Home

ATMOSYS User Manual

1. the solid circle has radius 1 the dashed circle has radius 0 5 The percentage of the stations that is inside the solid circle and that represent stations that fulfil the model quality objective criterion is shown in the upper left corner of the target plot In the upper right corner are the values of the parameters that are used for calculating the observation uncertainty that are read from the configuration file In addition the chosen period time aggregation in this case hourly and pollutant name are given in the subtitle Target plot Target plot Stations within crit T 1 100 NO2 Hourly values 01 01 2009 31 12 2009 2 0 56 100 0 BIAS gt 0 U 24 0 N Alpha 0 04 RV 200 0 ugin 3 1 8 1 6 1 4 1 0 Me m ES 0 8 0 4 0 2 BIAS 0 0 ay 0 2 0 4 i 0 8 j g 1 0 Be eet 1 2 1 4 1 6 1 8 BIAS lt 0 2 0 1 5 1 0 0 5 0 0 0 5 1 0 lio 2 0 CRMSE m 40HB23 e 41B004 Figure 33 Target plot Page 46 F ATM OSYS ATMOSYS User Manual sir Quality Mocellng System 6 16 Summary statistics In the summary statistics tab Figure 34 different statistics are computed for the stations The first column shows three different categories OBS Statistics for the given observations TIME Statistics computed for each of the stations using the given time transformed values see section 6 13 e SPACE Statistics computed for the whole model domain using
2. TMOSYS Air Quality Modelling System ATMOSYS User Manual ATMOSYS User Manual Ref Deliverable 7 v03 23 04 2015 Smeets Nele Van Looy Stijn Blyth Lisa Partners Lisa Blyth rm vito Email lisa blyth vito be Tel 32 14 33 67 57 2795 http www vito be http www life atmosys be Acknowledgements _ Laboratoire Central de Surveillance de la Qualit de l Air http www lcsaa or VITO http www vmm be http www irceline be http ec europa eu environment life ATMOSYS Alr Quality Modeling System 1 2 3 4 Introduction Home Navigation through the Web Application Interesting News Services Daily Forecast 4 1 4 2 4 3 4 4 4 4 1 4 4 3 4 5 Overview Parameters Changing the forecast date Visualisation of the Air Quality Forecast Maps Legend Changing the Map in the Map Viewer Changing the Pollutant Forecast Validation od 5 2 5 3 5 4 5 5 5 9 1 5 5 2 5 5 3 5 6 2 7 5 7 1 5 7 2 5 7 3 5 7 4 5 7 9 5 8 5 8 1 Overview Select Validation Period Select Pollutant Selection of Measurement Station Types Selection of Station Observation Data The Measurement Stations Map Viewer Map Viewer Legend Measurement Station Table Perform Validation Create Validation Charts Change Forecast Day Chart Functionality Chart type 1 Time Series Chart type 2 Scatter Plot Chart Values Calculation of the Validation Statistics Export Model Evaluati
3. 24h av EC conc starting at 01 01 2009 00 00 in ug m Figure 48 The Time Evolution maps page Page 66 5 ATMOSYS ATMOSYS User Manual 8 2 Loading New Maps By default the EC annual average concentration map for 2009 is shown in the main viewer In addition by default the first 14 days of the last available map is chosen New dates can be selected as explained in section 4 3 and shown below in Figure 48 Once the start date is chosen the end date automatically defaults to 14 days later Start date End date 01 01 2008 14 01 2006 January 2009 o Su Mo Tu We Th Fr Sa 1 3 8 H 1 16 17 P 24 30 31 Figure 49 Changing the date As for the previous services the pollutant can be changed via the pollutant drop down list as shown below Start date End date Pollutant 01 01 2005 14 01 2005 Send Link 8 3 Loading map data Once the period and pollutant are chosen click Load data to load the map and the 335 maps will be retrieved from the databank As this can take a few seconds a loading notifcation as shown in Figure 50 will appear and the the first hourly map 00 00hrs on the first day chosen will appear shortly after in the map viewer The title and legend for the first 24 hr moving average map 00 00hrs on the first day is given on the bottom left hand corner of the viewer Figure 51 Page 67 TMOSYS ATMOSYS User Manual i Modeling System Figure 50 Loading new Hourly Maps Page 68
4. ATM OSYS ATMOSYS User Manual Start date End date Pollutant 01 01 2009 14 01 2009 PMs Load Data Send Link K HA 24h av PM2 5 conc starting at A 01 01 2009 00 00 in ug m Figure 51 The first 24h average map 00 00hrs on the first day for PM for the period 1 January 2009 to 14 January 2009 appears in the map viewer 8 4 Viewing the Animation www To start the animation press the PLAY button above the map The 335 map views will be sequentially shown in the viewer age 09 ATM OSYS ATMOSYS User Manual The speed can be altered using the Speed slider bar shown at the top left hand side of the map To scroll through the 335 hourly views manually one by one use the fast forward and rewind 0600 Hereunder the 211 24h moving average map 19 00hrs on the 9th day thus showing the 24h average for the period starting at the 9 day 19 00 until the 10 day 18 00 for PM for the chosen period is shown The pollution episode is clearly visible in the north region of the map 24h av PM2 5 conc starting at 09 01 2009 19 00 in pg m Figure 52 The 211 24h moving average map 19 00hrs on the 9 day thus showing the average for the period starting at the 9 day 19 00 until the 10 day 18 00 for PM for the period 1 January 2009 to 14 January 20019 the episode is clearly visible in the north of the map Page 70 ATMOSYS ATMOSYS User Manual Alr Quality Modelling System 8
5. 26 01 2012 03 10 2013 a calendar component is shown Forecast base time 03 10 2013 i lt lt lt October 2013 gt gt gt x sun Mon Tue Wed Thu Fri Sat 40 29 30 1 2 3 41106 7 8 9 10 11 412 42 13 43 20 21 22 23 24 25 26 44 26 28 29 30 31 2 45 3 4 5 6 7 8 9 03 10 2013 Clean Today Page 12 ATMOSYS ATMOSYS User Manual 5 2 Air Quality Modelling System Another date can be selected by clicking on one of the available days Days outside the available period are disabled and cannot be chosen The arrows lt gt at the top can be used to navigate to the previous or next month The double arrows lt lt gt gt can be used to go to the previous or next year When a new date is chosen e g 02 10 2013 you must click on the Load button to load the corresponding maps The currently selected date is then shown in the title of the map shown in the main viewer Forecast 02 10 2013 Forecast day 0 NO2 Daily maximum Figure 5 Selection of another date 4 4 Visualisation of the Air Quality Forecast Maps Each day the service provider computes forecasts for three consecutive days starting from the current day E g on October 3 forecasts are computed for October 3 day 0 October 4 day 1 and October 5 day 2 At the top of the page the three images are shown that correspond to the three forecast days Each image shows the model output for the comp
6. 5 Elementary Carbon EC Nitrogen Dioxide NO and Ozone Q3 This demonstration system is established for the Belgian Inter regional environmental agency and contains at the time of writing only maps for the year 2009 The system is established so that future maps can be easily made and uploaded based on the 2009 set up These high resolution maps are generated using RIO an interpolation model coupled with IFDM a bi gaussian plume model The exception is the map for the pollutant EC which was computed using AURORA a chemical transport model coupled with IFDM In line with the air quality legislation several specific statistical indicator maps for each pollutant like the number of exceedences of the daily threshold allowed for PM are provided Page 60 ATMOSYS User Manual Services Annual air quality maps Reanalysis year Pollutant Indicator 2009 ec m Average Average 2009 air quality map EC Average pg m e a 7 E uw y Ma ty v 7 I Eh ds n ji m d D f Services Daily air quality forecast MyEnvironment VITO ATMOSYS is a LIFE Environment Policy amp Governance Forecast validation Copernicus project co financed by the European Commission The Exposure calculation FAIRMODE project began on the 1st September 2010 and will be Annual air quality maps MyAir Pasodoble completed at the end of 2013 Time Evolution maps Timeseries Model evaluation Figure 43 An
7. A fourth output of the exposure calculation is a cumulative distribution plot of the exposure On the horizontal axis the threshold value varies from O to the highest statistic value found in the calculated statistic The vertical axis denotes the percentage of the population exposed to a value above the corresponding threshold on the horizontal axis The threshold value given as an input parameter is not used here An example is shown in Figure 66 Cumulative distribution 100 de y gt 80 rm v v 60 v gt 2 40 C PE v CL X B c 20 o L 21 o a 06 10 20 30 40 50 60 exposure threshold value microgperm3 Figure 65 The Cumulative Distribution Plot of the Overall Average for NO for the chosen period 7 9 Output exposure summary As a final output a summary of the calculated human exposure is given This summary shows the total number of inhabitants in the region for which the exposure is calculated This region is the one for which both population density and calculated statistic data are available For the average exposure option the average exposure per inhabitant in that region is shown For the exceedings option it shows the no of exceedences of the daily average above the selected threshold value This last value represents a single point in the cumulative distribution plot Page 87 is TMOSYS ATMOSYS User Manual P sir Quality Modelling System Total number of
8. the bottom of the calendar and the data will be wiped clean The procedure for selecting the end time is the same as the one for selecting the start time Page 81 1 ATMOSYS ATMOSYS User Manual Alr Quality Modelling System The selected start and end times are both included in the exposure calculation For example to calculate the exposure for one day the start time must be 00 00 and the end time must be 23 00 The time stamps are expressed in universal time which is Greenwich mean time without daylight Saving time 7 3 Selecting a pollutant The pollutant can be selected by using the drop down box which shows the available pollutants Pollutant NOZ Q3 PM10 PM2 5 7 4 Selecting a statistic The exposure tool calculates the human exposure to a certain statistic calculated from the modeled pollutant concentrations Currently 2 statistics are available the overall average concentration over the selected interval in time and the number of exceeding over the daily average concentration over the selected interval in time The statistic to use is selected by using the drop down box If exceedings are selected the threshold value over which the daily average must rise to trigger an exceeding must be given as well If overall average is selected this field can be left empty as shown in the second figure below Statistic Overall average T Overall average Exceedings of daily Start time End time Pollutant 21 01 2013 00 00
9. 15 08 2013 31 08 2013 50 a a 45 40 E 3s B e e Pot a Station BETRO20 observations 3 10 gt Station BETRO20 modal output 5 e Station BETN121 observations 8 25 Station BETN121 model output t Station BETRE4 1 observations e 20 Station BETRS41 modal output 5 e Station BETM702 observations V 15 7 Station BETM702 model output A DA A s ET a gt A Q 21 08 2013 0 28 08 2013 0 Time hour Scatter plot Air quality forecasts for Belgium Forecast day 0 NO2 15 08 2013 31 08 2013 5tatton BETRO20 Station BETN121 Station BETRE41 s Station BETM 702 a ye2x y O05 x Regression Station BETRO20 Ri Q 27 Regression Station BETN121 Ri 0 28 Regression Station BETRB41 Ri 0 01 Regression Station BETM702 R 0 39 Model values 0 5 10 15 20 25 30 35 40 45 se Observations LJ Compute Dad averancs Figure 23 New Validation Charts for the daily Averge Page 32 F ATM OSYS ATMOSYS User Manual sir Quality Mocellng System The following possibilities are available daily average daily maximum e daily maximum of the eight hour averages monthly average when the selected period is longer than one month average day and night concentration 24 values where the i th value is the average of all observations model values corresponding to the i th hour of the day for all days in the selected period E
10. 25 01 2013 23 00 NOZ Statistic Exceedings threshold pq m Overall average X Calculate Exposure The threshold value is typed into the field by the user once the exceedings of the daily average value is picked In the figure below a value of 50 has been entered into this field Page 82 ATMOSYS ATMOSYS User Manual E Air Quality Modelling System Start time End time Pollutant 01 11 2013 12 00 05 11 2013 12 00 li PM10 Statistic Exceedings threshold ug m Exceedings of daily average 50 Calculate Exposure The exposure tool calculation is started after the user clicks the Calculate Exposure button A progress bar is shown while the calculation is in progress This can take a few moments depending on the period length chosen Exposure tool of PM10 from 22 01 2013 12 00 until 25 01 2013 12 00 7 5 Output population density The output is presented in 2 maps and 2 charts The first output of the exposure calculation is a map showing the population density in inhabitants per km The population is resampled on the same grid that is used by the air quality service This allows the user to compare the population data and the concentration map as well as the calculated statistic which is also computed on the same grid as described in the following section The exposure is calculated for the region for which both population density data and the calculated statistic are available The population map
