Home

User Guide - WRc Websites

Contents

1. Parameter Fitting Interface Smoothing period alculate SRG r Fit Parameters Parameters sr C Quarterly Read Existing Parameters ete Hell Hal Years Rainfall data Compare Filename csv Select Save Rainfall and Thresholds Exit iE f Calculate SRG Parameters You must click the E UETENUI DI button to fit the model parameters to the rainfall statistics calculated when you clicked on the site information form this process may take a long time Ensure that the Fit parameters option is selected When these parameters have been generated they will appear in a table at the bottom of the screen See Appendix A for examples Simulate Rainfall If your database contains existing parameters or if you have just fitted them to your sample rainfall statistics the parameters are produced after clicking the button click the Simulate Rainfall button to generate x number of years of SRG rainfall data This is saved to the SRGData table in the STORMPAC 4 1 database x is the Record Duration years specified on the site information form When rainfall has been simulated three graphs will appear on the right hand side of the screen These show a comparison of the simulated data and any historical daily data which has been imported in the form of a time series a cumulative frequency graph and a quantile plot which shows rainfall at points of equal probability of occurence These charts can be viewed
2. 97 STORMPAC 4 1 User Guide 1 THE STORMPAC USER GUIDE 1 1 STORMPAC 1 1 1 Overview STORMPAC is a software package combining a Stochastic Rainfall Generator SRG with Disaggregation models 9 and hourly rainfall processing capabilities The range of facilities available is i Generation of hourly rainfall time series for any location in the UK ii Threshold analysis of generated hourly rainfall and historic daily rainfall data iii Processing of hourly rainfall historical or generated into a chronological event database iv Sorting the chronological event database using storm depth intensity duration dry period Urban Catchment Wetness Index UCWI 30 day Antecedent Precipitation Index API30 or bathing season as the identifier v Disaggregation of hourly rainfall data into 5 minute rainfall intensity values using a new disaggregation model vi Five minute intensity values output in a format suitable for input to HydroWorks InfoWorks Rwin or SIMPOL 2 1 The current version of the STORMPAC Annually Regionalised SRG is not recommended for use with higher rainfall catchments where the average altitude is above 381 m or where there is a local microclimate In these situations local historical data should be used 1 1 2 Who will use STORMPAC STORMPAC has been specifically developed as part of the Urban Pollution Management UPM programme and is intended for use in conjunction with the other software package
3. Essential catchment location parameters Historical daily rainfall data Average Annual Rainfall SMD Data Need Source Selection Length of record historical Format of data historical Simulated records Format of data simulated The Event File Need Source Length of record Length of storm Format of data The WASSP PCD File Need Format DESCRIPTION OF MAIN FORMS AND BUTTONS 6 1 General 6 2 Parent Window 6 3 Site Information Form 6 4 SRG Parameter Fitting Interface Form 6 5 SMD Form 6 6 UCWI API30 Calculation Form 6 7 Storm Event Definition Form 6 8 Hourly Rain Disaggregator Form 6 9 File Export Form 6 10 Data views form 6 11 Multi RED analysis f SMD MODEL 8 FREQUENTLY ASKED QUESTIONS 8 1 Do need Microsoft Access to run STORMPAC 4 1 8 2 What is the minimum length of daily record should import 8 3 What should I do if have no data at my site to compare values 8 4 I m unsure if my SMD calibration is good enough 8 5 Should daily or annually regionalise 8 6 What return periods is STORMPAC 4 1 valid for 8 7 Developments in STORMPAC 3 2 compared to STORMPAC 2 0 8 8 What was new in version 4 0 8 9 What is new new in version 4 1 APPENDIX A WORKED EXAMPLES USING STOCHASTIC RAINFALL GENERATOR APPENDIX B FLOW DIAGRAM APPENDIX C REPORT FOR WRc ON A STOCHASTIC DISAGGREGATION PROCEDURE BASED ON A POISSON RECTANGULAR PULSES MODEL 85 APPENDIXD RED FILES FORMAT 83
4. 5 minute disaggregated data are stored in this table once they have been generated Data are only stored between the start and end times of events as defined in the events table This table contains the criteria used to define the events The check boxes chosen and values used are recorded within this table Also the number of events generated is updated as different filters are applied The way in which the filtered events were last sorted is also recorded A number will appear under the sorting parameter tab that will correspond to one of the parameters below e Time event number 0 e Total event depth 1 e Mean rainfall intensity 2 e Maximum hourly rainfall intensity 3 e Duration 2 4 e Dry period 5 e UCWI 6 e API30 7 Sort order as ascending or descending is also recorded e Ascending 0 e Descending 1 This table contains the statistics of the chosen events and is described in more detail in Section 5 4 20 WRc plc 2008 General Hourly SMD SMDParameters SRGData SRGParameters UCWlandAPI30 STORMPAC 4 1 User Guide This table contains various fields that describe the project you are currently working on such as Project Title Easting Northing Coastal Distance Altitude SAAR This table contains a DateTime field and a Depth field mm A chronological file of historical hourly data is stored in this table if hourly data have been imported This table contains Date and SMD mm fields w
