DBS Tailoring System
Contents
1. SICA SHAPEFILE HICK RASTER gt NETCDF gt b DODS IMAGE figure 16 The Export task tab Here you can enter the shapefile and data source files being used into the source and extra data source respectively After clicking on the target box you need to choose what type of file you are exporting the data to at present only shapefiles and ASCII files are supported The Export task requires two files 1 a shapefile captaining information on the data to export and 2 the source data which can be either a netCDF file or an ASCII file both time and id orientated files work In this version of the program it is possible to export data to two different file types namely shapefiles and ASCII files The ASCII files can be further divided into three subgroups time orientated ASCII files ID orientated ASCII files and discrete ASCII files Click on the task to open the task tab figure 16 The source data is the shapefile that will be used in the export Job click in the box and navigate the dropdown list to select the shapefile that you intend to use you need to have loaded the information into the system prior using a load task The extra data source 1s the file from which the data will be exported again you need to have loaded the information into the system prior using a load task 161Page A window will appear after you have selected the source file and the extra data source2. respectively asking you to define the data ranges figure 17 Here you can limit select the ranges of the data being used in the export task either selecting the ranges directly or as in the case for time related data you can set the start and end dates using these buttons in the ranges section You can also perform simple data manipulations i e unit conversions with the help of the data multiple or data add Additionally there is a box convert to normal calendar that you can select when exporting data from netCDF files that do not use a normal calendar If you do not select this box and the data you are exporting has a non normal calendar the exported data will retain the non normal calendar attributes Define data item Data name pr Output eks Ranges Data Time of 5 229 Data multiple 1 0 Data add 0 0 Significant digits 1 H 1 turns off Nr decimals 1 H 1 turns off L convert to normal calendar Data modifier from other resource dick to add figure 17 After selecting each of the files for the source and the extra data source you will be asked to define the data item Here you can limit select the ranges of the data being used in the export task You can also perform simple data manipulations i e unit conversions with the help of the data multiple or data add There is also a box convert to normal calendar which you can select i
3. Climate Change and Environmental Objectives 2 DBS Tailoring System a reference manual of the software version 1 0 Kean Foster Swedish Meteorological and Hydrological Institute HI2 0_D342_DBS_Tailoring_System_a_reference_manual_ version_1 Senast ndrad 2012 07 03 2lPage Contents CONTENTS 2 i aaa 3 THE DISTRIBUTED BASED SCALING METHOD 4 1 INTRODUCTION ni se se ennenen 4 2 DA OB EE sean eee cies ceeie ete cece soe cee utc nieces cece eee eee 4 2 1 The DBS Tailoring system 4 2 2 ag S21 01 162 NOM anim inondation ni 5 2 3 FTemperatUr Cras aac aaa oan aon asec lle lle lei den 5 3 REQUIREMENTS wicacicsecestasscccetecacecacseacsvatavaccsatanbesvedavadevetevstevstaescawetacads T 3 1 System T QUIF M NES scr Men i E AE EEEE 7 3 2 Programi esiliati 7 3 3 bala requiremenisia dillo 8 4 THE USER INTERFACE anal areali aeree 9 4 1 DAN Tes ee ee AEE ee ee ii ia 10 4 2 DAV C10 ere 10 4 3 Ea E A E E A A E E A A T 10 4 4 Cedric 10 4 5 Add sigilli iaia 10 4 5 1 LOAD 11 4 5 2 PLOT 14 4 5 3 EXPORT 16 4 5 4 DBS 18 4 5 5 FLIPPER 22 5 EXAMPCES cadet needa ace ede beat ee ca because gece vudabaca veda bacaeedsaece ous eeeaeentebeinee tales 23 5 1 Typical DBS job process diagrams 23 6 2 samplexmi lesene 25 Acknowledgements The development of the DBS tailoring system was co funded by the national research programs CLEO CLimate change and Environmental Objectives
4. 11 11 7 49 10 42 22100 12 38 11 89 10 42 9 14 19810104 6 69 9 69 6 47 9 14 figure 3 A sample of typical time orientated data left and the same sample data in id orientation The files are tab delaminated ASCII files and can be created using any suitable software e g Excel if you need to create them manually from data sources not supported by this system 8lPage 4 The User Interface The user interface allows users to create DBS jobs by building and editing XML files that run the system When you first open the DBS tailoring system interface you will see a blank workspace figure 4 if you have not loaded an existing xml file sno last save gt rta amp lt no last savez figure 4 The blank user interface left The action wheel can be opened by right clicking in the work space right Here you can build and edit tailoring jobs with the help of the action wheel to access the action wheel right click anywhere on the empty workspace The action wheel has five alternatives Save saves the xml file you are working on to you work folder If you have not previously saved the xml file before it will open the Save To option instead Save To opens a standard window for saving files where you can enter change the name of the xml file and the location Load for loading existing xml files into the system Clear clears the workspace Add opens the new task window By right clicking on the task