11. EC PM Emission Inventories Since this system focuses on modelling air quality in city hotspots one of the first tasks of the project was to establish the first EC emissions inventory for Flanders An overview of the methodology used is provided under this EC PM emission inventories page The EC emissions were geographically distributed over the Flanders region for input for the AURORA model for the provision of both annual assessment maps for EC and daily forecast maps The maps are available for view in the retrospective services annual air quality maps and daily air quality forecast services 2 Data assimilation techniques Within ATMOSYS two different DA techniques were implemented for the correction of the results from the regional chemical transport model AURORA For correcting the air quality forecasts a Kalman Filter KF based air quality forecast bias adjustment proved to be the most suitable data assimilation technique The final daily air quality forecast services shown are calculated using the KF data assimilation technique The Optimal Interpolation OI technique based on the Hollingsworth L nnberg method was used to correct the annual air quality maps produced with AURORA the chemical transport model 3 High resolution air quality modelling High resolution air quality modelling is required to access in detail the air quality over cities Following rigorous testing it was found that the best annual air quality maps for Belgi
12. Mechelen Background Page 25 ATMOSYS ATMOSYS User Manual E Air Quality Modeling System Validate Once the stations are chosen and you are ready to validate click the button 5 7 Create Validation Charts Once the validation occurs the charts window will automatically open showing two charts that have been computed These charts are as shown in Figure 18 The top one is a time series chart showing the measurement and modeled concentration values for the chosen pollutant over the chosen period The other is a scatter plot for the modeled values vs the observation values Further detail is given below in section 5 7 4 Page 26 ATMOSYS ATMOSYS User Manual Air Quality Modelling System services Air quality forecast validation Start validation End validation Pollutant Station type 31 08 2012 3 NO2 wl Alle types Show stations Validation 15 08 2013 31 08 2013 Forecast day 0 NO2 Stations Charts Statistics Export Time series Air quality forecasts for Belgium Forecast day 0 NO2 15 08 2013 31 08 2013 110 100 sa m Station BETRO20 observations Staion BETRO20 model output Station BETN121 observations Staion BETN12 1 modal output Station BETRS4 observations Station BETRZ41 model output e Station BETM702 observations Stadion BETM702 model output Concentration ug m 21 08 2013 0 28 06 2013 0 Time hour Scatter plot
13. Requested Point 51 21380 N 4 48231 E Zoom id Bd im im 3m 6m m 6m 1 iv Max ec0 no20 030 pm100 pm250 01 01 2010 00 00 180 JN i UM 4 M MM M MU MIA 1 tt TATE WI dd ait Wi bs MM m VUL Th m Figure 56 Timeseries data for the chosen point 51 21380 N 4 48231 E for the year 2009 showing hourly data for May to December 2009 Page 74 P ATM OSYS ATMOSYS User Manual a Quality Modeling System 9 4 Analysis of the data This chart contains the series data for the whole year in this case 2009 for each of the 5 pollutants available The x axis shows the time and the y axis the concentration values At the top right of the chart the colour of the 5 pollutant lines in the graph is shown pm10 41 39 0545 50 noz 30 71 ec 1 64 pm25 9 97 The values beside each point is the concentration value at a particular time which relates to where the mouse pointer lies on the time series chart at that moment In the right corner the date and time for that particular moment where the mouse pointer is is 10 12 2009 06 00 shown On the top left hand side of the map there are various zoom options These allow the user to zoom in and visualise a particular period slice such as a day 1d a month 1m 3 months 3m or a year 1y For example by clicking on 1m only the data for a month is visible Thus now in Figure 57 only the data for December 2009 is shown What appears in this month
14. Timeseries data slice for October 2009 showing only the PM and PM concentration results 9 6 Export the data The time series data in CSV format for the whole year chosen can be downloaded using the export E button shown at the top right of the viewer For each hour of the year the modeled concentration values for each pollutant for that chosen point are available The CSV file uses the English settings which uses the comma as delimiter and the period as decimal separator An excerpt from a CSV file is given below in Figure 60 Page 77 ATM OSYS ATMOSYS User Manual Abr Quality Modelling Syste time ec no2 03 pm10 pm25 2009 01 01 00 00 00 0 742999970913 62 2729988098 0 0 103 206001282 71 2089996338 2009 01 01 01 00 00 0 742999970913 66 3369979858 0 0 85 0999984741 61 5810012817 2009 01 01 02 00 00 0 742999970913 59 0250015259 0 0 88 6529998779 65 1169967651 2009 01 01 03 00 00 0 742999970913 9 10200023651 26 5650005341 78 4550018311 59 9230003357 2009 01 01 04 00 00 0 638999998569 9 55099964142 27 1009998322 81 8690032959 60 858001709 2009 01 01 05 00 00 0 773999989033 9 28499984741 27 8740005493 79 0299987793 57 1529998779 2009 01 01 06 00 00 1 04200005531 6 05399990082 29 5620002747 75 5510025024 53 90599823 2009 01 01 07 00 00 1 40600001812 7 10699987411 27 8050003052 70 9980010986 54 3409996033 Figure 60 CSV file excerpt for the chosen point for the 5 pollutants Page 78 z T ATM OSYS ATMOSYS User Manua
15. User guide P Thunis E Georgieva A Pederzoli Joint Research Centre Ispra Version 2 04 April 2011 Page 96 Air Quality Modelling System ATMOSYS User Manual Appendix 1 Statistics for model evaluation DELTA tool From 2 o List of available statistical indicators Factor of modelled values within a factor of two of observations Centred Root Mean Square error g lyx o uisus N inl N iwl o M LY the limit vaiue concentration LV and M E rhe correzpondingiy ranked modeiied concenmation MRDE Max RDE over 90 a RDE where O is the closest observed concentration to M aes 2 3 where p i the percentile corresponding 10 the allowed L4 number of exceedances of the limit value MRPE Max RPE over 90 of stations Am 0 5 S 0 2 n 0 elze FAC Zn with CRMSE yh M lo oT Mode Efficiency Score MEF 1 RMSE Page 97 ES Air Quality Modelling System ATMOSYS ATMOSYS User Manual Appendix 2 Change log version 3 March 2015 Compared to version 2 the functionality of the benchmarking tool was aligned with version 5 1 of the JRC DELTA tool These changes required a revision of chapter 6 of this manual The main changes are the configuration of the tool is read from the goals criteria oc dat configuration file that is also used by the JRC DELTA tool This is to ensure that eventual changes to the JRC Delta tool can be adopted in the online tool wi
16. average values per station The second column contains six different statistics which are described in the following sections In the case of the TIME statistics the third column contains a green circle when more than 9096 of the blue dots are inside the green or orange range in the fourth column For the SPACE statistics there is only a single dot and the circle will therefore only be green if this single dot is inside the green or orange range in the fourth column The fourth column shows the calculated statistics For the categories OBS and TIME a statistic is computed for each station and each blue dot corresponds to a single station For the category SPACE only one spatial statistic is computed so there is only one blue dot The blue dots for stations for which the model performance criterion is fulfilled lie within either the green or the orange shaded areas If a dot falls within the orange shaded area the error associated with the particular statistical indicator is dominant albeit still small enough for compliance Page 47 MOSYS ATMOSYS User Manual Air Quality Modeling System Summary statistics NO2 Hourty values 01 01 2009 31 12 2009 Nb of stations groups 2 valid 4 selected INDICATOR OBS Mean Exceed 200 D ugm 3 TIME Bias SPACE Corr g e o o o o oO Info about plot data pow Performaces Criteria satesiied Error dominated by D TIME gt 90 of stations fulfils the Performance Critona SPACE D
17. chosen simulation results and given population data in inhabitants per km for region I These tools offer the capability to access visualize and analyse an archive of historic air quality maps for a specific region country The user friendly interface also allows the user to extract the complete set of hourly maps to obtain a better insight into specific pollution episodes To access these individual service demonstrations click on the Service header or the More info tab at the end of each service description Daily forecast This service offers a three day forecast of Particulate Matter PM10 PM2 5 Elementary Carbon EC Nitrogen Dioxide NO2 and Ozone O3 which is designed to cover more European hotspots on request Here you can view a demo of the real time forecasting chain for the Belgian Interregional Cell for the Environment More info Figure 3 Navigation from the Services Page The final service shown is the Help Service Here you will find this user manual some information on the Delta benchmarking tool and examples files for the ATMOSYS model evaluation tool Page 9 ATMOSYS ATMOSYS User Manual E Air Quality Modeling System 4 Daily Forecast 4 1 Overview Under the air quality forecast service tab a demonstration of a 3 day forecast for Particulate Matter PM10 PM2 5 Elementary Carbon EC Nitrogen Dioxide NO2 and Ozone 03 is provided The maps shown here at the launch of the system a
18. different parameters used in the measurement uncertainty calculation are listed that correspond to that pollutant name alpha U RV as well as the limit values for the number of exceedances and the data availability criterion These parameter values are read from the goals criteria oc dat configuration file which is also used by the JRC DELTA application pi Target plot Step 1 Read period and pollutants from input files Step 2 Choose pernod and poliutant and compute target plot Start validation End validation Pollutant Figure 31 Target plot tab screen after parameters have been computed 6 11 Compute Target Plot The computation of the target plot can be started by clicking the Compute Target Plot button This button is disabled until the user clicks the Read file Parameters button After clicking the button all uploaded model value and observations files are read and the selected validation period is extracted In the selected validation period at least 75 of the model observation tuples for a given station should be effective otherwise the station is ignored When a model value file and observation file have the same name then they are associated with each other When a model file has no corresponding observation file then this file is ignored and the other way round When all files are read the target plot and summary statistics are computed for the stations for which both a model value and observation file is availa
19. dragging Scale The scale of the map is shown in the bottom left corner Transparency The transparency of the concentration map can be changed by using the slider in the header bar above the map viewer Figure 7 The transparency slider 4 4 3 Changing the Map in the Map Viewer By default forecast day 0 for the pollutant NO2 daily max concentration is selected in this map viewer When the user wants to study one of the three forecast days in more detail he clicks on one of the three images The corresponding map is then shown in the map viewer below Page 14 TMOSYS ATMOSYS User Manual Air Quality Modeling System Figure 8 and Figure 9 show forecast day 1 and forecast day 2 respectively The selected forecast day is surrounded by a dotted border and the title above is shown in a larger font e 2i Expertise About Contact ATMOSYS Home fur Quabty Modeling System Services Air quality forecasts for Belgium Forecast base time Pollutant yg m 03 10 2013 gt NO2 Load Available 26 01 2012 03 10 2013 Daily maximum Forecast day 1 Forecast day 2 Forecast day 0 Forecast 03 10 2013 Forecast day 1 NO2 Daily maximum emjee we a a NET VIE e Y m VIRE 4 h V4 CON Figure 8 Visualization of forecast day 1 in the main viewer 4 October Page 15 Air Quality Modeling System ATMOSYS Home News MER Expertise About Contact Aor Quality M