5. C2 THE POISSON RECTANGULAR PULSES MODEL 85 WRc plc 2009 STORMPAC 4 1 User Guide C2 1 Model definition and background In the Poisson rectangular pulses PRP model rain cells have arrival times that occur in a Poisson process with rate A Each rain cell has a random lifetime which is distributed as an independent exponential random variable with parameter n The intensity X of each rain cell remains constant throughout the cell lifetime we will take X to be an independent Weibull random variable with parameters a and 6 The total rain intensity at any point in time is the sum of the intensities of all cells alive at that point Rodriguez lturbe et al 1987 compared the PRP model based on an exponential distribution for cell intensity to the Neyman Scott and Bartlett Lewis Poisson clusters models The Poisson cluster models were shown to outperform the PRP model when fitted to full records of hourly data However as we shall be using the model for disaggregating wet sequences only without attempting to fit long dry sequences between storms this is not regarded as a reason to reject the model Furthermore Cowpertwait et al 2004 found that the spatial PRP model performed well when disaggregating hourly data for input in urban catchment models Consequently it seems reasonable to use the PRP for temporal disaggregation also as this is just a special case of the spatial PRP model C2 2 Model properties and fitting procedure Let Y
6. 31 10 2046 05 00 Figure 5 3 STORMPAC Form Used to Define Storm Events 5 4 2 Source Initially STORMPAC processes hourly rainfall into an event file ordered chronologically As described above these data can then be sorted by any of the variables included in the events summary Other sources of rainfall can also be imported to the package and processed at this stage once converted into the correct format 5 4 3 Length of record The maximum length of record that can be processed in STORMPAC is 100 years 5 4 4 Length of storm There is no maximum length of storm event that can be produced by STORMPAC 5 4 5 Format of data Each record in the Events table in the project database represents a storm event The columns are defined as shown in Table 5 6 49 WRc plc 2009 STORMPAC 4 1 User Guide Table 5 5 Column Headings Used in the Events Table of the Project Database Column Variable Number Start Date Time End Date Time Depth mm Mean Intensity mm hour Maximum Intensity mm hour Duration hours UCWI mm API30 mm Antecedent Dry Period hours ON Oa FW NY 5 5 The WASSP PCD File 5 5 1 Need WASSP PCD files can be used directly with WASSP The data contained within them are disaggregated rainfall STORMPAC has the facility to import PCD files populating the Disaggregation table in the project database The five minute values are also summed to obtain hourly rainfall da
7. table and hourly stochastic rainfall SRGData table These are presented either as monthly or threshold analysis and can be used to compare the generated hourly data with imported historical daily data Selecting threshold analysis gives the average number of days above any given threshold in a single bathing season Monthly average rainfall shows the average monthly SRG rainfall from the SRGData table compared to the average monthly rainfall in the Daily table If both SRG and Daily historic data are available both are shown so that a comparison can be made 61 WRc plc 2009 STORMPAC 4 1 User Guide Clicking the Graph button allows you to view these results as a graph 60 50 E 3 P S s Daily Average number of days in a bathing season I o 0 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 Threshold rainfall mm Clicking the Table button shows them as a table of data This data can be copied to MS Excel 62 WRc plc 2009 STORMPAC 4 1 User Guide Query Graph j i Average Average days per days per Threshold year year mm threshold threshold exceeded exceeded SRG historic 6 11 Multi RED analysis This requires selection of one or more Stormpac databases with data covering the same time period and either all stored in the Hourly tables or all stored in the SRGData tables Use to add databases multiple databases can be selected if th