5. chosen reference period and the reference precipitation data again 5 Calculate the statistical parameters related to the simulated temperature This step requires an ASCII file containing simulated temperature for a chosen reference period and the results from the scaling of the precipitation data 6 The scaling of the simulated precipitation This step requires an ASCII file containing modelled temperature for the entire period of interest and the following files the parameters from steps 2 4 and 5 and the results from the scaling of the precipitation data 4lPage The DBS tailoring system helps to create the required files for the different steps described above and controls the DBS motor in each step The DBS tailoring system uses XML files to define how to perform these tasks These XML files are constructed with the help of the user interface see sections 4 and 5 for more details 2 2 Precipitation In the case of precipitation the DBS approach uses two steps 1 spurious drizzle generated by the climate model is removed to obtain the correct percentage of wet days and 2 the remaining precipitation is transformed to match the observed frequency distribution To obtain the percentage of wet days correctly a threshold is identified for each sub basin and season Days with precipitation amount larger than the threshold value were considered as wet days and all other days as dry days There are various theoretical distributions avai
6. file open in the interface you will be asked if you want to add the xml file to the end of open group If you choose yes the tasks in the xml file that you are loading will be added to those already in the user interface to create one long group If you choose no the tasks in the xml file that you are loading will be added as a new group after those already in the user interface to create two groups of tasks The difference between one or more groups 1s that the tasks in each group are independent of those in the other groups This means that if a Job crashes in a critical task in one group the system will skip to the next group and continue running If you only have one long group the entire job will crash if a critical task crashes 4 4 Clear This option on the action wheel allows you to clear the user interface of all tasks instead of having to delete the individual tasks one at a time 4 5 Add This option on the action wheel allows you to add the different individual tasks and in that way build up an xml file The individual tasks used to build an xml file are load plot export DBS and flipper figure 6 10 Page Add new task Choose task type to add Loan por EXPORT FLIPPER figure 6 New task window for adding one of the five base tasks to the workspace 4 5 1 LOAD This base task allows you to load file information for different file types into the system Left clicking
7. files created in the first two DBS tasks to scale the simulated precipitation dataset figure 19 If you are using DBS engines versions 1 2 or 3 it is recommended to perform this task twice once to scale the simulated precipitation for the reference period and a second time for the entire simulated dataset This is because scaled precipitation data for the reference period is needed in DBS step 5 and it is not yet clear if the DBS engine selects the correct period from the full data set Darin i Lele Simulated precipitation dataset E click to edit Parameters For reference Precipitation click to edit Parameters for Simulated Precipitation click to edit optional Save parameter difference scale strength to click to edit Save results to click to edit figure 19 The third DBS task scaling of the simulated precipitation dataset using the parameters calculated in DBS tasks 1 and 2 Here you need to reenter where the DBS engine and work folder are located the reference period and redefine the seasons if needed done in previous BDS tasks 191Page As with DBS tasks 1 and 2 all the information needs to be re entered for the DBS engine directory seasons etc only this time you need to enter the start and end date for the scaling 1 e the reference period for the first run and the entire simulation period for the second run You then need to select the necessary files for t