20. inhabitants in the region 5886374 people Average exposure per inhabitant 47 136 pg m Figure 66 Summary Statistics Output Page 88 ATMOSYS ATMOSYS User Manual Air Quality Modelling System 8 Help Service This user manual can be downloaded from the help page of the web application A link to the document is put on the help tab In addition the following documents can be downloaded e Two articles about the JRC delta tool e Anexample CSV file for the benchmarking model evaluation tool Figure 67 The help page contains links to the user manual articles and example files ATMOSYS Help Please oownioad our user manual of contact us if you need more mdormabon on now to use me Almosys Web Sil User manual Articles Sample files amp Links Links Project Ivemaromnent VITO ATMOSYS 1 a LIFE Emtienmeen Polic amp Govemance project co financed dy ne European Commission The protect began on the 121 Septemmer 2010 anc wel be completed si me end of 201 Pee RME Hn M Lime eT DE mg rey Figure 67 Help page Page 89 ATM OSYS ATMOSYS User Manual Alr Quality Modeling System 9 Expertise This page presents an overview of the scientific studies which are the backbone on which the ATMOSYS system was developed Within each section further detailed information is available for download or contact information is provided if more information is required There are 5 main topics 1
21. shows the concentration in ug m The model results are connected by a spline The period that is shown corresponds to the validation period that was selected by the user The title of the chart shows the name of the service the forecast day the pollutant and the validation period 5 7 4 Chart type 2 Scatter Plot Scatter plot Air quality forecasts for Belgium Forecast day 0 NO2 15 08 2013 31 08 2013 Station BETRO20 Station BETN121 Station BETRE41 a Station BETM702 Model values vV 0 5x Regression Station BETRO20 R 20 27 Regression Station BETN12 1 R 0 28 Regression Station BETRS4 1 Ri 0Q 01 Regression Station BETM702 R 0 39 25 30 Observations Figure 21 Scatter plot example An example of a scatter plot is shown in Figure 21 For more information about this type of chart http en wikipedia org wiki Scatter plot This chart contains both observations and model output for the selected station s Each point represents a model concentration and the corresponding observation for the same timestamp The value of the model concentration is shown on the Y axis the value of the observation is shown on the X axis The chart shows data for the selected validation period The identity line y x is drawn as a reference Two other lines y 2x and y 0 5x are also shown For each station a line of best fit or trend line is shown in the same color as the m
22. 2 9 2 Choosing a Location 73 9 3 Loading the time series 74 Page 3 F ATM OSYS ATMOSYS User Manual sir Quality Modelling Syste 9 4 Analysis of the data 75 9 5 Changing Viewing of the Pollutants 77 9 6 Export the data 77 10 Exposure tool 79 10 1 Overview 79 7 2 Selecting start and end time 80 7 3 Selecting a pollutant 82 7 4 Selecting a statistic 82 7 5 Output population density 83 7 6 Output calculated statistic 84 7 7 Output histogram 85 7 8 Output cumulative distribution 87 7 9 Output exposure summary 87 8 Help Service 89 9 Expertise 90 10 About 93 11 Contact details 94 12 References 96 Appendix 1 Statistics for model evaluation DELTA tool 97 Appendix 2 Change log 98 Page 4 ATM OSYS ATMOSYS User Manual Air Quality Modeling System 1 Introduction The ATMOSYS web application is developed as part of the LIFE ATMOSYS project http www life atmosys be The core goal of the ATMOSYS LIFE project is to demonstrate innovative solutions and knowledge specifically designed to evaluate and analyse air pollution in European hotspot regions This demonstrative system has being established for the European hotspot region of Flanders north Belgium by VITO Belgian scientific research institution together with the Flemish Environment Agency VMM This system is a test environment for a multitude of state of the art air quality services After a process of rigorous testing some will become part of the off
23. 5 Share the Animation Video To share a particular episode video use the send link button to obtain the link which can be copied and sent by mail Send Link Copy this link to share this video hwww atmosys eu atmosys faces services hourly maps jsp startDate 01 01 2 Figure 53 After clicking the send link the link appears in a box to copy and share Page 71 Ed ATMOSYS ATMOSYS User Manual Alr Quality Modelling System 9 Time Series 9 1 Overview Environmental agencies are often asked to provide detailed concentration data for specific pollutants for a particular location within their region To assist the agencies with this task a tool was developed which allows this data to be easily extracted and visualized from an archive of hourly air quality concentration maps which were generated using air quality models Within ATMOSYS this tool is linked to the hourly archive of maps produced for the Belgian Interregional agency Using this tool the hourly concentration data for a whole year can be accessed The user can then zoom in on specific time intervals and compare the different pollutants with each other ATMOSYS Services Services Project Partners vito Figure 54 The Time Series Service Page Page 72 ATMOSYS ATMOSYS User Manual Air Quality Modeling System 9 2 Choosing a Location The first step is to choose the location This is done by clicking on a specific location on the OpenStre
24. 7 TMOSYS Air Quality Modeling System 3 Services ATMOSYS User Manual Here a short overview of each of the seven services is given The services are split into two main categories Forecast and Assessment Services as shown by the selection pane on the left are OSY Air Qualty t aleling iere Services Forecast Service Daty forecast Forecast yvalctaton Assessment Service Annus maps Time Evolution maps Timesenes Mode evaluator Exposure cak wiator Heip Home News Services Expertise Abou X Contact ATMOSYS Forecast Service This service offers demo of me daily ai quality forecasting that was estabasned for the Beigian imerregional Cet for the Environment and is designed to cover more European hotspots on request A ihrer day forecast of Particulate Matter PM 30 PM2 5 Elementary Carbon EC Nitrogen Dioude NO2 and Omne 03 5 supplied together wif a validation tool and a human exposure calculation Daily forecast This service offers a tree day forecast of Particutate Matter PM10 PMZ 5 Elementary Carbon EC Nitrogen Diode NO2 and Ozone O3 which ss designed to cover more European hotspots on request Here you can view a demo of tie real time forecasting chain for the Beigian Inferregionat Celi for the Environment More info Forecast validation The forecast validation service allows a user to evaluate the concentrations focecasted by the ar quality model using measurement data E is d
25. A Govaemances provedt co hnanced be e European C mrmssion The Died began on Ihe 1st September 2010 and will be Inda sam guzy mar Partners PA vito Figure 61 Exposure Calculation Page Before Starting a Calculation 7 2 Selecting start and end time The start time can be set as follows After clicking the calendar icon next to the start time field a calendar appears on the screen as shown in the figure The arrows on top can be used to navigate through the months A date can be selected by clicking on it Only the dates for which model data is available can be selected Page 80 P ATMOSYS ATMOSYS User Manual Alr Quality Modelling System Start time 05 11 2013 00 00 Imm lt lt November 2013 gt gt gt x Sun Mon Tue Wed Thu Fri Sat 44 1 12 zi uum 45 3 4 M3 45 AT 48 49 05 11 2013 Clean 00 00 Today The hour of the day for which the exposure calculation must start can be refined by clicking the 12 00 time value on the bottom of the calendar A pop up as shown in the figure below appears and the time can be fine tuned using the arrows next to the fields Since the model data consists of hourly values the selected minutes in the time do not matter and are ignored Start time 05 11 2013 00 00 Tl lt lt lt November 2013 gt gt gt x Ln 00 a 00 OK Cancel To fix the chosen time click the OK button To cancel and start again press the FAI button at
26. Air quality forecasts for Belgium Forecast day 0 NO2 15 08 2013 31 08 2013 110 Station BETRO20 Station BETN121 Statton BETR 41 Station BETM702 Yeux y 2x Model values y 05 x Regression Station BETRO20 R1 0 14 Regression Gtation BETN12 1 R 0 15 Regression Gtation BETRE4 1 R1 0 06 Regression Station BETM702 R1 0 21 E 10 20 3 4 50 0 70 9 90 190 100 Observations Honori value Em Compute Figure 18 The Validation charts for the period 15th August until 31st August 2013 Page 27 Ed ATMOSYS ATMOSYS User Manual 5 7 1 Change Forecast Day By default the first forecast day Forecast day O is selected When the user wants to validate another forecast day he selects one of the radio buttons at the top of the Stations tab see Figure 19 Validation 15 08 2013 31 08 2013 Forecast day 0 NO2 Stations Charts Statistics Export Forecast day 0 Forecast day 1 0 Forecast day 2 Figure 19 Selection of the forecast day When clicking the Validate button after this the charts and statistics will use the selected forecast day The forecast day that is used in the computations is shown in the map Validation 15 08 2013 31 08 2015 Forecast day 0 NO2 and chart titles ime series Air quality forecasts for Belgium Forecast day 0 NO2 15 08 2013 31 08 2013 Page 28 ATMOSYS ATMOSYS User Manual Alr Quality Modell
27. I project Is a ist environment tra of the oficial forecasts and assessments on amblant alr quality In Belptum as published by Belgian interregiona M vironment xi RCEL CELNE and Flanders as published by the Flemish Environment Agency VMM Copyright 2 Figure 1 Home page Page 6 Ed ATMOSYS ATMOSYS User Manual 2 Air Quality Modelling System Navigation through the Web Application From this page the various services and information can be accessed The web application is spilt into 5 main parts News Services Expertise About Contact Po E The simplest way to access each part is by using the banner above News Services Expertise About Contact However you can also click on one of the main headings in the first half of the page to reach respectively the Services Expertise and About pages Discover our solutions Learn more Background Finally you can also take a short cut to one of the services by clicking on one of the individual services in the list at the bottom of the page Services Daily air quality forecast Forecast validation Exposure calculation Annual air quality maps Time Evolution maps Timeseries Model evaluation Interesting News In the second half of the page the most recent interesting items relating to the project and or application are shown At the time of writing the ATMOSYS film is available to view and download alongside a glimpse of the latest news items Page
28. IENNE INDUSTRIAL BETECCI BETECCI UNKNOWN senEut 41HEU1 MEUDON BACKGROUND BETMEOZ AZMBOZ ANTWERPEN NOUSTRIAL BETNO43 41NO43 HAREN WOUSTRIAL BETNOTO 43N07O MONS BACKGROUND BETRIO7 AGREST ANTWERPEN NOUSTRIAL news ES exer About Contact ATMOSYS User Manual Figure 14 Validation Page showing all the NO2 measurement stations available for the Forecast 0 day for the period 15 August until 31 August 2013 Page 22 ATMOSYS ATMOSYS User Manual E Air Quality Modelling System 5 5 1 The Measurement Stations Map Viewer The map shows the available measurement stations and modeled time series output for the chosen pollutant the specific forecast day day 0 1 2 and time period see Figure 14 above The geographical data is visualized in a GIS viewer that allows the user to perform typical GIS operations as explained in section 4 4 2 5 5 2 Map Viewer Legend There are different types of stations Each type has a different color purple industrial orange background blue traffic and green unknown These colors correspond to the colors that are used on the European Air quality database website EIONET AirBase The corresponding legend is shown on the top right hand corner of the map as shown in Figure 15 industrial trafik background unknown Figure 15 The legend for the measurement stations 5 5 3 Measurement Station Table The table underneath this map shows the same list of statio
29. NSPIRE compatibility The ATMOSYS daily operational air quality forecasting system must be capable of obtaining measurement data from the Belgian Interregional Environment Agency s IRCEL CELINE http www irceline be database for use in the data assimilation scheme and in publishing and sharing the resulting air quality maps according to INSPIRE guidelines Here a brief overview is given as well as how it fits into the Belgian Interregional Environment Agency s IRCEL CELINE overall eReporting and INSPIRE compliance services work 6 Microscale air quality modelling To analyse air pollution at local it is often needed to zoom in at the level of individual streets buildings and human beings This level of detail is referred to as the micro scale Within ATMOSYS several developments were made to our micro scale models and case studies were performed to demonstrate the need and potential of micro scale air quality modelling e Studying of the behavior and concentration of ultrafine particles in cities e The effect of street canyon trees on annual pollutant concentrations e The effectiveness of noise screens to improve air quality at an industrial site Page 91 TMOSYS F Air Quality Modeling System ATMOSYS Air Quality Modelling System Expertise City and Highway Measurement Campaigns EC PM Emission Inventories INSPIRE 3D environmental analysis High resolution modelling Data Assimilation Page 92 ATMOSYS Us