8. 0 0 0 000 12 1993 1 10 0 0 0 0 000 12 1993 1 11 0 0 0 0 000 12 1993 1 12 0 0 O 0 000 12 1993 1 13 0 0 O 0 000 12 1993 1 14 0 0 0 0 000 12 1993 1 15 0 0 O 0 000 12 1997 12 20 0 0 0 9 800 12 1997 12 21 0 0 0 0 000 0 1997 12 22 0 0 0 0 000 0 1997 12 23 0 0 0 0 000 0 1997 12 24 0 0 0 3 700 12 1997 12 25 0 0 0 9 300 12 1997 12 26 0 0 0 11 900 12 1997 12 27 0 0 0 28 700 12 1997 12 28 0 0 0 10 900 12 1997 12 29 0 0 0 9 400 12 1997 12 30 0 0 0 7 300 12 1997 12 31 0 0 0 4 800 12 1998 1 1 0 0 0 17 000 12 Quality Types 0 Good 2 Edited 3 Snow 4 Suspect 8 Missing Man Entered Quality Types 10 Good 12 Edited 13 Snow 14 Suspect 18 Missing Dig Quality Types 20 Good 22 Edited 23 Snow 24 Suspect 28 Missing M O Quality control 30 Good 32 Edited 33 Snow 34 Suspect 38 Missing 40 WRc plc 2008 STORMPAC 4 1 User Guide Table 5 3 Example of STM Format for Hourly Rainfall Data RAINFALL EVENT FILE Filename dum STM Location dum Number of years 10 Hourly rainfall file dum zim SMD file system smd Event selection criteria events satisfy all four criteria 1 Total rainfall depth gt or 1 00 mm 2 Mean rainfall intensity gt or 10 tron 3 Max Hourly rainfall intensity gt or 10 mnm 4 Dry period defining end of event gt or 1 hours YEAR M D ST DEN TOT
9. 1 mm day in Winter and 3 mm day in Summer or the evaporation values that you have just calculated during the SMD calibration stage For this exercise use the default values of Soil Class 1 and Winter 1 mm day Summer 3 mm day After these boxes have been filled in click the Calculate UCWI API30 button to start the calculations 29 WRc plc 2008 STORMPAC 4 1 User Guide If data exist in the database you will be asked Do you want to overwrite existing UCWI API30 data Clicking Yes will overwrite the UCWI data in your database whilst clicking NO will cause the UCWI Calculation data form to unload On clicking Yes a plot of the calculated UCWI API30 and rainfall will be displayed after the calculations have finished along with a message box stating a T ASUED The purpose of the plot is to provide the user with a visual plot of the UCWI and API30 values in relation to the rainfall Confirm that when rainfall values increase the UCWI values increase as well Clicking No will take you back to the UCWI Calculation data form where you will be allowed to re enter the preceding five day rainfall and SMD at end of day values Clicking Yes means that the UCWI data will be written to the database Click Yes for this exercise Click OK on the message box stating UCWI API30 Calculation Complete and then Exit on the UCWI Calculation data form to return to the Control Interface form We have now reached step 4 on Figure 4 1 4 8 Step 4
10. 6 0 23 2 2 17 1 3 00 00 00 00 00 0 0 0 0 00 00 00 00 00 00 00 00 00 0 0 0 0 2000 1 4 0 0 00 00 00 00 00 00 00 00 00 0 0 0 0 0 0 00 00 00 00 00 00 00 00 00 0 0 0 0 2000 1 5 0 0 00 00 00 00 00 00 00 00 00 0 0 0 0 0 0 00 00 00 00 00 00 02 01 07 14 1 3 2000 1 6 1 2 02 00 00 00 00 00 00 00 00 0 0 0 0 0 0 00 00 00 00 00 00 00 00 00 0 0 0 0 2000 1 7 0 0 0 1 03 02 00 00 00 00 00 00 24 1 3 0 0 00 00 00 00 00 00 00 00 00 0 0 0 0 2000 1 8 0 0 00 00 00 00 00 00 00 00 00 0 0 0 0 0 0 00 00 00 00 00 00 00 00 00 0 0 0 0 2000 1 9 0 0 00 00 00 00 00 00 00 00 00 0 0 0 0 0 0 00 09 03 00 00 00 00 00 00 0 0 0 0 2000 1 10 0 0 00 00 00 00 00 1 7 27 1 7 1 1 14 0 9 0 8 09 13 0 7 07 0 7 06 03 0 1 01 00 0 0 Format line 1 XXXXXXXXxx2000xxxxx 1xxxxx 1 line 2 Xxx 0 0xxx 0 0xxx 0 2xxx 1 2xxx 2 3xxx 2 2xxx 1 7xxx 0 8xxx 0 1xxx 0 0xxx 0 0xxx 0 0 Key 2000 Year 1 Month 10 Day x Space _ Space for extra digit if required 0 8 mm rainfall hour 13 39 WRc plc 2008 STORMPAC 4 1 User Guide Table 5 2 Example of SCF Format for Hourly Rainfall Data GAUGE REF _ 562150 M O GAUGE REF 562150 STATION NAME Wayoh Res No 2 Year Month Day Hour Minute Second Rainfall Quality 1993 1 2 0 0 0 0 000 8 1993 1 3 0 0 0 0 000 12 1993 1 4 0 0 0 0 000 12 1993 1 5 0 0 0 0 000 12 1993 1 6 0 0 0 0 000 12 1993 1 7 0 0 0 0 000 12 1993 1 8 0 0 0 0 000 12 1993 1 9 0
11. 73 0 55 39 80 5 80 3980 5804 8 2000 2 27 1 443 22 0 35 7 80 220 7 80 220M 3 2000 3 2 16 101 27 0 34 9 30 290 930 290 10 2000 3 3 22 15 13 0 15 2 00 200 200 200M 11 2000 3 10 10 155 25 0 26 6 60 340 660 3404 12 2000 3 16 21 142 14 0 24 3 30 240 330 2404 13 2000 3 24 15 184 36 0 18 6 50 290 650 2904 14 2000 4 3 1 212 53 0 52 27 30 2 20 2730 2204 15 2000 4 Ei 67 0 11 4 20 310 420 3104 10 nnn A 2 E 07 01 2000 000002 7 1 2000 00 00 07 01 2000 23 00 00 3 E_08_01_2000_000003 8 1 2000 00 00 09 01 2000 14 00 00 100 1 00 4E 13 01 2000 000004 13 1 2000 00 00 14 01 2000 19 00 00 100 23 1 00 5E 27 01 2000 000005 27 1 2000 00 00 29 01 2000 08 00 00 100 33 1 00 6 E 03 2 2000 000006 3 2 2000 00 00 04 02 2000 05 00 00 100 33 1 00 E_06_02_2000_000007 6 2 2000 00 00 09 02 2000 01 00 00 100 33 1 00 8 E 27 02 2000 000008 27 2 2000 00 00 27 02 2000 22 00 00 100 33 1 00 3E 02 03 2000 000009 2 3 2000 00 00 03 03 2000 03 00 00 100 33 1 00 10 E_03_03_2000_000010 3 3 2000 00 00 03 03 2000 13 00 00 100 33 1 00 11 E 10 03 2000 000011 10 3 2000 00 00 11 03 2000 01 00 00 100 33 1 00 12 E 16 03 2000 000012 16 3 2000 00 00 16 03 2000 14 00 00 100 33 1 00 13 E 24 03 2000 000013 24 3 2000 00 00 25 03 2000 12 00 00 100 99 1 00 14 E 03 04 2000 000014 3 4 2000 00 00 05 04 2000 05 00 00 100 39 1 00 15 E 07 04 2000 000015 7 4 2000 00 00 08 04 2000 14 00 00 100 38 1 00 16 E 18 04 2000 000016 18 4 2000 00 00 18 04 2000 15 00 00 100 99 1 0