8. of the Swedish Environmental Protection Agency and HY DROIMPACTS2 0 Hydrological climate change impact scenarios developing the tools for a new generation by the Swedish Research Council Formas 3lPage The Distributed based scaling method 1 Introduction Today there is a huge demand for knowledge on climate change and the impact on our environment One source of information on possible outcomes of climate change is the numerical global circulation models GCMs The GCMs model the climate for the entire globe from the past into the future Different assumptions on how the greenhouse gas emotions will evolve can thus be tested within this framework However the scale of the information obtained from those models are often to coarse for any impact study on a regional scale To downscale the information from the GCMs either statistical methods or regional climate models RCM are used Both methods have their advantages and disadvantages The RCMs hold more physics and might give a better representation of the geographical variation The statistical methods on the other hand are more straightforward and don t require the same amount of computational time as the RCMs To make the regional information useful in hydrological impact studies the information has to be even further downscaled and bias corrected The DBS Tailoring System presented in this reference manual is an interface to a statistical method to do this downscaling and bias correction
9. required files need to entered It is not possible to save the scaling strengths for the temperature scaling in this version You need to enter in the required files for the scaling the simulated temperature data set parameter files for the simulated and the reference temperature datasets the parameters for the simulated precipitation dataset and the scaled simulated precipitation data Lastly you need to enter the path and filename for the scaled simulated temperature This version of the DBS tailoring system does not support saving the scaling strengths for temperature scaling 21 Page 4 5 5 FLIPPER This task allows you to change the orientation of the ASCII files The DBS engine requires that all ASCII files are time orientated that is to say they need to have the dates on the x axis and the IDs on the y axis HYPE on the other hand requires that the files are ID orientated There are two ways to transpose the ASCII files e You can use an export task and export the ASCII file as another ASCII file but with the required orientation e You can use the flipper task to transpose the data in the ASCII file being flipped Using the flipper task is much simpler all you need to do is enter the file to be flipped and set the path and name of the flipped or transposed file figure 22 figure 22 The flipper task transposes ASCII files from ID orientated to time orientated and vice versa Enter the input file the file
10. year but does not have leap years 12lPage Data of variable time lt 30 read around center of dimensions but always first and last 2923 0 0 29023 0 26099 29024 0 26100 29025 0 26101 Attributes of variable tinme 1 units days since 1950 01 01 00 00 00 2 standard name time 29026 0 26102 29027 0 26103 29028 0 26104 3 long_name time 29029 0 26105 4 calendar standard 29030 0 26106 5 bounds time_bnds 29031 0 26107 figure 9 Right clicking on a variable and choosing Show attributes opens a window with a summary of the netCDF variable s attributes left Similarly by choosing Show data opens a window that shows a sample of the variable data right Information about the start date for the data and the calendar type can be seen in the Variable attributes window figure 10 and calculated from information found using the and Show data function described above figure 8 and figure 9 Although this information is optional it is advisable to enter it otherwise all data will have relative time step labels starting at 1 1 e 1 2 3 etc as opposed to dates e g 19610101 19610102 19610103 etc NOTE it is important to check that the calendar setting corresponds to the calendar used in the netCDF to avoid temporal shifts You can only proceed when all the relevant fields are set 1 e the ok button will only become active when this step is complete C
11. TA contains three selectable options Discrete Timeserie and Idserie Select the option that is relevant to the data you are loading information from Choose the option that is relevant to the file you want to load information for this will open a normal windows load window where you can choose the ASCII file you want to load information for When you have done this another window will open figure 12 Here you need to confirm the type of ASCII information the file contains data series or discrete values and the orientation of the file whether it is Id oriented or time oriented choose Id for discrete data Further down there are two fields where you enter information regarding headings in the file Header at row refers to which row the header is on normally this is row 1 and the second Skip first rows refers to how many rows need to be skipped to reach the data again this is usually 1 At this point there is no need to make any changes to the last three options These are used to import information about geographical data that may be saved in an ASCII file Here you would enter which columns contain the id information latitude information and longitude information Define ASCII data file CI contains data series contains discrete values has geography gt HBV time serie gt contains DBS parameters or orientation Id Time Comments regex W Deliminator r