30. Parcecia Ern Target Limit value pg m Number of exceedances 2 Compute European code Name cO MBias RMSE 428029 9472222 1523 2 18 73 435121 OFFAGNE 432 425541 BETES 18 7 MECHELEN 4441792 BETNITOS 1499 EATVEL CE Figure 24 The validation statistics tab Page 34 ATMOSYS ATMOSYS User Manual E Air Quality Modelling System 5 8 1 Export The validation results can be downloaded in the Export tab The results consist of e The selected parameters O service name O Start date o end date o pollutant o forecast day 0 1 or 2 O Stations e The observations and model values for the selected stations In both cases the timestamp and corresponding value are given e The data used for the scatter plot for each selected station The observations are in the first column the corresponding model values are in the second column e The values of the different statistics computed for each station Validation 15 08 2013 31 08 2013 Forecast day 0 NO2 Stations Charts Statistics Export A the following link to download the validation date es C lownload report Dutch setting lownload report I nouan settings The Gate ie exported as CS mme Seperated Valves file Different settings are foreseen The fht CSV file utes the sami cole 8s delimiter and the comme es decimal separalor When ihe regional settings of ine users machine are in Duteh this file can be automatically opened in Exce
31. Re BU css IABO csv Ra Al eroyve AIHB2 ce ADHERI cer Remove Famoyve 416004 ca 418004 cow Penove MK arnavi When the format of one of the files is invalid then an error message is shown above the upload component The message is shown in red and contains the name of the file the line number and a description of the error Model values Error in file 40AB02 csv on line 8759 The given file should contain 8760 rows This file contains only 8758 rows Upload file 6 5 Remove uploaded file An uploaded file can be removed by clicking the Remove button to the right of the file 404B01 csv Remove 40AB02 csv Remove 40HB23 csv Remove Page 40 ATMOSYS ATMOSYS User Manual Alr Quality Modelling System 6 6 Remove all uploaded files By clicking the Remove all button all uploaded model values files and observations files are removed at once Upload Targetplot Summary statistics Export Extra validation Info Check format Remove model value files Remove observation files 6 7 Remove all model value files By clicking the Remove model value files button all uploaded model value files are removed Upload Targetplot Summary statistics Export Extra validation Info Check format Remove all Remove model value files Remove observation files 6 8 Remove all observation files By clicking the Remove observation files button all uploaded observation files are removed Upload Targetplot Summary statist
32. T D chengtadbaxch Ne Awo C n uv Em w UJ UJ fh ty hJ hJ D g Y Figure 47 The map legend The legend shows the meaning of the colors that are used in the maps Page 64 1 ATM OSYS ATMOSYS User Manual Air Quality Modeling System 8 Time Evolution Maps 8 1 Overview One of the user requirements during the latter iterations of the system was to provide a functionality that would allow visualization of specific episodes Instead of loading specific episodes it was decided to provide this functionality for the annual maps that are already loaded into the archive These annual maps are essentially made up of 8760 hourly maps which were generated using RIO an interpolation model coupled with IFDM a bi gaussian plume model Again for EC the maps were produced using AURORA a chemical transport model coupled with IFDM For the time evolution visualization 24h moving average maps are calculated from the hourly map database for the selected period The 24h period over which the average is made starts at the selected hour and ends 23 hours later For maximum performance the period chosen is restricted to two weeks a maximum of 335 hours from 00 00 day 1 to 23 00 on day 14 With a user friendly interface the user is able to scroll through these maps to get an hourly view of a specific air pollution episode that occurred during a period Page 65 7 ATMOSYS ATMOSYS User Manual ATMOSYS m Ti
33. ble and that contain more effective tuples Page 43 ATMOSYS ATMOSYS User Manual Air Quality Modelling System in the selected validation period than required by the data availability criterion which is currently set at 7596 This can take some time 6 12 Stations summary The stations summary panel contains an overview of the uploaded files Figure 32 There are three categories OK Stations for which both a model value and observation file are uploaded For the selected period more than 75 of the model observation tuples are effective e Missing file Stations for which the model value file or observation file is missing Insufficient tuples Stations for which both a model value and observation file are uploaded For the selected period less than 75 of the model observation tuples are effective OK Missing mode value of observation hie nsufficient tuples lt 75 m selected penod Figure 32 Stations summary 6 13 Time transformation Before computing the target plot and the summary statistics the following transformations are computed which are determined from the settings in the configuration file goals criteria oc dat and which are currently set as follows Moving 3 hour average using the values at hours 2 1 O Hourly values raw data Daily maximum of eight hour moving average using the values at hours 7 6 5 4 3 2 1 0 PM10 Daily average 6 14 Minimum data a
34. described in section 5 functionalities to further analyse the uploaded observations and model values Thus Page 55 B ATMOSYS ATMOSYS User Manual ir Quality Modeling System to generate scatter plot and time series charts and run other statistics The statistics will be computed on the time transformed tuples described in section 6 13 This tab only becomes available once the previous tabs Upload Target Plot and Statistics have been used i e only after uploading one or more files in the Upload step section 6 2 checking the format of these files and computing the target plot the Extra validation tab will be enabled By default the last week of the selected period is selected for the pollutant that is selected in the target plot tab The user can select a start and end date and then click the Start validation button When clicking this button the charts and statistics are computed and shown in the Charts and Statistics tabs underneath These charts and statistics correspond to those computed in section 5 Upeoad Farget plot summary Stalstics Export Extra validation In Avallable period Start validation End validation Pollutant Charis Statistc Figure 39 Extra validation tab Page 56 TMOSYS ATMOSYS User Manual Air Quality Modeling System Uposs Targetpot Summayetstetcs Epot extravalidation nto Extra validation Sy clicking fne Start validator ution ackitional Charts time series and
35. e included The online model evaluation tool is composed of six main tabs Upload Target plot Summary statistics Export e Extra validation Info The different functionalities will be discussed in the following sections 6 2 Upload model values and observations The Upload tab see figure 26 allows the user to upload files containing model values at the left and observations at the right The format of these files is explained in section 6 3 Each file contains data for one station The file name should reflect the station name Model value files and observation files are matched based on their file name so files containing data for the same station should have the same name Upload Figure 26 The upload tab Page 36 TMOSYS ATMOSYS User Manual Air Quality Modeling System When clicking on the Upload file button a file dialog Figure 27 is opened which allows the user to select one or more CSV files on its local file system Date mnadiied 24 11 2017 18 29 Microsoft Excel C 4411 2017 14 29 Macrmsen tare C M TL 2 18 29 Micron face C 31 2017 18 79 Mieroveft Excel C 1292 20015 1557 Microscft Excel C 102 2015 1557 Microsoft Encel C 192 2015 1557 Microsoft Encel E Local Duk C ica loca Dak Dx de SRECVCLE BIN air week de AIRINFORM Figure 27 When clicking on File upload a file dialog is opened After selecting the files Figure 28 the Open button shou
36. er Manual Home News Services Expertise About Contact City and Highway Measurement Campaigns EC Typically the existing measurement network of a city 19 is insufficient to provide detailed information to comprehensively assess the air quality in a city hotspot To gain a better insight into the how the concentrations of PM and NO2 vary within a city hotspot between several city hotpots and nearby a busy highway an intensive city and highway campaign was carried out in Belgium in the period July 2011 until February 2013 For a 12 month period in the cities of Ghent Bruges Antwerp and Ostend air quality measurements were taken at 3 different city locations an urban background a regional road and in a street canyon As a result of this study a comprehensive measurement database is available which is being used to gain important knowledge of the air quality variation throughout the city hotspots For instance this study shows that staying under the daily limit for PM10 is still a significant problem in the city s street canyons The measurement database is also being used to validate some of the high resolution air quality modelling assessments being carried out in the region To find out more information visit VMM s website a PMs NO EC PM Emission Inventories A key part of the project was to establish the first elemental carbon EC emissions inventory for Flanders and to generate geographically distributed EC and Part
37. et map provided in the GIS viewer To pin point the location exactly use the typical GIS operations Zooming Zooming is done by selecting the map and then using the mouse wheel to zoom in or out An alternative method is to use the slider at the left of the map window By clicking on the plus or minus sign or by dragging the slider up or down you can zoom in or out Panning Panning can be done by clicking and dragging The scale of the map is shown in the bottom left corner Hover over the chosen location with the mouse and click to pinpoint the location A red pointer will appear at your chosen location Under the map the GPS coordinates of your requested location will be shown alongside the pollutions available for this location as shown in the map below E MOSS Services Timeseries Polutants ug m EC NO2 Op PMi Phas Requested Point 5 2 08 N 4 48174 E Figure 55 Visualisation of the chosen point 51 21168 N 4 49174 E By default all the pollutants available are selected and at this stage cannot be unselected Page 73 TMOSYS ATMOSYS User Manual Air Quality Modeling System 9 3 Loading the time series Click the load button to upload the time series data chart as shown below Figure 56 Services Timeseries amp Ee ot vers D Na Bev ren vA J ee 19 hy UN g A gt gt ot i E Le 4 Me PAN REINE fh JA Pollutants yg m EC M NO2 M O 4 PMs PM
38. eveloped 10 allow a user to validate recent model forecasting results idealy over a few days automaticaly on the fy More info P 7 amp do amp Mo o de mw dt im i he n omo 2 39 2222222 81z1 ATMOSYS Assessment Service This servie offers a demonstrative tool fo access visuaive and analyse an archive of histonc ar quality mags for a speci region country The user triendly interface also allows the user to extract the complete set of hournty maps to obtain beter insignt into specife potubon episodes Annual Maps The service offers annual average concentration maps of Partculate Matter P410 PM2 55 Elementary Carbon EC Figure 2 Services Page Page 8 ui ATMOSYS ATMOSYS User Manual Alr Quality Modelling System Under the Forecast Services there are two main services 1 Demonstration of an harmonized AQ Forecast Service for Belgium 2 Interactive Forecast validation tool that can be used to validate the demonstration forecast The Assessment Service section offers five different demonstrative tools for exploration 1 Archive of annual concentration maps of AQ pollutants for Belgium Time evolution maps of the above 3 Time series tool to extract a time series for a particular pollutant for a specific location from the maps above 4 Model evaluation tool to evaluate AQ models based on the FAIRMODE Delta tool 5 Exposure calculation tool to calculate human exposure to a chosen pollutant based on the