12. 9 What is new new in version 4 1 Rainfall series of up to 100 years can now be generated Calculated SRG parameters and upper and lower parameter limits can be viewed Graphical views comparing simulated results with historical data are produced Updates to Data Views function Create output file function for multi site RED generation speeded up 71 WRc plc 2009 72 STORMPAC 4 1 User Guide WRc plc 2009 STORMPAC 4 1 User Guide REFERENCES 10 11 12 13 14 15 16 17 Cowpertwait P S P Metcalfe A V O Connell P E Mawdsley J A and Threlfall J L 1991 Stochastic Generation of Rainfall Time Series Foundation for Water Research Report No F0217 December 1991 Foundation for Water Research 1994 Urban Pollution Management UPM Manual FR CL 0002 Cowpertwait P S P and Threlfall J L 1994 Further Developments of the Stochastic Rainfall Generator Foundation for Water Research Report No FR0438 DOE NWC Design and Analysis of Urban Storm Drainage The Wallingford Procedure 1981 STC Report No 28 NWC Hydraulics Research Ltd 1991 WALLRUS User Manual Danish Hydraulic Institute 1990 MOUSE User s Guide amp Technical Reference Threlfall J Cowpertwait P S P Strandner H O Connell P E Kilsby C G and Mellor D 1998 Adaptation of Rainfall Generation Model Technology Validation Project IN101871 WRc Report Number UC3254 Cowpertwait P S P 1
13. be the rainfall depth in the i th time interval which for our purposes will be of duration 5 minutes Then the following properties are given in the literature the first two in Rodriguez Iturbe et al 1987 and the last follows as a special case in Cowpertwait 1998 The mean p E Y A E X n 1 The autocovariance y Cov Y Y 2 Ax E X n 2 The third moment E Y 6 E X n 2 ne 2e yjg 3 In the above E X a F 1 r 8 A n e 1 and A 7 1 e e k gt 0 The model can be fitted by matching the above properties to their equivalent values taken from the sample which can be achieved using a minimisation procedure based on the squared differences see Cowpertwait et al 1991 2004 C3 FITTED MODEL C3 1 Parameter estimates Using the above properties u yo Y1 the parameters were estimated for each calendar month for a 31 year record of 5 minute data from Farnborough UK Following Cowpertwait et al 2004 only wet hourly sequences were included in the estimation of the sample properties The parameter estimates are given in Table C1 86 WRc plc 2009 STORMPAC 4 1 User Guide Table C1 Parameter Estimates for Farnborough 1941 71 Month estimate A n 0 a 1 0 5771 0 4605 0 4890 0 0153 2 0 7480 0 4037 0 4409 0 0079 3 0 8003 0 6081 0 4663 0 0130 4 0 3899 0 6153 0 5878 0 0403 5 0 3480 0 7712 0 6006 0 0727 6 0 6499 0 7470 0 4357 0 0212 7 0 87
14. checked by default and we recommend that this should always be used as it replicates five minutely data sets to a higher standard than the Ormsbee method particularly for extreme values The Ormsbee method should only be used to replicate sets of rainfall created in previous versions of STORMPAC up to and including 3 1 and here the Ormsbee option needs to be selected Click OK when the correct file type and disaggregator required have been selected A progress bar will appear informing the user that the disaggregation calculation is proceeding Once completed the user is returned to the Control Interface form The data has been disaggregated to five minute interval values You are now in a position to export the disaggregated data to HydroWorks InfoWorks To stop a disaggregation that is taking too long to process click the Cancel button If a dataset of 10 years takes several hours to process and you want to speed up the process click the Cancel button and select the Accelerated Poisson option and Click OK If this processing takes too long then click the Cancel button and reduce the first or second iteration values to speed up the disaggregation See section 2 3 3 for details 32 WRc plc 2008 STORMPAC 4 1 User Guide 4 10 Step 6 Create Output File To export data to a HydroWorks file click the Create Output File button on the Control Interface form The File Export form will appear as shown in Figure 4 16 You can save data to e