12. e DBS engine and work folder seasons dates etc needs to be re entered Then you need to enter in the reference temperature and precipitation datasets into DBS task 4 so that the parameters for the reference temperature dataset can be calculated For DBS task 5 you need to enter the simulated temperature dataset and the results from the scaling of the simulated precipitation dataset for the reference period as discussed above in DBS task 3 And again enter in the path and filenames for the parameter files to be calculated 4 5 4 4 DBS task 6 DBS task 6 1s the last of the DBS tasks figure 21 scaling the temperature dataset Once again the initial data in the header needs to be entered as per the previous tasks except that the period is now just the period of the simulated data There are more files to enter as the temperature scaling is dependant information from the precipitation files Simulated Temperature dataset L click to ed Parameters For simulated Temperature click to edit Parameters For reference Temperature click to edit Parameters for simulated Precipitation click to edit Result from precipitation scaling Click to edit optional Save parameter difference scale strength to click to edit Save results to figure 21 DBS task 6 scaling the temperature dataset The initial information needs to be re entered in the header the period is the entire simulation period Then all the
13. eed to enter where the DBS engine and work folder are located and define the reference period You can also change how the seasons are defined here The seasons are each three months as default you can change the number and length of the seasons by selecting the season number and then clicking on the month you want in that season You need to enter the period for this task That is the date when the DBS engine needs to start calculating and the end date These dates are entered in a popup window with the following format 18lPage YYYY MM DD As the first task is calculating the parameters for the reference precipitation dataset the time period you need to enter is the reference period being used period of the reference data The relevant datasets are entered into the corresponding boxes by navigating the dropdown menus to the relevant files you have to have loaded this file info into the system before you can select 1t In DBS task 1 you need to enter the reference precipitation dataset in DBS task 2 you need to enter the simulated precipitation dataset for the reference period and the parameters for the reference precipitation dataset calculated in DBS task 1 The name and location of the parameter files are entered in the save boxes furthest down the task tab Note that every time you add a new DBS task you need to re enter all the search paths dates and seasons etc 4 5 4 2 DBS task 3 The third DBS task uses the two parameter
14. egex its J Header at now oH Skip first rows of ld column 0 Lon column 1 H Lat column 1 H figure 12 After selecting the ASCII file for which to load information about this window appears in which you need to select the type of data the file contains data series or discrete values indicate the orientation select Id for discrete data which row the header is on and how many rows to skip to reach the data 4 5 2 PLOT This task allows you to create images of shapefiles Load in the information for the shapefile you want to plot as described in Loading shapefiles above Once you have added the plot task left click on the task to open the task tab figure 13 Source refers to the shapefile to be plotted click 14lPage in the source box select task output and choose the shapefile from the dropdown menu as shown in figure 13 PLOT task 2 1 eport Criticaltask figure 13 The Plot task tab to choose the shapefile you wish to plot from navigate the dropdown lists and click on the desired shapefile A popup menu will appear listing the different variables in the shapefile figure 14 here you can choose lt all items gt to create plots for all variables or choose a specific variable from the dropdown menu to create a plot for just that variable figure 14 A popup with a dropdown menu with the different variables that can be chosen Here you can choose lt all items gt if y