39. extreme events e g exceedances a model performance criterion for percentiles was introduced HPercNorm Mperc E Operc 2U Operc in which M and Operc are respectively the observed and modeled value corresponding to a certain percentile and U O is the observation uncertainty corresponding to the observed value for a specific percentile For pollutants and time aggregation values for which limit values for number of exceedances have been set by legislation these limit values can be used to determine appropriate percentiles In other cases the 95 percentile is used HPercNom 6 23 SPACE Corr Norm First the average observation and average model value is computed for each station Using these average values the normalized correlation is computed as described in section 6 20 The result of the computation is only one value so only one blue dot is shown 6 24 SPACE StDev Norm First the average observation and average model value is computed for each station Using these average values the normalized standard deviation is computed as described in section 6 21 The result of the computation is only one value so only one blue dot is shown 6 25 Formulas Page 50 E ATMOSYS ATMOSYS User Manual RMSU kU a a 02 o2 a RV The parameters k u a RV are read from the configuration file goals_criteria_oc dat 6 26 Configuration file goals criteria oc dat The computation of the different sta
40. f the Overall Average for NO for the chosen period This calculation is based on the modelled simulations 7 7 Output histogram A third output of the exposure calculation is a histogram of the calculated statistic This is a bar chart expressing the percentage of the population exposed to a certain value of the calculated Statistic In the case of the overall average this is the of the population exposed to the calculated overall average values over the region of interest For the exceedings of the daily average value this is the of the population exposed to a number of exceedings of the daily average concentration above the chosen threshold value An example is shown in Figure 64 overleaf Page 85 i TMOSYS ATMOSYS User Manual ES Air Quality Modelling System Histogram 10 population exposed 40 45 50 exposure microgperm3 Histogram 60 Ww L eo o population exposed 95 NJ o 0o 0 5 1 0 1 5 2 0 exposure exceedings of daily average over 25 microgperm3 Figure 64 Top Histogram of the of the population exposed to the calculated overall average NO for the chosen period and model domain Bottom Histogram of the 6 of the population exposed to a number of exceedings of the daily average concentration above the chosen threshold value for the chosen period and model domain Page 86 E ATMOSYS ATMOSYS User Manual Alr Quality Modelling System 7 8 Output cumulative distribution
41. for download This contains an overview of the selected parameters the stations summary the values of each station in the target plot and values of the summary statistics table including the range values The time transformed tuples that are used in the target plot and summary Statistics are listed at the end of the CSV file The values are listed per station This file can be downloaded by clicking on the URL labeled Download report as CSV Upload Targetplot Summary statistics Export Extra validation Info Export Click the following link to download the model evaluation report as PDF Download report as PDF Click the following link to download the model evaluation report as CSV Download report as CSV Click the following link to download the summary statistics as PNG Download summary Statistics as PNG Finally the summary statistics can be downloaded as a single PNG by clicking the last URL Upload Targetplot Summary statistics Export Extra validation Info Export Click the following link to download the model evaluation report as PDF Download report as PDF Click the following link to download the model evaluation report as CSV Download report as CSV Click the following link to download the summary Statistics as PNG Download summary Statistics as PNG 6 28 Extra validation Forecast Validation Tool Functionality The Extra validation tab Figure 39 allows the user to use the forecast validation tool as
42. g when the selected period consists of 8 days then the i th value is the average of 8 values each corresponding to the i th hour of the day 7 values where the i th value is the average of all observations model values corresponding to the i th day of the week where Monday is the first day of the week Tuesday is the second day for all days in the selected period This is only meaningful when the selected period is longer than one week By selecting a new value in the drop down list and by clicking the Compute button the charts will be updated to show the selected option 5 8 Calculation of the Validation Statistics Under the statistics tab a list of available statistic calculations is given This list is based on the core statistics available within the DELTA tool Appendix 1 gives an overview of these statistics The user is able to dynamically choose the statistics that he wants to use A list of all available statistics is shown in the first left hand frame Stations Charts Statistics Export Centred Root Mean Square Error pH Copy all n Factor of modelled values within a factor of two of observations Index of Agreement Mean Bias Mean Fractional Bias bul From this list the user can choose the ones that he is interested in This can be done by double clicking on the statistic or by selecting the statistic and then clicking on the Copy button Multiple statistics can be selected by using CTRL click o
43. icial forecasts and assessments on ambient air quality in Belgium as published by Belgian Interregional Environment Agency IRCEL CELINE and Flanders as published by the Flemish Environment Agency VMM The demonstrative website allows potential users to explore the various services and information available in a real test case environment There are 6 main parts to the website Home e News e Services Expertise About Contact Each of them will be described in a separate chapter A chapter will also be dedicated to each of the services The ATMOSYS web application is available at the following URL http atmosys eu pages ATMOSYS ATMOSYS User Manual Air Cluality Modeling System 2 Home When opening the web application in a browser the home page of the application is shown see Figure 1 This home page contains some general information about the web application ATMOSYS News Services Expertise About Contact lualty Modal Check online air quality forecast What type of air quality can we expect in the next few days View onkne Discover our solutions Learn more Background The ATMOSYS system offers different spatial ATMOSYS provides valuable information for air The core goal of the ATMOSYS LIFE project tools to explore quality experts on the following topics is to demonstrate the solutions and Air pollution forecast 3 days ahead EC emission inventories knowisdge Sp cifica
44. ics Export Extra validation Info Check format Remove all Remove model value files Remove observation files Page 41 ATMOSYS User Manual ATMOSYS Alr Quality Modelling System 6 9 Target plot tab The Target plot tab allows the user to compute a target plot for the uploaded observations and model values There are three panels Parameters Stations summary Target plot The three panels will be discussed in the following sections Upload Target plot Summary statistics Export Extra validation Info Parameters Step 1 Read period and pollutants from input files Read File Parameters Step 2 Choose period and pollutant and compute target plot Start validation End validation Pollutant T Available Stations summary Target plot No data Figure 30 The target plot tab 6 10 Parameters Select amp Validate Parameters To verify what data is available relative to the loaded files the user must first click Read file Parameters The total period and the pollutants available will then appear in the parameters window Figure 31 Page 42 O 3 F ATM OSYS ATMOSYS User Manual sir Quality Modeling System The user can alter the validation period by selecting a start and end date Changing a date is explained in section 4 3 The pollutants available will be shown as options in the pollutant drop down menu Underneath the pollutant name shown in the pollutant drop down menu the values for the
45. iculate Matter PM emission maps The EC inventory will help policymakers better determine the extent of traffic pollution in the country s city hotspots as EC is a more suitable indicator of traffic pollution than PM Furthermore the EC emission dataset is being used to produce the first EC air quality assessment maps and forecasts for Belgium INSPIRE Services for Atmospheric Monitoring and eReporting During the ATMOSYS project the in house ICT infrastructure at the Belgian Interregional Environment Agency IRCEL CELINE was remodelled to conform to the system design Figure 68 Expertise page TMOSYS ATMOSYS User Manual Air Quality Modeling System 10 About This page presents some general information on the project To find out more visit the project website www life atmosys be N ATMOSYS About ATMOSYS amp a LFE Environment Potty amp Governance proe co financed by the European Commission Within les project a genen web based service i developed contaimng novel elements specifita Gesgnecd to evaluate and analyse air pollution in European not spot regions The system is established for the hotspot repon Flanders m the North of Belgium but can be deployed in other regions as wel The established system 16 based on advanced techno ay ncluding prognostic 3 D atmospheric computer models results trom recent and on gomg air pollution me ssurement Campaigns in urban agglomerations anc generic OGC com
46. idate recent model forecasting results ideally over a few days automatically on the fly The tool uses the same basic validation statistics as those developed by JRC under the FAIRMODE SG4 working group model benchmarking initiative which led to development of the Delta tool In consultation with JRC an on line version of the desktop Delta tool has also been developed chapter 6 within ATMOSYS The on line model evaluation tool in contrast to this quick check validation page is initially aimed at users who wish to evaluate long term retrospective simulations Validation is done using non validated observations of IRCEL CELINE that are published using a SOS service Page 18 TMOSYS ATMOSYS User Manual Air Quality Modeling System 5 1 Overview Home News ETS Expertise About Contact Services Air quality forecast validation Start validation End validation Pollutant Station type 28 09 2013 7 02 10 2013 T NO2 Alle types Available 26 01 2012 03 10 2013 Show stations Click the Show stations button to show the stations corresponding to the selected parameters Validation 28 09 2013 02 10 2013 Forecast day 0 NO2 Stations Charts Stati Forecast day 0 Forecast day 1 Forecast day 2 _ 7 0 AY m DCR n U industrial 4 Ely 7 r y d r naria a Le Zu lo l j SLE Ud CA flic y 4 r MOMOGOOp Y 4 Ed of tra 4 z m i t 4 f i x gt CN Operatic i gt ET z P T i u
47. ing System 5 7 2 Chart Functionality Both charts have the following functionalities For all selected stations the corresponding observations and model values are added to the time series chart and the scatter plot The period shown in the charts corresponds to the period that is selected in the upper left corner of the page The same color is used in both charts for the same station e The legends of the charts contain the name of the station and make clear whether the corresponding data represent observations or model output By right clicking on a chart and selecting Save Picture As or Save Image As depending on the browser the user can save the chart as PNG and use it in its own reports In the following paragraphs the two chart types are discussed in more detail 5 7 3 Chart type 1 Time Series Stations Charts Statistics Export Time series Air quality forecasts for Belgium Forecast day 0 NO2 15 08 2013 31 08 2013 Concentration g m Time hour Figure 20 Time series chart example An example of a time series chart is shown in 15th August until 31st August 2013 Figure 20 Page 29 E ATMOSYS ATMOSYS User Manual E Air Quality Modeling System e This chart contains both observations points and model output lines for the selected station s Observations and model output that correspond to the same station have the same color The X axis shows the time the Y axis
48. ion results 5 2 Select Validation Period The period that is validated can be changed by changing the start date and end date at the top of the page see Figure 11 By default the last week of the available period is selected Start validation End validation Pollutant 28 03 2013 m 02 10 2013 m NO Available 26 01 2012 03 10 2013 Dates can be changed using the drop down calendar as described in section 4 3 Start validation End validation 17 09 2013 T 24 09 2013 Available 26 01 2012 03 10 2013 Figure 11 Change the validation period 5 3 Select Pollutant The pollutant is selected using the pollutant drop down menu box 5 4 Selection of Measurement Station Types There are four classes of station types given as demonstrated in Figure 12 The user can select all of the station types or they can select only one of the classes Page 20 Ei ATMOSYS ATMOSYS User Manual E Air Quality Modelling System Station type Alle types IT pes Background Industrial Traffic Unknown Figure 12 Change the validation period 5 5 Selection of Station Observation Data Show stations After all the selections click on the button so that all the available measurement stations or for that chosen class for the chosen pollutant the specific forecast day day 0 1 2 and time period will be made available for the validation step In this step a query of the SOS Sensor Observation Service service at the Belgian In
