizontal diffusion FyT represents the effect of vertical mixing and dry convective adjustment and w is dp dra 9 where dt t w AD dp p m Op z The expression for Cpm Cp 1 0 8q where c is the specific heat at constant pressure for dry air and q is the mixing ratio of water vapor Hydrostatic Equation The hydrostatic equation is used to compute the geopotential heights from the virtual temperature Ty Og de qr fi PONE ss py ge A 11 Sinto pulp i 11 where T T 1 0 608q dv qc and q are the water vapor cloud water or ice
52. t introduce the model s grid configuration The modeling system usually gets and analyzes its data on pressure surfaces but these have to be interpolated to the model s vertical coordinate before input to the model The vertical coordinate is terrain following Figure 1 meaning that the lower grid levels follow the terrain while the upper surface is flatter Intermediate levels progressively flatten as the pressure decreases toward the top of the model A dimensionless coordinate is used to define the model levels where p is the pressure p is a specified constant top pressure p is the surface pressure o ee 1 Ds Pt It can be seen from the equation and Figure 1 that is zero at the top and one at the surface and each model level is defined by a value of The model vertical resolution is 1Y 1 Y JX 1 1 J gt 1 JX Figure 2 Schematic representation showing the horizontal Arakawa B grid staggering of the dot and cross grid points defined by a list of values between zero and one that do not necessarily have to be evenly spaced Commonly the resolution in the boundary layer is much finer than above and the number of levels may vary upon the user demand The horizontal grid has an Arakawa Lamb B staggering of the velocity variables with respect to the scalar variables This is shown in Figure 2 where it can be seen that the scalars T q p etc are defined at the center of the grid box while the eastward u
53. tarring regcm tar gz will create a main directory called RegCM and several subdi rectories containing all the files needed for pre processing running the model and post processing Preprocessing programs are in the The RegCM PreProc Terrain and RegCM PreProc ICBC sub directories the model source code is in the Reg CM Main sub directory and the postprocessing programs are in the RegCM PostProc and RegCM PostProc v5d sub directories 6 2 Pre processing Several pre processing steps are necessary before running a simulation These steps involve setting up the model domain and creating the necessary initial and boundary conditions files 6 2 1 Setting up the domain The first step is to define the domain and interpolate elevation and land use data to the grid This is done in the RegCM PreProc Terrain sub directory For this practice run we use a south Asian domain of 6300 km x 7440 km size centered over India 21 09 N 78 0 E and a horizontal grid point spacing of 60 km The domain parameters are defined in the domain param file and the values used for practice run are listed in Table 12 27 STEP 1 Link the necessary data files to the RegCM PreProc DATA sub directory go into the DATA subdirectory cd RegCM PreProc DATA edit the datalinker script using a text editor such as xemacs xemacs datalinker x execute the datalinker script datalinker x STEP 2 Go into the Terrain sub directory and edit the domain param
54. te of first simulation idatel restart date idate2 end date of restart simulation nslice number of days for next model run Timestop parameters radfrq time step for radiation model abemh time step for LW absorption emissivity abatm time step for lsm dt time step for atmosphere model ibdyfrq lateral boundary conditions frequency ifsave save output for restart savirq time interval to save output for restart hr iftape save atmospheric output tapfrq time interval to save atmospheric output hr ifrad save radiation output radisp time interval to save radiation output hrs ifbat save surface model output batfrq time interval to save surface model output hrs ifprt printer output prtfrq time interval for printer output hrs kxout k level of horizontal slice for printer output jxsex j index of the north south vertical slice for printer output iotyp Output format 1 direct access 2 sequential ibintyp 1 big endian 2 little endian ifchem save tracer model output chemfrq time interval to save tracer model output hrs 33 Table 7 List of physic options in regcm in file Physics paramotor iboudy ibltyp icup igcc ipptls iocnflx ipgf lakemod ichem SUBEX parameter neld fcmax qck1lland qckloce gulland guloce rhmax rh0oce rh0land cevap caccr lateral boundary conditions 0 fixed 1 relaxation linear 2 time dependent 3 time and inflow outflow dependent 4 sponge 5 relaxati