15. f you are exporting data from a netcdf file that has a non normal calendar and wish to have the data converted to a normal calendar 17 Page 4 5 4 DBS This task prepares the info DBS _ generator txt file which controls the DBS engine DBS_generator exe which performs the scaling To perform a scaling job you will need to set up six different DBS tasks three for each variable The first two tasks prepare parameter files for the observed and simulated datasets respectively and the third uses these parameter files to perform the scaling 4 5 4 1 DBS tasks 1 and 2 The first two DBS tasks calculate the parameters for the reference and simulated precipitation dataset to be used in the scaling of the scenario data later The DBS engine must be saved in a work folder where all the necessary files for the scaling tasks are will be saved In figure 18 below you have the first two of the DBS tasks where the parameters for the reference precipitation and the simulated datasets are calculated You need to enter where the DBS engine and the work folder are located as mentioned above these should be in the same folder click to edit click to edit Parameters for reference Precipitation click to edit Save parameters for Precipitation to click to edit l clickto edit figure 18 The first two DBS tasks perparing the parameter files for the reference observed and the simulated data sets Here you n
16. he scaling process the simulated precipitation dataset first a file for only the reference period and then the file for the entire period the parameters for the reference precipitation dataset and the parameters for the simulated precipitation dataset You can choose to save the scaling strengths a measure of how much the DBS scaling has affected each data point which can be useful for analysing the scaling results To save the scaling strengths you must click on the box for saving the scaling strengths and select a path and file name Lastly select the path and file name for the scaled precipitation data 4 5 4 3 DBS tasks 4 and 5 DBS tasks 4 and 5 are similar to DBS tasks 1 and 2 in that they prepare the parameter files for the reference and simulated temperature datasets They differ though in that they use files related to precipitation in the process figure 20 click to edit Lelick to edit Reference Precipitation dataset Result from precipitation scaling Save parameters for reference Temperature to click to edit figure 20 DBS tasks 4 and 5 are similar to DBS tasks 1 and 2 only that they need files related to precipitation in the process Again all the initial information needs to be re entered as per DBS tasks 1 2 and 3 then the relevant files for calculating the paramete files are entered as well as the path and names of the parameter files being calculated 20 Page The initial data the path to th
17. ify Netcdf contents Rieht click on on item to set a field dota field must be set Eime field is optiondi rotated pole rion Mon 0 84 LON ton frfat 0 94 rion 0 84 rat riat 0 94 LAT tat riat 0 94 rion 0 84 time imeks time_bnds time Longitude field DATA pr time 0 4 Latitude field Data field Show attributes Show data figure 8 After selecting a netCDF file you need to set the required fields longitude latitude the data field and time This is done by right clicking on the variable and setting it to the appropriate field Selecting Show attributes or Show data options displays the netCDF variable attributes and a sample of the data respectively figure 9 Selecting either Show attributes or Show data options from the menu as seen in figure 8 opens a window with the attributes of the netCDF variable and a sample of the variable data respectively figure 9 This can be very useful for determining the calendar type and variable units which are important in later steps When you right click on time another window will appear here you will be asked to input optional time related information Here it is possible to set the start date for the netCDF data and change the calendar type used in the netCDF figure 10 The default calendar setting is NORMAL The option ONLY360 refers to a calendar that uses a 360 day year while NOLEAPS refers to calendar that uses a 365 day
18. lable to describe the probability distribution function PDF of precipitation intensities A commonly used distribution is the gamma distribution because of its ability to represent the typically asymmetrical and positively skewed distribution of daily precipitation intensities Wilks 1995 Haylock et al 2006 Yang et al 2008 A single gamma distribution was therefore considered as the first choice in the DBS method but then expanded as discussed below The gamma distribution is a two parameter distribution whose density distribution is expressed as _ x B exp 1 8 gt p BT x where a is the shape parameter is the scale parameter and is the gamma function The distribution parameters were estimated using maximum likelihood estimation MLE Daily precipitation distributions are typically heavily skewed towards low intensity values As a result the distribution parameters will be dominated by the most frequently occurring values but may not be able to accurately describe the properties of extreme values To capture the main properties of normal precipitation as well as extremes the precipitation distribution was divided into two partitions separated by the 95th percentile The resulting distribution is hereafter referred to as the double gamma distribution Two sets of parameters a p and amp os Pog Were estimated from observations and the RCA3 output in the control period 1961 1990 These parameter sets were in turn u