49. is calculated based on population data that we received from VMM This data describes the location of houses and the number of inhabitants in each house for Flanders and Brussels Page 83 ud ATMOSYS ATMOSYS User Manual Air Quality Modelling System Exposure tool Overall average of NO2 ug m from 21 01 2013 00 00 until 25 01 2013 23 00 Population density inhabitants km gt 30000 10000 5000 e pst 1000 gr Y I s Hertogenbosch j Broda 500 Tilburg Te psg coena vrestfalen kim Krefeld TN l F Tx pbsselsort 2 X mM X Sina i 50 Amiens and Strasburg 750973 Sfi sit 6447951 23007 M Figure 62 Population Density map 7 6 Output calculated statistic A second output of the exposure calculation is a map showing the calculated statistic which is computed on the same grid as the air quality service The exposure is calculated for the region for which both population density data and the calculated statistic are available Page 84 ui ATMOSYS ATMOSYS User Manual Air Quality Modelling System Overall average of NO2 ug m Colchester TIN de MWeramerkvWesttalen f x N NT A T4 t 1 Haute Norrmangie J i E ceu S TUM X F f i F yo Paristnt Za 4 MEL Data CCBy Ae ty Der ireetMap LAPS HL f Z R f s UN be aspi g D 427340 5 f608 6281909 70351 f Li 1 P gp Piede France Figure 63 The calculated Statistic in this case a map o
50. l sir Quality Modeling System 10 Exposure tool 10 1 Overview The exposure tab allows the user to calculate human exposure to a chosen pollutant based on specific model simulation results and population data in inhabitants per km for that same region In this prototype demonstration version population data has been provided by the partner VMM for the region of Flanders and Brussels Currently for this demonstration prototype the modelled simulation concentrations are taken from the daily forecast service as described in section 4 However any appropriate database of modelled concentrations could be used The following exposure calculations are provided 1 thetotal number of inhabitants in the region for which the exposure is calculated the average exposure per inhabitant in that region and 3 the percentage of the population in that region that is exposed to a value above the selected threshold value This last value represents a single point in the cumulative distribution plot p On first opening the page only the top panel is visible Here the user selects the parameters to start the calculation After the calculation has finished the results will appear underneath this panel Page 79 E ATMOSYS ATMOSYS User Manual Alr Quality Modelling System ATMOSYS Services Exposure calculation Start time End time Pollutant Exceedings threshold ug m Project ATMOSYS is a LIFE Enuronmoent Pobcy
51. l The second CSV file uses the comme ai Gelimiler and the period as decimal separator When the regional settings of Ihe users machine are in Engiien this file can be automatically opened in Excel Figure 25 The Export tab Two versions of the CSV file are made available The first CSV file uses the Dutch settings which uses the semi colon as delimiter and the comma as decimal separator When the regional settings of the user s machine are in Dutch this file can be automatically opened in Excel The second CSV file uses the English settings which uses the comma as delimiter and the period as decimal separator Page 35 p ATM OSYS ATMOSYS User Manual Air Quality Modeling System 6 Model Evaluation Tool 6 1 Introduction The ATMOSYS model evaluation tool is based on the desktop Delta tool http agm jrc ec europa eu DELTA index htm that is being developed by JRC in the context of the FAIRMODE working group on air quality assessment http fairmode jrc ec europa eu wg1 html The target plot and the summary statistics table developed for the Delta tool for one year of hourly values are included in the ATMOSYS online tool The current ATMOSYS implementation corresponds to version 5 1 of the Delta tool 1 At this moment the tool is configured for the pollutants NO O4 PM and PM s but by changing a configuration file goals criteria oc dat which is the same as for the JRC DELTA tool other pollutants can b
52. ld be clicked The selected files will then be uploaded to the server RE ee EE eraras arta m I01 24351553 Musvustt Excel C DUNS 1S 57 Mertest u Py et 1 4 Aisay fore 4 Tome calected Late meshas ATLL IR Date meded 230022005 12 22 Aw 112 MB Flememe tiada MABI cov MOST en ANT Figure 28 The user selects one or more CSV files The selected files are listed in the Upload tab as shown in Figure 29 Page 37 B ATMOSYS ATMOSYS User Manual ir Quality Modeling System Uptogd Figure 29 The names of the uploaded files are listed alphabetically 6 3 File format The model values and observations should be included in CSV files in which individual fields are separated by semi colons Each file contains hourly values for one year for one station and for one or more pollutants The name of the CSV file is equal to the name of the station followed by the extension csv The model values and observations are included in different files The first line of the file should be a header year month day hour pollutant for one pollutant or year month day hour pollutant1 pollutant2 for multiple pollutants where pollutantX corresponds to one of the pollutant names in the configuration file and should with the current version of the configuration file be one of NO2 O3 PM10 PM25 Page 38 ATMOSYS ATMOSYS User Manual The other lines are in the case of one pollutant have the struct
53. lease contact one of the following persons as shown on the contact page For general queries regarding the project and for specific queries for VITO Lisa Blyth VITO Environmental Modelling Unit lisa blyth vito be For specific queries for IRCEL CELINE Elke Trimpeneers Kunstlaan 10 11 B 1210 Brussels trimpeneers irceline be For specific queries for VMM Edward Roekens Vlaamse Milieumaatschappij Dienst lucht Kronenburgstraat 45 B 2000 Antwerpen e roekens vmm be Page 94 Home hy Mapelbing Sym Contact Please use the contact form if you have any questions or requests conceming our services or if you just want to De informed if new updated services are available Thanks ATMOSYS User Manual News Services Contact VITO Lisa Blyth Boeretang 200 8 2400 Mol Tet 32 14 33 67 57 Emai Contact VMM Edward Roekens Kronenburgstraat 45 B 2000 Antwerpen Bekpum Tel 432 32 44 12 22 Ema Contact IRCEL CELINE Elke Trimpeneers Kunstiaan 10 11 B 1210 Brussel Begum Tel 32 22 27 57 01 Email Figure 70 The contact page Page 95 e ATMOSYS ATMOSYS User Manual E Air Quality Modelling System 12 References 1 DELTA Version 5 0 Concepts User s Guide Diagrams P Thunis C Cuvelier Contributors A Pederzoli E Georgieva D Pernigotti M Marioni Joint Research Centre Ispra February 2015 2 The DELTA tool and Benchmarking Report template Concepts and
54. lete region covered by the selected service These three images can be used by the user to compare the different forecasts in a quick way 4 Forecast day 0 Forecast day 1 Forecast day 2 Under these images the forecast day 0 for the current day is presented in the main viewer The pollutant indicator unit selected date and selected forecast day are shown in the map title the blue header above the map The viewer offers standard GIS functionalities such as zooming and panning The scale of the map is shown in the bottom left hand corner The transparency of the concentration map can be changed by using the slider above Page 13 F ATM OSYS ATMOSYS User Manual sir Quality Miodle System 4 4 1 Legend The legend of the maps is shown at the right of the viewer Figure 6 Unit yg m gt 400 2 0 200 150 110 Figure 6 The map legend The legend shows the meaning of the colors that are used in the maps The units are shown above by the pollutant 4 4 2 Main Map Viewer GIS Operations The geographical data is visualized in a GIS viewer that allows the user to perform typical GIS operations Zooming Zooming is done by selecting the map and then using the mouse wheel to zoom in or out An alternative method is to use the slider at the left of the map window By clicking on the plus or minus sign or by dragging the slider up or down you can zoom in or out Panning Panning can be done by clicking and
55. ly designadto tud s Historical annual maps Data assimilation techniques and analyse air pollution in Europesn Time Evolution maps High resolution and micro air quality hotspot regions The established system is Human exposure to air pollution modelling based on advanced technology including 3D and Model validation tools INSPIRE compatibility atmospheric computer models results from recent air pollution measurement campaigns i urban agglomerations and generic OGC communication protocols to facilitate a fully automated web based service ATMOSYS deliver r quality information from m hes regional vel up to detailed street level ATMOSYS Final Conference r 4 m4 On November 19 20 2013 jointly organize Brussels with a Quality Programmes More info Green Week The ATMOSYS project was presented during Green Week to illustrate the contribution of a LIFE air quafity project to air quality in the EU More info June 1 Services Links Project Daily air quality forecast MyEnvironment VITO ATMOSYS is a LIFE Environment Policy amp Governance Forecast validation Copernicus project co financed Dy tne European Commission The Exposure calculation FAIRMODE project began o Me 16 September 2010 anc willl be Annusl air quality maps MyAir Pssodobis completed at the eno of 2013 Time Evolution maps Timesenes Model evaluation Partners vto Mit CEL Vi ALAMOR BORAT RCRA PRY one of the main results of the ATMOSYS
56. ly slice can be changed by moving along the slicer which is shown under the time series chart Show here under in red RA TH HE LH AUN DRM DNA L i LH in iy LR ON tm M iE LT 401 Lo TE 12 13 14 15 16 Oct17 18 WA PJ Page 75 vor ATMOSYS User Manual l Air Quality Modeling System Polutants ug m WiEC W NO WiO M FMe M PMax Requested Point 51 IAIN 4 49117 f Zoomils zi ie An r i Mas ori 2018 519 pnt0 pmi d 9101 2010 00 04 i re e sf i i aM baal lal 1 ky M Voy M Vy i 23 ca ee LR Lil mi gt LA de CEE MA Figure 57 Timeseries data slice for December 2009 Potutants Guam M EC MNO MO M PMi M Pis Requested Port 5 21055 M 3 40117 E Esport ESS Zoom 15 22 1m 1n n is Maz 1 9 Lj 7 ra em bi Figure 58 Timeseries data slice for October 2009 is now shown by moving along the slicer to the left All the time a month period is visible Page 76 P ATMOSYS ATMOSYS User Manual sir Quality Modelling System 9 5 Changing Viewing of the Pollutants Once the time series is loaded the various pollutants can be selected for viewing or not Thus to visualise only the PM results click off all the rest Now below we see that only the concentrations for PM and PM are visible Pollutants ugm EC NO2 O4 PM PM Requested Point Export i r arid 4 je Bee Sor 15 enw 14 16 foowris 33 Le 2e in 1x uc M Figure 59
57. measurement station then use the CTRL button when clicking on the additional stations The stations are automatically selected in the table underneath the map selected stations in the table are indicated in blue 2 Click on a measurement station in the table The selected row is indicated in blue When you want to select more than one station then use the CTRL button when clicking on the additional stations The stations are automatically selected in the map It is also possible to select a range of stations by clicking on one item and then shift clicking on another stations all stations in between will be selected For example in Figure 17 four stations are selected BETRO20 BETR841 BETM702 and BETN121 Page 24 vor ATMOSYS User Manual Air Quality Modelling System Start validation End validation Pollutant Station type 15 08 2013 nt 31 08 2013 noe pa Alle types Show stations Validation 15 08 2013 31 08 2013 Forecast day 0 NO2 Stations Charts Sues Export Forecast day 0 O Forecast day 1 Forecast day 2 ps amp LA AL 7 ge io A d MSN E bl PES AT nd Ne ehe LE pt More DC 3 cs iy B 7 G e s LA 2d ES Ak PTN an D C LILAS hmc aai gt Monee ts 22 vase Figure 17 Four stations are selected on the map and in the table BETM702 Ertevelde Industry BETN121 Offagne Background BETRO20 Vilvoorde Traffic and BETR841