55. th and on the atmospheric sta bility in the surface layer The surface evapotranspiration rates depend on the availability of soil water BATS has 20 vegetation types Table 2 soil textures ranging from coarse sand to intermediate loam to fine clay and different soil colors light to dark for the soil albedo calculations These are described in Dickinson et al 1986 11 2 2 3 Planetary Boundary Layer Scheme The planetary boundary layer scheme developed by Holtslag et al 1990 is based on a nonlocal diffusion concept that takes into account countergradient fluxes resulting from large scale eddies in an unstable well mixed atmosphere The vertical eddy flux within the PBL is given by Fe K r 12 where y is a countergradient transport term describing nonlocal transport due to dry deep convection The eddy diffusivity is given by the nonlocal formulation z2 Ke kw z 1 J 7 13 where k is the von Karman constant w is a turbulent convective velocity that depends on the friction velocity height and the Monin Obhukov length and h is the PBL height The countergradient term for temperature and water vapor is given by der wih where C is a constant equal to 8 5 and is the surface temperature or water vapor flux Equation 14 is applied between the top of the PBL and the top of the surface layer which is assumed to be equal to 0 1h Outside this region and for momentum y is assumed
56. ution of the global terrain and land use data 1 1 degree 144 5 minute 4 30 minute 400 3 minute 36 10 minute 900 2 minute vegetation dataset 13 MM4 Vegetation 20 GLCC Vegetation must set ntypec 36 map projection LAMCON Lambert Conformal POLSTR Polar Stereographic NORMER Normal Mercator ROTMER Rotated Mercator h2ofrac if water fraction lt h2ofrac then land else water ifanal true perform cressman type objective analysis false perform 16 point overlapping parabolic interpolation smithbdy nent IDATEI IDATE2 SSTTYP SST dataset OISST OLNC GISST OLWK for FVGCM RF AQ global analysis dataset ECMWF FVGOM NNRP1 FNEST NNRP2 igrads true output GrADS control file ibigend 0 little endian binary computer PC LINUX DEC 1 big endian binary computer SUN SGI IBM PC LINUX with PGI or IFC IOMEGA use vertical winds in boundary conditions FUDGE land sea mask fudge true or false 31 Table 4 List of output variables from Terrain DOMAIN ht Surface elevation m hisd landuse xlat lon diat dion nap dmap coriol snowam Table 5 List of variables in ICBCYYYYMMDDHH files sac hPa Specific moisture kg kg Surface pressure hPa Surface air temperature K 32 Table 6 List of restart timestep and output parameters defined in regcm in file ifrest true or false for restart simulation idateO start da
57. water vapor The model has a vegetation layer a snow layer a surface soil layer 10 cm thick or root zone layer 1 2 m thick and a third deep soil layer 3 m thick Prognostic equations are solved for the soil layer temperatures using a generalization of the force restore method of Deardoff 1978 The temperature of the canopy and canopy foilage is calculated diagnostically via an energy balance formulation including sensible radiative and latent heat fluxes The soil hydrology calculations include predictive equations for the water content of the soil layers These equations account for precipitation snowmelt canopy foiliage drip evapotranspiration surface runoff infiltration below the root zone and diffusive exchange of water between soil layers The soil water movement formulation is obtained from a fit to results from a high resolution soil model Climate Processes and Climate Sensitivity 1984 and the surface runoff rates are expressed as functions of the precipitation rates and the degree of soil water saturation Snow depth is prognostically calculated from snowfall snowmelt and sublimation Precipitation is assumed to fall in the form of snow if the temperature of the lowest model level is below 271 K Sensible heat water vapor and momentum fluxes at the surface are calculated using a standard surface drag coefficient formulation based on surface layer similarity theory The drag coefficient depends on the surface roughness leng
Download Pdf Manuals
Related Search