19. lick on the green light to save settings and close the tab Enter start date as calendar if known fima Enter start date as calendar if known or leave empty or leave empty Variable attributes J Variable attributes 1 units days since 1950 01 01 00 00 00 1 units days since 1950 01 01 00 00 00 2 standard_name time a 2 standard name time 3 long_name time D 3 long_name time 4 calendar standard A 4 calendar standard 5 bounds time_bnds 5 bounds time_bnds Ps Date YYYY MM DD YY YY MM D DI VAT Date YYY MM DD 1958 01 01 Calendar type NORMAL w Pesa Cancel Ok Canc ONLY360 figure 10 Information about the start date and calendar type for the data contained in the netCDF can be entered here The NORMAL calendar is the default setting with two other options ONLY360 and NOLEAPS If no date is entered for the data in the netCDF file then relative time steps will be used starting at 1 4 5 1 3 Loading ASCII files To load information for an ASCII file click on the source box and choose ASCIIDATA A secondary drop down menu will open with three choices figure 11 Discrete non continuous data such as parameter files Timeserie time orientated time series 1 e they have time on the x axis 13IPage Idserie id orientated time series 1 e they have id on the x axis pac SHAPEFILE _ RASTER figure 11 The secondary drop down menu for ASCIIDA
20. o load automatically on start up This is not necessary and can be left blank You can also add the term 7 Page nogui to the end of the command this will automatically load and run the xml file entered without opening the user interface java Xss2048k Xms512m Xmx1024m jar DBS jar tasklist xml nogui 3 3 Data requirements This version of the DBS Tailoring system works with precipitation and temperature data An important note it is important to choose the correct temperature variables from the scenario to scale Some models have variables that represent non grid averaged data which is a better representation of observations than grid averaged data So if you are using observation data as a reference it is encouraged that you try to use a non grid averaged variable such as tasol temperature near surface over open land from the RCA model On the other hand if you are using modelled data as a reference e g Eralnterim data then it is encouraged that you use a grid averaged variable e g t2m 2m temperature and tas temperature near surface The DBS stage of the system requires ASCII files to be time oriented that is to say with time on the x axis and id on the y axis fig ID TIME 19810101 19810102 19810103 19810104 ID TIME 21619 21620 22099 22100 21619 10 49 9 51 8 26 6 69 19810101 10 49 12 87 10 14 12 38 21620 12 87 12 58 11 11 9 69 19810102 9 51 12 56 9 14 11 89 22099 10 14 9 14 7 49 6 47 19810103 8 26
21. of two variables important for hydrological modelling namely precipitation and temperature DBS is an acronym for Distribution Based Scaling the method was developed at the Swedish Meteorological and Hydrological Institute 2 Theory 2 1 The DBS Tailoring system The DBS tailoring system is a tool used primarily for facilitating bias correction downscaling using the DBS approach see the theory sections Precipitation and Temperature below The DBS tailoring system is used both to prepare the different data and control files needed to perform a DBS bias correction downscaling and run the DBS motor A typical DBS job can be broken down into six general steps 1 Calculate the statistical parameters related to the reference precipitation This step requires an ASCII file containing observed precipitation or the desired analogous substitute for a chosen reference period 2 Calculate the statistical parameters related to the simulated precipitation This step requires an ASCII file containing modelled precipitation for a chosen reference period and the parameters from step 1 3 The scaling of the simulated precipitation This step requires an ASCH file containing modelled precipitation for the entire period of interest and the parameters from steps 1 and 2 4 Calculate the statistical parameters related to the reference temperature This step requires an ASCII file containing observed temperature or the desired analogous substitute for a
22. on the task in the workspace will open a tab where you can choose which file you want to use in the system At the moment the system can only work with three types of files figure 7 shapefiles see Loading shapefiles netCDF files see Loading netCDF files and three types of ASCII data see Loading ASCII files The other file types are unavailable in this version but will be activated in future versions This step creates a posmap file a file that contains information regarding the data contained in the file i e number of data points data ranges etc These POSMAP files with their information regarding how the data is stored in each file allows the system to identify where in the ASCII files the is stored which allows for faster reading of the data If you have prior tasks that have created data files clicking on the source box will give you the option to load a system file or a task output By choosing system file you will enter the menu as seen in figure 7 below while choosing task output will open a drop down menu with all the task outputs that can be loaded at that stage Click on the green light to save settings and close the tab By clicking on the red light will cancel any changes made and close the tab EA lt no last save SHAPEFILE SouICE RASTER Gener NETCDF pops IMAGE GRIB kl Timesene PTGRID figure 7 Left clicking on the task opens the task tab Here you can choose the data