58. mit value of hourly average concentration of 200 ppm 2009 air quality map NOz hours gt 200 ug m ug m eh D TS 7 aa a TT a T7 LT al m E go Tini PEN 11a 1 M 5 eet Le t a ee Figure 45 Annual Air Quality Maps Page showing above the drop down indicators available for NO2 and below the resulting NO map after choosing to visualize the map showing the number of 3 exceedences of the European limit value of 200 ug m Page 63 P ATMOSYS ATMOSYS User Manual ir Quality Modeling System 7 3 Map Viewer GIS Operations The maps are visualized in a GIS viewer that allows the user to perform typical GIS operations Zooming Zooming is done by selecting the map and then using the mouse wheel to zoom in or out An alternative method is to use the slider at the left of the map window By clicking on the plus or minus sign or by dragging the slider up or down you can zoom in or out Panning Panning can be done by clicking and dragging Scale The scale of the map is shown in the bottom left corner Transparency The transparency of the concentration map can be changed by using the slider in the header bar above the map viewer Figure 46 The transparency slider 7 4 Map Legend The legend of each map is shown at the right of the viewer Figure 6 The unit can be found within 2009 air quality map PM 10 Average ug m the map title together with the pollutant q ty map mS ge ug m 4 uss 4
59. munic ation protoc oss to Tac Mate a fully automated web based service 3D atmospheric computer medei The ATMOSYS system offers the opportunty to expire forecasted as well as li historical data records of air potubon af nexgnbormmood scale by means of animated Air qua ity support system spa maps time senes or Statistics of me nmeevant air pollution indicators ozone yu Polo manado moce li partkutale matter NOZ elementai carbon Regional scale ar p ion simulations are European air pollution hotspots combined wih high resolution street level modets in order to capture all relevant spatial i f Scales of Ihe air podution phenomenon in SION to ihe well estadished Aeg monitoring EU directive compliance network in Beigum dedicated measurement campaigns were setup in vanous o Re ansiyuis service Flemish cihes i order to assess me ar quality differences between urban background urban and street canyon locations Near rea time monitorning Gata and model results ciuding data assimilation techniques are made avaulabie in a web platform by means of INSPIRE compliant ICT technology to offer citizens and policy makers a multitude of Gate about local air pollution m European cities Day forecast maps Onine utser interia Figure 69 About Page Page 93 ATMOSYS ATMOSYS User Manual Alr Quality Modelling System 11 Contact details When you have questions about the ATMOSYS web application that are not answered in this user manual p
60. n Square Emor Mean Fractonai Bias Pearson Coretaton Coemcient Retatve Directve Error ROC CRHMSC 092 0 12 877 poo 0 19 67a Figure 41 Statistics tab containing uploaded data 6 29 Info Page 58 The Info tab Figure 42 contains information about the functionality and use of the model evaluation tool TMOSYS ATMOSYS User Manual Air Quality Modelling System Upkez Targepht Summam votsi c Expat Ecavaldabon into Manual The functionality of the modes evaluation foul is explaned m chapter E Of he user murus which can be found im the help section of Ihe Almmosys toci Desclagmer The content of te Model Evanston section 15 based on the work conducted on the oped ty the JAC in Ihe frame of me FAIAMODE activity The deveopmee of evaluation parameters and graphice representation is 518 com nuing anc should not be considered neither fired nor responding to any requaremen connected to the AQ Orecwe m its present and future state The conten of the section preserted and the model evaluation report are net the model benchmarsiing procedure described m ihe Framework contest and ne model evaluabon repart mareh reports the results of the Model evaluation section Figure 42 Info tab Page 59 ATMOSYS ATMOSYS User Manual Alr Quality Modelling System 7 Annual Air Quality Maps Archive 7 1 Overview The annual air quality maps service is an archive of annual average concentration maps of Particulate Matter PM PM
61. nknown vx A Fe wus Py eter Nordiierrs VVestfaten K a NT Le N Haute Normande st INN y Va y e oy gt x LA DUI Ss tt Zz 3 SHepeF ranice Station Id Station Type Figure 10 Forecast validation page By default the end validation day is the last week of the available period is selected The end date is thus most often today s date 3 October 2013 and the start date 28th September 2013 is 7 days earlier Figure 10 shows the main parts of the validation web page At the top of the page the user can select the validation period and the parameters that he is interested in Page 19 ATMOSYS ATMOSYS User Manual E Air Quality Modeling System In the main window there are four tabs containing a map stations tab charts validation statistics and export functionality for the model output corresponding to the selected parameters At the top of the Stations tab the user can select the day that he wants to validate by selecting one of the radio buttons Forecast day 0 Forecast day 1 Forecast day 2 The map underneath shows all measurement stations that are in the service s region and for which the service provider has computed time series In the Charts tab two charts are included The first chart shows one or more time series of model data and observations The second chart is a scatter plot In the Statistics tab some statistics are shown The Export tab allows the user to export the validat
62. ns This table shows the station ID the station name and the station type Use the scroll bar on the right to view the rest of the stations Station ld Station Name Station Type BETR240 43R240 ENGIS INDUSTRIAL BETR amp 01 42R801 BORGERHOUT TRAFFIC BETNO40 42N040 ST P LEEUWG BACKGROUND BETRS02 45R502 LODELINSART BACKGROUND BETBOO4 41B004 STE CATHERI TRAFFIC BETNOSUO BETNOSO UNKNOWN BELLDOZ 40LD02 LAAKDAL BACKGROUND BETB006 416006 PARL EUROPE BACKGROUND BETRS12 A5R512 MARCHIENNE INDUSTRIAL BETEOOS BETEOOS UNKNOWN BETMEU1 41MEUT MEUDON BACKGROUND Figure 16 The measurement stations selected The 3 headings categories can also be filtered in ascending or descending order Page 23 E ATMOSYS ATMOSYS User Manual E Alr Quality Modelling System aa SE 7 kh a4 k Pr SES Si ae aa 195525 25 ENEE Station Id U Name I Ststion Type BELGKOS M Sort Ascending INDUSTRIAL BELHB22 z Sort Descending BACKGROUND BELLDO1 INDUSTRIAL 2 BELLDO2 f Columns b BACKGROUND BELOBO1 200601 DOS TROZEBER INDUSTRIAL BELSZ01 405201 STEENOKKERZ TRAFFIC BELSZ02 408202 STEENOKKERZ BACKGROUND 5 6 Perform Validation The next step is to choose the measurement stations that you wish to use for the validation This can be done in two ways 1 Click on a measurement station in the map The size of the triangle increases to indicate that the station is selected When you want to select more than one
63. nual Air Quality Maps Page Page 61 TMOSYS ATMOSYS User Manual Air Quality Modelling System 7 2 Load New Maps Choose Pollutant amp Indicator By default the EC annual average concentration map for 2009 is shown in the main viewer As for the previous services the pollutant can be changed via the pollutant drop down list Figure 45 FarMosvs News Services Expertse About Contact At Quality Midsibng Sytari Servic s Annual air quality maps Reanalysis year Pollutant indicator 2009 EC Average NO eae Os PM 2009 air quality map EC Avera ZRN NT These maps are generated using RIO an intermeosation model coupled with FOM D gaussa n Hume mode Excent EC WNO is based on AURONA a chemical iranspoht mode coved with fF DM Figure 44 Annual Air Quality Maps Page Pollutant Options Likewise the required specific statistical indicator can be chosen via the indicator drop down menu see Figure 45 To load the new map in the main viewer click LOAD Under each indicator an explanation is provided For example AOT60ppb max8u is the cumulative sum of the daily maximum 8 hour average concentration exceeding 60 ppb Page 62 ATMOSYS User Manual Average hours y gm P99 79 2009 air quality map NOz Average ug m Reanalysis year Pollutant indicator 200 s no hours gt 200 yom el Number of exceedances of European li
64. odel observation values Page 30 E ATMOSYS ATMOSYS User Manual Alr Quality Modelling System e The correlation coefficient R is also shown in the chart and in the legend e The title of the chart shows the name of the service the forecast day the pollutant and the validation period 5 7 5 Chart Values At the end of page there is the option to change the averaging period By default the hourly values are for the selected period are shown However if you wish to validate the daily average daily maximum values etc this is done using the drop down menu as show in Figure 22 Hourly value Compute Daily average Daily maximum Daily maximum of eight hour averages 24 values for each hour of the day f values for each day of the week 2 values for day and night Figure 22 Drop down menu of the different averaging periods avialable Compute When you press the button the charts for that new averaging are computed In Figure 22 the daily average values are given for the chosen period above are given Page 31 ATMOSYS Air Quality Modeling System ATMOSYS User Manual Start validation 15 08 2013 Avallahis 29 21 2012 25 10 2213 End validation Pollutant 31 08 2013 2 NO2 Station type Alle types m Show stations Validation 15 08 2013 21 08 2013 Forecast day 0 NO2 Stations Charts Statistics Export Time series Air quality forecasts for Belgium Forecast day 0 NO2
65. odeling System services Air quality forecasts for Belgium Forecast base time Pollutant pq m 03 10 2013 b NO2 Available 26 01 2012 03 10 2013 Daily maximum mE Forecast day 0 z Forecast day 1 Forecast 03 10 2013 Forecast day 2 NO2 Daily maximum pM Fs y A 7l re 2 DLL Bruxedies Brussel TN a Figure 9 Visualization of forecast day 2 5 October Page 16 Moechuenooonaes t n MOSYS ATMOSYS User Manual ATMOSYS ATMOSYS User Manual Alr Quality Modelling System 4 5 Changing the Pollutant By default the NO2 daily maximum map for the most recent forecast day available normally today is shown in the main viewer To visualize another pollutant concentration map click on the pollutant drop down menu box to view the other pollutants available For ozone it is possible to view the 1h maximum or the 8h maximum concentration maps Pollutant pg m E EC O3 1h max O3 8h max PM25 Forecast day 2 When you have chosen the new pollutant click on the load button to upload the corresponding maps The legend shown in the main viewer will change depending on the pollutant being shown Page 17 ATMOSYS ATMOSYS User Manual Alr Quality Modeling System 5 Forecast Validation The validation tab allows a user to validate the model output of a chosen ATMOSYS air quality forecast using measurement data It is developed to allow a user to val
66. on Tool 6 1 6 2 6 3 6 4 6 5 6 6 Introduction Upload model values and observations File format Check the format of the uploaded files Remove uploaded file Remove all uploaded files Page 2 ATMOSYS User Manual CON N OU 10 10 12 12 13 14 14 17 18 19 20 20 20 21 23 23 23 24 26 28 29 29 30 31 33 35 36 36 36 38 39 40 41 ATM OSYS ATMOSYS User Manual me Air Quality Modelling Syst 6 7 Remove all model value files 41 6 8 Remove all observation files 41 6 9 Target plot tab 42 6 10 Parameters Select amp Validate Parameters 42 6 11 Compute Target Plot 43 6 12 Stations summary 44 6 13 Time transformation 44 6 14 Minimum data availability 44 6 15 Target plot 45 6 16 Summary statistics 47 6 17 OBS Mean 48 6 18 OBS Exceed 49 6 19 TIME Bias Norm 49 6 20 TIME Corr Norm 49 6 21 TIME StdDev Norm 49 6 22 TIME HPerc Norm 50 6 23 SPACE Corr Norm 50 6 24 SPACE StDev Norm 50 6 25 Formulas 50 6 26 Configuration file goals criteria oc dat 51 6 27 Export 51 6 28 Extra validation Forecast Validation Tool Functionality 55 6 29 Info 58 7 Annual Air Quality MapsArchve oo 60 7 1 Overview 60 7 2 Load New Maps Choose Pollutant amp Indicator 62 7 3 Map Viewer GIS Operations 64 7 4 Map Legend 64 8 Time Evolution Maps 65 8 1 Overview 65 8 2 Loading New Maps 67 8 3 Loading map data 67 8 4 Viewing the Animation 69 8 5 Share the Animation Video 71 9 Time Series 72 9 1 Overview 7
67. ot fulfils tw Performance Criteria e TIME 30 of stations fulfils he Performance Critena SPACE Dot does not fut fe Performance Criteria Figure 34 Summary statistics table 6 17 OBS Mean For each station the mean observation is computed N _ 1 O 0 i 1 The resulting means for each of the stations are presented as blue circles Page 48 m ATM OSYS ATMOSYS User Manual 6 18 OBS Exceed For each station the number of exceedances of a limit value is computed The limit value is dependent on the selected pollutant and is read from the configuration file It is shown in the second column 6 19 TIME Bias Norm For each station the normalized mean bias NMB is computed in the following way M O0 2 x RMSU The limit values for the normalized mean bias are 1 NMB lt 1 6 20 TIME Corr Norm For each station the normalized correlation is computed in the following way 1 R U 2 RMS CorrNorm The limit values for the normalized correlation are 0 CorrNorm 1 6 21 TIME StdDev Norm For each station the normalized standard deviation is computed in the following way OM 00 StDevN evNorm 2 RMSU The limit values for the normalized standard deviation are 1 and 1 1 x StDevNorm lt 1 Page 49 ATM OSYS ATMOSYS User Manual Air Quality Modeling System 6 22 TIME HPerc Norm To provide insight on the model capability to reproduce