15. consultation with the Met Office A representative soil type and land use value for the study catchment should be considered in selecting the data 5 3 4 Length of record historical SMD data are required in either a long term end of month average for each month of the year or a long time series of daily values if historical hourly rainfall data are to be processed and UCWI values calculated The latter is used by STORMPAC to create long term end of day averages for each day of the year The Met Office can supply long term end of month average SMD values 5 3 5 Format of data historical The Met Office supply SMD data in tabular form usually as an MS Excel worksheet STORMPAC requires monthly data to be in the form of a series of 12 rows one for each month An example file SMD XLS is saved on your computer when STORMPAC is installed Each row contains the month in column 1 and the long term average SMD value in column 2 STORMPAC will ignore the header information if present at the top of the worksheet The format of daily SMD time series is the same as that for daily rainfall data A full description of this is given in Section 5 2 3 Note the SMD data file should have the file extension XLS 47 WRc plc 2009 STORMPAC 4 1 User Guide 5 3 6 Simulated records Simulated SMD data are calculated from the imported rainfall data either daily or hourly or the simulated SRG hourly data Long term end of day averages for these
16. especially in regard to the extreme values where the Ormsbee method tended to underestimate Therefore it is recommended as a suitable replacement to the Ormsbee disaggregator Cowpertwait 2005 9 Appendix C presents a comparison of the new Poisson and old Ormsbee methodologies evaluating them both against historic data using quantile plots This highlights the improvement seen using the new Poisson method It should be noted that although the option is still available to use the Ormsbee method it should only be used to replicate rainfall sets created in previous STORMPAC versions up to and including 3 1 10 WRc plc 2008 STORMPAC 4 1 User Guide Disaggregation Procedure The disaggregation procedure can be summarised as follows A 1 hour rainfall depth is read in from the database which contains the hourly series to be disaggregated A 5 minute series is simulated using the fitted model The 5 minute simulated series is summed and the 1 hour total of the simulated series compared to the 1 hour total that was read in The simulated series is discarded if the absolute difference between the simulated 1 hour total and the total read in exceeds 5 The process is repeated until the totals are in agreement to within 5 The simulated series then represents a possible realisation of 5 minute data representative of the 1 hour value that was read in Following this procedure a record of 1 hour rainfall depths can be disaggregated Some addition
17. existing Multi RED i database database analysis DASTORMPACNdb mdb Site Statistics Disaggregated Import Rainfall Data i p l Daily Data Hourly Data Data Figure 4 6 Control Interface Form during runtime If you click Site Statistics without first importing some data you will only be able to annually regionalise your rainfall series e However if you import rainfall data as daily hourly or disaggregated data see Section 5 for a description of the different data formats you will be given the options of annually or daily regionalising your data or calculating statistics from your hourly historical data e UCWI values cannot be calculated until you have imported or generated some hourly data You have now reached step one as shown in Figure 4 1 4 5 Step 1 Import Historical Hourly Data The starting point for this exercise is historical hourly rainfall data The example data files RAIN SIM SMD XLS supplied with STORMPAC can be used for this exercise Click on the button and find the file RAIN SIM in the folder where STORMPAC was installed 23 WRc plc 2008 STORMPAC 4 1 User Guide Select the RAIN SIM file and click Open to import the historical rainfall into your database There should be one year of rainfall data in this file year 1988 When using other data files it is important that they are in the correct format see Section 5 4 6 Step 2 Calculate Site Statistics The Site Information form i
18. far right hand column 64 WRc plc 2009 STORMPAC 4 1 User Guide Events can be sorted to help select a set of events to export as RED files or changing the export order by clicking on column headers to cycle between sort ascending and sort descending To reduce the number of events that will be exported as RED files select a block of rows then click on the Note select deselect The lower grid displays the chosen events and allows you to change the following Event name C N gt tick box far right hand column to select deselect a block of events if you select a mixture of selected and deselected rows they will toggle rather than all Default UCWI to be used if the individual Stormpac files not have UCWI values Default API30 values to be used if the individual Stormpac files not have API30 values 4 Default Antec values if the individual Stormpac files do not have Antec values MultiRED files Select STORMPAC databases te event criteria Select events Create output files Close Save as text Save as XLS Export all events Inter E vent Max Include eve intensity mm hr Avg Intensity mm hr 1 2000 1 i 6 70 6 70 2504 2 2000 1 0 18 4 20 220 420 2204 3 2000 1 8 gt 2 x 147 44 40 5 60 44 40 5604 4 2000 1 13 7 78 43 0 15 6 60 210 660 2104 5 2000 127 3 301 56 1 02 57 00 7 60 57 00 7 601 6 2000 2 3 14 135 23 0 38 10 30 4 70 1090 470 2000 2 6 1 42