23. ou wish to plot all of them otherwise chose the one you are interested in Clicking on OK will automatically fill the Data source box with your choice Next click on the Target box here you enter the location and name of the images that will be created It is advised to save the images as PNG files this is done by adding png as the file extension when entering the file name When this is done a popup window will appear figure 15asking you to enter the image width The larger the number the larger the Image so you are free to choose the image size a width of 800 is suggested for a standard sized image 15lPage figure 15 Popup box to eneter in the image width a width of 800 can be used as standard 4 5 3 EXPORT This task allows you to export data from a file The task can be divided into two main types of export 1 e using geographic information to select the relevant data and using unique ids to select the relevant data The former type is mostly used to extract data from netCDF files to be in the scaling process The export task uses the geographic data in the shapefile to select the nearest corresponding point in space from the netCDF file and assigns the id number from the shapefile to the data at that point The latter type 1s more versatile and can be used for a number of different tasks e g export statistics for the data to a shapefile for analysis purposes or extract a subset from an existing dataset
24. s not on the empty work space you will see the following drop down menu figure 5 With this drop down menu you can delete or edit the task you clicked on insert a new task before the task you clicked on delete the entire list or delete large parts of the task list 1 e all tasks before excluding this task or all tasks after including this task with the last two options Another helpful feature are the arrows that appear if you move the cursor over the left margin of the task boxes These allow you to move that task up or down the list 9lPage amp no last save Delete Edit Insert new prior Remove group Remove rest including Remove prior excluding figure 5 By right clicking on the tasks you will get this menu Here you can delete or edit the task you clicked on insert a new task before the task you clicked on delete the entire list or delete large parts of the task list with the last two options 4 1 Save This option saves the xml file you are working on If you have not saved it for the first time a standard windows save popup window will appear where you can enter the file name and location 4 2 Save to This option allows you to save the xml file you are working on to a different location or under a different name It is the same as the save as option in many other programs 4 3 Load This option allows you to open existing xml files saved on your system If you already have an xml
25. s kwt Dx pry wetSin kwt k 1 i K 0 Dry Wet x x o t pry wet e ai gt Cx Dry wet cos kwt dx pry wetSin kwt k 1 where Qo Ax Dx Co Ck and dg are the Fourier coefficients is the day of the year w equals 2p n where n is the time units per cycle and k stands for the nth harmonic used for describing the annual cycle of adjusted daily temperature TDBS Theoretically t 2 1 harmonics are able to represent a complete cycle Five harmonics were used and found to be sufficient The DBS parameters for temperature were calculated for both observations and RCM simulated data series They are denoted Uops Sons Herr and Gcr and are used to scale the daily temperature Tpgs F71 ops Hops F Treas Oct Ucri 6lPage 3 Requirements 3 1 System requirements To run the DBS tailoring system you need to have Java 7 installed on your system The necessary files can be downloaded from http www oracle com technetwork java javase downloads index html 3 2 Program files Create a work folder on your hard drive and copy the DBS tailoring system files to this folder and a lib subfolder Your folder should look like figure 1 and the lib folder should look like figure 2 The program files can be downloaded from this address from within SMHI winfs proj proj klimathydrologi DBS program_files or you can send a request for a download link to kean foster smhi se if you do not have access to the SMHI network Name Da