68. r SHIFT click All statistics can be copied at once by using the Copy all button Copied statistics will appear in the list at the right Copied statistics can be deselected i e moved to the list at the left in a similar way by using the Remove buttons or by double clicking in the list at the right Page 33 B ATMOSYS ATMOSYS User Manual Air Quality Modelling System When clicking on the Compute button a table is shown containing the selected statistics columns with values for each selected station rows Two of the statistics Relative Directive Error and Relative Percentile Error make use of a limit value and the maximum number of exceedances of a certain limit value Since these values change in time and since different organizations are interested in different limit values these values are specified as input values This information is available in the web application by clicking on the question mark 5 ATMOSYS Services Air quality forecast validation Start validation End validation Pollutant Station type 15 08 2013 31 08 2013 NO Alle types Show stations Validation 15 08 2013 31 06 2013 Forecast day 0 NO2 Stations Charts Statistics Export y ee w e an C5 das a Foot Mesan Squere Ever epon Conelssmon CosMoert een Biss Sais Cf pytlemehs 28 Utsystemane RASE z Starca Qeviption moselied valve Felative Coe Enc S a 2a73 Cevistion chartres values Beiptive
69. re taken from a test service established for the Belgian Interregional Cell for the Environment By default the NO2 daily maximum map for the most recent forecast day available normally today is shown in the main viewer as shown in Figure 2 Above the main viewer the maps for the three consecutive forecast days day 0 day 1 and day 2 are shown Page 10 MOSYS ATMOSYS User Manual Air Quality Modeling System ow ATMOSYS Home News Expertise About Contact fur Qualty Modetkng System Services Air quality forecasts for Belgium Forecast base time Pollutant pq m 03 10 2013 2 NO2 Available 26 01 2012 03 10 2013 Daily maximum Forecast day 1 Forecast day 2 Forecast day 0 Forecast 03 10 2013 Forecast day 0 NO2 Daily maximum 3 3 D 7 7 Figure 4 Visualization of the forecast service for Belgium forecast day 0 Page 11 ATMOSYS ATMOSYS User Manual ES Air Quality Modelling System 4 2 Parameters At the top part of the page the current selected forecast date the period for which data is available the chosen pollutant the indicator and the unit are shown Forecast base time Pollutant ug m 03 10 2013 NO2 m 02 ar 0 Avadable 26 01 2012 03 10 21 Daily f 4 3 Changing the forecast date By clicking on the current date or on the icon to the right of it Forecast base time 03 10 2013 Available
70. scalier pois and statistics are computed for the upioaded data Available period Start validation End validation Pollutant Dpt 01 2009 31 12 2009 24 12 2000 31 12 2008 INO2 _ Start vasdaton Hourly values Charts ststetce Time series NO2 24 12 2009 31 12 2009 110 100 4 90 80 70 60 50 ey 40 Concentration pym 30 20 10 24 12 2009 Oh 26 12 2009 Oh 28 12 2009 Oh 30 12 2009 Oh 01 01 201 Time hour Model values ECYDLSEsSLESSSUISSSLSSE p Observations Page 57 ATMOSYS User Manual Figure 40 Extra Validation tab after the validation Extra veidtation By chctang ihe San vasadabon billion addons chars ame Series and scafie piots Available period 01 01 2009 Start validation End validation Start validabon Charts Statishes Fraction of mocelied values within a tactor of iwp of observator Acid ail index of Agreement i Mean Bias Mean Fractional Error Mean modeted value Mean observed value Modet Efioency Score Ratio of systematic and unsystematic RMSE Raliatus Parrontiio Error T it Remove all Limit value po m 450 Number of exceedances 4 Compute turopean code 40HB23 40HB25 216004 416004 an 1321201935405 are compused for Tie uptoade c data Pollutant NOJ Howy value Centred Root Mea
71. ser Manual Alr Quality Modelling System Figure 35 Export tab When clicking on the PDF download link a dialog is opened which allows the user to save the PDF file or to open it using an appropriate program Figure 36 Figure 37 and Figure 38 show an example of such a PDF file It contains an overview of the selected parameters the stations summary target plot and the summary statistics table VITO 18 03 2015 15 31 Page 1 of 3 Model evaluation report 1 Parameters Start validation 01 01 2009 00 00 End validation 31 12 2009 23 00 Pollutant NO2 Indicator Hourly values 2 Stations summary OK 40HB23 41B004 Missing model value or observation file insufficient tuples lt 75 in selected period 40ABO1 40AB02 Figure 36 Exported PDF file page 1 Page 52 ATMOSYS User Manual VITO 18 03 2015 15 31 Page 2 of 3 3 Target plot Target plot Stations within crit T 1 100 NO2 Hourly values 01 01 2009 31 12 2009 BIAS 20 45 1 0 05 00 0 5 1 0 1 5 2 0 CRMSE E 40HB23 41B004 Figure 37 Exported PDF file page 2 Page 53 MOSYS ATMOSYS User Manual me Air Quality Modelling Sysbem VITO 18 03 2015 15 31 Page 3 of 3 4 Summary statistics NO2 Hourly values 01 01 2009 31 12 2009 Nb of stations groups 2 valid 4 selected Figure 38 Exported Pdf file Page 3 Page 54 ATMOSYS ATMOSYS User Manual Alr Quality Modelling System A CSV file is also available
72. terregional Agency is undertaken A SOS service is a web service interface which allows querying observations sensor metadata as well as representations of observed features In the SOS for each of these measurement stations hourly concentration values are available for all days in the available period and for each of the three forecast days This enables the user to carry out validations for each of the forecast days and for each of the days in the available period Figure 13 Retrieving the data from the SOS Validate Show stations is completed Once this query is complete the message appears in place of the show stations button In addition the available stations are now visible both in the map viewer see Figure 13 and in the table underneath E Page 21 Pervosrs Parwoss Services Air quality forecast validation ERE POOO E ES Sac L 15082033 O aean Im we ou Alle types ne Avallable 29 07 2072 03 12 2015 a OS Validation 15 08 2012 31 08 2013 Forecast day 0 NO2 Forecast day i Forecast day 2 4 I 2 m A D a ke uw ht r E En NT NL UL Xe LE ac SR GE DONC SEA AVS Y mes Een Name Savon Tyee BETRZ4C 43RZ4C ENGS SETRI 429801 BORGERHOUT TRAFEIC BETNOAO 42ND40 ST P EEUNG BACKGROUND BETRSOZ 45R502 LODEUNSART BACKGROUND BETECCA 41BCC4 STE CATHER TRAFE SETNOS0 BETNOSO UNKNOWN BETECOS 418008 PAR EUROPE BACKGROUND SETRI 458512 MARCH
73. thout requiring any recoding of the tool parameter values used in the analysis are displayed on the target plot the formulation of some of the model performance criteria presented in the summary Statistics table was changed the summary statistics table only considers the case where the model performance criterion is fulfilled green red circle the intermediate case orange circle is not considered any more the summary statistics table now also indicates which error type is predominant the summary statistics table no longer contains a value for the relative directive error RDE a performance criterion for percentile values was added to the summary Statistics table to allow assessment of the models capability to predict extreme events when exceedances of limit values occur Page 98
74. tistics uses different values that are different for each pollutant These values are read from the configuration file goals criteria oc dat This file is also used by the JRC Delta tool As the development of the benchmarking procedure within the FAIRMODE community is an ongoing activity the current values in this configuration file could still change It was therefore decided to also base the ATMOSYS benchmarking tool on this same configuration file to ensure that changes to the parameters in the JRC Delta tool can easily be adopted in ATMOSYS A description of this file can be found in the Delta tool manual 1 The current values in the file can be found in the table below NO2 hourly value 200 24 004 IT ENT EDITT 0 018 PM25 dac us 0 035 O3 daily maximum 8 hour 120 12 6 0 62 average 6 27 Export The export tab Figure 35 allows the user to download a PDF file or CSV comma separated values file that contains the results of the model evaluation tool including the target plot and the summary statistics Upload Targetplot Summary statistics Export Extra validation Info Export Click the following link to download the model evaluation report as PDF Download report as PDF Click the following link to download the model evaluation report as CSV Download report as CSV Click the following link to download the summary statistics as PNG Download summary statistics as PNG Page 51 ATMOSYS ATMOSYS U
75. um are produced using RIO an interpolation model coupled with IFDM a bi gaussian plume model Information on the developments undertaken in ATMOSYS and the full methodology used to generate the best possible annual air quality maps for Belgium are given here 4 City and Highway Measurements To gain a better insight into how the concentrations of PM and NO2 vary within a city hotspot between several city hotpots and nearby a busy highway an intensive city and highway campaign was carried out in Belgium in the period July 2011 until February 2013 This had led to a comprehensive measurement database for cities in Flanders which provides a valuable insight into the dynamics of these pollutants across the cities It has also been used to validate the high resolution IFDM model used to produce the ATMOSYS hourly and annual air quality maps for Belgium Aside from the measurement campaigns one of the tasks within ATMOSYS is to identify which of the permanent VMM measurement network stations should be actually used for validation of the AURORA model and for use in the Data Assimilation In order to do so the Spatial representativity Page 90 F ATM OSYS ATMOSYS User Manual sir Quality Mocellng System of each of VMM s monitoring stations had to be determined This was done by applying the work of Spangl et al to derive a method to assess the spatial representativity of VMM s monitoring stations More information can be found here 5 I
76. ure 2007 01 01 00 123 2 2007 01 01 01 243 7 2007 12 31 23 25 8 where the fields separated by semi colons from left to right are e year 4 digits e month 1 12 day 1 31 e hour 0 23 Leap years are not taken into account so a file always contains 8760 365 24 values Some other rules If data are missing the gaps should be filled by 999 Each blank row or rows beginning with or will be discarded No spaces are permitted between the fields Decimal separator is An example CSV file can be downloaded from the Help tab see section 8 6 4 Check the format of the uploaded files The format of the uploaded files can be checked by clicking the Check format button Upload Info Check format Remove all Remove model value files Remove observation files Depending on the number of files to check checking the uploaded files can take up to several minutes During upload the hour glass appears When all files are checked and valid then a green message is shown above the upload component indicating that the validation is ok Page 39 ATMOSYS ATMOSYS User Manual Air Quality Modelling System Upioad Target plot Summary statisecs Export Extra vabdation Inte heck orma Remove a Remoye model value nias Remove cocervation ies Model valves Obseryations aiudation ok Validation on Uiptoad file Upload fe Uploaded model value files Uploaded observahon fies 404801 csv DABO I csy Remove
77. vailability As described above at least 75 of the tuples in the selected validation period should be effective Otherwise the target plot and summary statistics cannot be computed for a station Page 44 ATM OSYS ATMOSYS User Manual Air Quality Modelling Syste This 75 limit is also applied when computing the time transformations E g when computing a daily average at least 18 of the 24 hourly values should be effective If not the daily average cannot be computed for that day When computing an eight hour moving average at least 6 of the 8 hourly values should be effective otherwise the moving 8 hour average for that hour cannot be computed 6 15 Target plot The target plot Figure 33 shows two statistics for each station Each station is represented by its own color and or shape in the plot which is shown in the legend below the target plot The x value is given by where The value of CRMSE is always positive The following sign function determines whether the x value is positive or negative When 09 O o ol 2090m 1 R i then the sign is 1 Otherwise the sign is 1 The formula for the Pearson correlation coefficient R for the standard deviations and for RMSU can be found in section 6 25 The y value is given by MBias 2 RMSU where N 1 MBias M E Oi i 1 Page 45 ATMOSYS ATMOSYS User Manual Alr Quality Modelling System The target plot shows two circles

ATMOSYS User Manual

Contents

Download Pdf Manuals

Related Search

Related Contents