19. functions If fit is poor note the values of the calculated parameters compared with the upper and lower bounds for the five parameters In Figure 2 2 Beta is mostly at the lower bound and Nu and Lambda are mostly at the upper bound Upper bound 0 5000 5 000 30 00 0 0400 0 510 100 0 0010 0 0398 3 13 30 0 04 0 0801 2 97 30 0 04 0 0783 0 041 2 14 30 0 04 0 0708 0 041 2 83 20 8 0 04 0 05 0 152 2 65 26 2 0 0232 0 0504 0 329 3 48 26 4 0 0186 0 05 0 195 2 92 14 2 0 0393 0 0503 0 041 2 59 30 0 04 0 05 0 041 2 23 30 0 035 0 0692 2 44 25 8 0 0391 2 37 28 7 0 04 Figure 2 2 Calculated SRG parameters Weighted Estimates At some sites there may be some limited historical data available in which case it would be appropriate to use weighted averages of the regression estimates and site estimates taken from the site data A check box is available to switch this function on or off Statistics are calculated using the available historical data and regionalised model It is typical to use this function if only a limited amount of hourly data is available at a site Generation The rainfall time series is produced using random numbers with the regionalised model algorithms Although STORMPAC can produce 100 years of simulated rainfall it is recommended that series that are the same length as historical rainfall series are produced the maximum length for a historical series is 50 years The generated rainfall statistics cannot b
20. necessary user inputs are given in Sections 2 4 and 5 Two versions of the SRG exist in STORMPAC 4 1 an Annually Regionalised version and a Daily Regionalised version It is also possible to produce weighted averages of site statistics if short less than 10 years daily or hourly historical rainfall records are available at a site The SRG model generates storm events where the storm event contains a series of possibly overlapping rain cells The storms arrive as a Poisson process with rate parameter Lambda per hour Each storm generates a number of rain cells taken from a Poisson distribution with the mean number of cells per event being given by the parameter 1 Nu The average time between rain events within the storm is given as an exponential distribution with parameter Beta per hour Each rain event is exponentially distributed with parameter Eta per hour Finally the intensity of the rainfall is taken as an exponential distribution with parameter Shape hours per mm 7 WRc plc 2008 STORMPAC 4 1 User Guide Annually Regionalised The Annually Regionalised version of the SRG utilises average annual rainfall data to help in the regionalisation process This version has not been developed for use for catchments with a microclimate or with an average altitude above 381m and corresponding high SAAR values User experience suggests the Annually Regionalised SRG should not be used where the SAAR is above 1500mm The use of histori
21. rainfall is disaggregated and make sure the Poisson option is selected to use the Poisson disaggregator Then click OK A progress bar will appear informing the user that the disaggregation calculation is proceeding Once complete the user is returned to the Control Interface form You have now disaggregated the data to five minute values and are ready to export the disaggregated data to either HydroWorks SIMPOL or RWIN format files Export To export data to a SIMPOL file you need to display the File Export form by clicking the Create button on the Control Interface form You can save data to a SIMPOL file format by selecting this option from the Export File Type drop down box For the purposes of this exercise you require all storms sorted in chronological order You have already sorted the storms in the Event Definition form To select all storms leave the box blank Select the Simulated option because the events were from simulated rainfall Click Browse to select where the exported files will be saved and to name the file Enter a name in the file name box and click Save Click Save to save the files to the specified directory when the message box SIMPOL file export complete appears click OK and Exit to exit the File Export form Click STOP on the main menu bar to exit STORMPAC 81 WRc plc 2009 82 STORMPAC 4 1 User Guide WRc plc 2009 STORMPAC 4 1 User Guide APPENDIX B FLOW DIAGRAM Flow diagram showi
22. reference Line 3 Station name Line 4 Year Month Day Hour Minute Second Rainfall Quality Line 5 to end of data Comma separated data corresponding to order of headings Beyond last line of data Definitions for Quality values 37 WRc plc 2008 STORMPAC 4 1 User Guide 5 1 5 4 HIS files Same format as SIM files 5 1 5 5 STM files These are events output files from STORMPAC2 STORMPAC 4 1 extracts the hourly rainfall data and corresponding dates To ensure a continuous hourly rainfall series zero values are inserted for all hours between events The hourly time series can then be used to redefine events or create output files An example of this format can be seen in Table 5 3 5 1 6 Missing data Missing data points are expected to take the form of either 999 99999 or 99999 If these values are imported from any file format zeros are substituted at the relevant date time position An error message will appear stating a gap has been found if there are any chronological gaps in the rainfall data from a SCF file The import process will then continue Once complete however the user should correct the fault in the data and reload the file into STORMPAC before performing any further analysis Careful consideration should be given to missing data within the data file These values need to be edited to show 0 0 although a global search and replace is not recommended The most prudent procedure to follow is to