26. sed to correct the RCA3 outputs for the entire projection period up to 2100 using the equation a gt 0 f x Pops F aops Bobs E P acri Bert if P lt 95th percentile value F amp 0p5 95 Bons9s F P acros BcrLos if P lt 95th percentile value Ppps where Obs denotes parameters estimated from observations and CTL denotes parameters estimated from the RCA3 output in the control period F represents the gamma probability distribution The DBS parameters for daily precipitation were optimized separately for each sub basin to preserve spatial variability of the RCA3 outputs To take seasonal dependencies into account they were also optimized for each season DJF Dec Feb MAM Mar May JJA Jun Aug and SON Sep Nov 2 3 Temperature Compared to precipitation temperature is more symmetrically distributed It can be accurately described by a normal distribution with mean m and standard deviation s SlPage 2 f x e Wee ov 21 The mean and standard deviation of daily temperature on a Julian day were smoothed over the control period using a 15 day moving window Considering dependence between precipitation and temperature temperature time series were described with distribution parameters conditioned by the wet or dry state of the day Fourier series were used to smooth the seasonal mean and standard deviation of temperature for wet and dry days K P Ao Dry Wet n A u t pry wet gi Ax pry wet co
27. te modified Type d lib 2012 03 26 09 13 File folder DBS Jar 2012 02 10 17 16 Executable Jar File CHI DES generator exe 2012 02 14 14 13 Application 5 Run_DBS bat 2012 01 18 14 07 Windows Batch File figure 1 Image of the work folder showing the required program files The contents of the lib folder are shown in figure 2 below Name Date modified Type Le appframework 1 0 3 Jar 2012 02 10 17 16 Executable Jar File Le beansbinding 1 2 1 Jar 2012 02 10 17 16 Executable Jar File Poot 012 02 10 17 16 Executable Jar File 2012 02 10 17 16 Executable Jar File 012 02 10 17 16 Executable Jar File Jhlabs jar o cu E 4 Cu ro j ee a Pa Le Library_dataform jar Le netcdfAll 4 2 jar 2012 02 10 17 16 Executable Jar File Le swing worker 1 1 jar 2012 02 10 17 16 Executable Jar File figure 2 Image of the contents of the lib folder The system can be opened by double clicking on DBS jar however it is advised to use the batch file Run_DBS bat to open the system as it increases the available computing resources to the system and relative information regarding the scaling job can be read in the command window If you do not have access to the batch file you can make your own using a text editor and typing in the following command java Xss2048k Xms512m Xmx1024m jar DBS jar tasklist xml The yellow highlighted text above is the name of the xml file you would like the system t
28. to fllip transpose and the output file 22 Page 5 Examples 5 1 Typical DBS job process diagrams Scaling process using modelled reference data E Hype 23 Page Scaling process using observed reference data S Hype 241Page lt eJouar SSa901d Buljeos JOSE EP D9d9 F1 UM EJEP 39UAJ9JAI JO UONIALIOI U0122109 eeg i F F DT 3 i s D Buisn HOSY 0 eyep ASBHLd AIBuIq ynu uol2e11X3 eeg 5 2 Sample xml files As mentioned earlier The DBS Tailoring system the XML files are used to run the DBS tailoring system Here are two examples of xml files The first file is for when you already have the 25 Page reference data in ASCII form and the second 1s for when you have to extract reference data from a netCDF file i e when using EraInterim data Please note that these are examples and are not generic for every situation you will need to at least need to input your own data filenames dates etc Il tasklist_a xml Il tasklist_b xml 26 Page
29. you want to load into the system by right clicking in the source box the currently supported file types are shapefiles netCDF files and ASCII files discrete data time series with time as the X axis header lllPage or time series with ID number as the X axis header The other options are not available in this version but will be available in the next version of the system 4 5 1 1 Loading shapefiles To load information for a shapefile click on the source box and choose SHAPEFILE This will open a normal windows load window where you can choose the shapefile you want to load Click on the green light to save settings and close the tab NOTE shapefiles should have a latitude longitude projection preferably WGS84 4 5 1 2 Loading netCDF files To load information for a netCDF file click on the source box and choose SHAPEFILE A secondary drop down menu will open but only one option is selectable at this stage Choose single this will open a normal windows load window where you can choose the netCDF file you want to load information for NOTE netCDF files should have a latitude and longitude as field variables After selecting the netCDF file the following window will appear figure 8 where you need to identify the netCDF variables that are needed Right click on the relevant variables and select from the menu to set the field To use a netCDF file you need to set four fields 1 e longitude latitude the data field and time Ident
Download Pdf Manuals
Related Search