23. stored in the project database and will be available if the project is subsequently re opened for further analysis Location input To regionalise the SRG some local catchment and geographical input details are required Essential inputs include grid reference distance from coast km and altitude of study catchment m Sources and formats Grid reference A 4 figure easting and northing grid reference is required for the local catchment The first number relating to the National Grid sheet reference can be obtained from an ordnance survey OS map or by looking at Figure 5 1 and selecting the value that corresponds to the study catchment It should be noted that the map shown only provides the first figure of the each 4 figure grid reference The remaining three figures are obtained from an OS map The range of grid reference values that STORMPAC recognises is shown in Table 5 4 If the site under consideration is outside the allowable range STORMPAC will generate a warning message to inform the user However the SRG parameters will then still be calculated using an extrapolation procedure It should be noted that the resulting SRG parameters may not be as reliable as ones calculated for sites within the allowable range 42 WRc plc 2009 STORMPAC 4 1 User Guide Figure 5 1 Map of the UK to Identify the First Figure of the Northing and Easting Grid Reference for the Study C
24. time period Typically the first four criteria will be used to sort the data 2 1 6 Selection Once the events have been sorted the first x storms can be selected where x is a number between one and the maximum number of storms identified and sorted 2 1 7 Disaggregation The entire historical or SRG hourly data series can be disaggregated down to a five minute intensity value 2 1 8 InfoWorks HydroWorks Rwin SIMPOL STORMPAC formats the data following disaggregation into files compatible with InfoWorks HydroWorks Rwin and SIMPOL hydraulic models 2 2 When and how the SRG can be used 2 2 1 SRG Testing Testing of the output from the SRG model in STORMPAC 2 0 as a substitute for historical rainfall data in sewer models has been undertaken and reported in an FWR report Recommendations from the report are shown below They are intended to increase the level of confidence that the user can have when using rainfall series 2 2 2 Report Recommendations 1 Long complete hourly rainfall historical data series if available will provide maximum accuracy for any modelling study The necessary length of this series will depend on the user requirements For pollution management studies at least ten years should be used 6 WRc plc 2008 STORMPAC 4 1 User Guide If specific extreme events are required for example up to a ten year return period a series in excess of 20 years is necessary 2 Accuracy in the SRG model can be improved if
25. 0 17 E 25 04 2000 000017 25 4 2000 00 00 27 04 2000 13 00 00 100 33 1 00 65 WRc plc 2009 STORMPAC 4 1 User Guide PRODUCING OUTPUT Finally a location can be selected for where the output files are to be created Note that the selected folder will have an open folder displayed if the folder is displayed as closed you will need to click on it again to select it MultiRED files Select STORMPAC databases Define event criteria Select events Create outputfiles Close m Create output Sc v ERE Program Files wre EI Stormpac 32 Create output j 155 files left of 225 66 WRc plc 2009 STORMPAC 4 1 User Guide 7 SMD MODEL A sinusoidal model radians approximating the magnitude of potential evaporation over one year is given by the following equation PE F inf D Fy A 365 2 where F isthe seasonal amplitude factor typical values 0 1 to 10 X is the seasonal offset factor days typical values 30 to 30 A is the amplitude shift factor typical values 1 to 5 PE is Potential Evaporation any negative PE values are set to zero t is the day number from 1 1 January to 365 31 December F determines the amplitude of the shifted sinusoid function which has values between 0 and 2 F For example with A and X set to zero and F set to 1 the peak will occur half way through the year with a value of 2 which is equivalent to 2 mm A shifts the value of PE mm up neg
26. 0 255 1 152 2 092 2 582 3 456 2 304 3 382 4 256 5 406 5 154 4 8896 7 008 8 448 8 448 TIUS 6 624 6 045 4 032 3 456 2 592 4 016 4 880 2 660 1 440 0 564 1 152 2 3504 2 016 1 726 1 152 2 582 2 304 2 304 1 728 2 016 2 016 2 016 2 304 2 464 2 354 2 880 2 192 2 304 1 440 1 152 1 728 2 016 0 564 1 440 1 440 1 440 1 152 2 304 2 3504 2 320 1 744 1 152 0 864 0 664 0 664 0 206 0 000 2 EVENT z 20 00 5 170 84 VALUES UCWI 151 0 30 151 1 64 300 630 600 0 000 0 000 0 000 0 514 0 314 0 800 0 256 0 266 0 000 0 000 0 000 0 000 0 000 0 000 0 000 0 000 0 400 0 400 0 000 0 000 0 000 0 000 0 000 0 590 0 266 0 2568 0 000 0 206 1 360 1 360 1 646 1 152 1 726 2 016 1 440 0 664 1 440 1 440 1 856 1 250 1 565 0 864 1 440 1 440 2 016 1 152 0 576 0 564 0 564 0 564 1 456 1 744 1 744 1 726 1 440 0 664 1 056 1 632 2 206 1 726 1 440 1 152 1 152 0 576 0 285 0 800 O 512 O 512 0 000 0 000 0 256 0 256 0 000 0 000 0 000 0 000 51 304 564 596 440 564 464 016 744 HE DO PO OW BOM Oo 400 000 576 152 UB 564 2255 oOorrdaaon WRc plc 2009 52 STORMPAC 4 1 User Guide WRc plc 2009 STORMPAC 4 1 User Guide 6 DESCRIPTION OF MAIN FORMS AND BUTTONS 6 1 General The exit button on most forms allows you to unload the form from memory Cancel Calculations The cancel calculations button on most forms allows you to stop the current calculation Parent Window E Once you have loaded STORMPAC 4 1 click thi
27. 08 STORMPAC 4 1 User Guide Storm Event Definition r Data Type m Storm Criteria mm C Historic f Simulated v Rainfall depth for event 1 Number of years f Mean hourly rainfall intensity 1 Events filename Select Mar hourly rainfall Eel intensity D Program Files Stormpac 3 2NDem M Minimum inter event dry period hours Filter Save Events Events Close r Sort Criteria Sort by one of following Date Depth Meanint Maxint Duration UCWI API30 Antec Dry Period C e C e C 2 C C Ascending Descending M Select only in Bathing Season May to Sept Number of Filtered Events 143 r Filter by date Filter by date Stat 03 12 2039 02 00 z End 31 10 2046 05 00 z Figure 4 14 Storm Event Definition Form Now click the Filter Events button and a progress bar and event counter will appear After completion a message box will appear stating Events table successfully generated Click OK and another form is displayed on top of the Control Interface form which shows the events as they are stored in the database At this point you could select different sorting criteria or different storm criteria specifically the rainfall depth mean hourly rainfall intensity and maximum hourly rainfall intensity This changes the order of the events in the database To save your set of events select a destination for the CSV file that will contain statistics of
28. 09 STORMPAC 4 1 User Guide MultiRED files Clear Select STORMPAC databases Define event criteria Select events Create output files Close For catchments with more than one Stormpac database set up at many rain gauge sites multi profile RED files can be produced for use in Info Works catchment modelling Events are defined using either historic or simulated hourly data and RED files created from the disaggregated 5 minute data amp set of Stormpac databases can be selected as long as they each have hourly data covering the same period UCIW 4PI30 have been calculated and a disaggregation has been carried out Create Create Output name UCwl API30 profile profile ly Talysam 2008Jeremy r7 wi Bl E El g Database name D STORMPACTests T alysarn 2008J eremy mdb SELECTING EVENT CRITERIA Having specified all the Stormpac files to be used in analysing for rainfall events you then move on to specifying what will constitute an event All the Stormpac files must have the same start year and duration for the rainfall events otherwise a message to this effect will be produced and you will have to drop or modify the relevant Stormpac databases E 111111111111111 AE MultiRED files Select STORMPAC databases Define event criteria Select events Create output files Close r Event Definition To define event start time both Starting Starting average intensity mm h 02 Average Intens
29. 37 0 8149 0 3351 0 0092 8 0 3078 0 6642 0 4947 0 0575 9 0 2686 0 6790 0 6282 0 0902 10 0 3088 0 7365 0 5831 0 0821 11 0 3649 0 5605 0 5657 0 0462 12 0 2863 0 4616 0 6445 0 0479 C3 2 Simulation tests Using the fitted model 50 years of 5 minute data were simulated for each month The historical and simulated distributions of 5 minute rainfall were compared on quantile plots for each month separately examples for January and July are given in Figures C1 and C2 respectively In a quantile plot points of equal probability of occurrence are plotted providing a good way of comparing two distributions the ideal is all points lying close to the line Some discrepancies can be seen in our plots e g the differences in the highest points Some of these discrepancies could in fairness be attributed to sampling error as the largest points have relatively high standard errors Quantile plots are a rigorous visual comparison of two distributions with a tendency to exemplify differences in the distribution tail i e extreme values so that overall the quantile plots add credence to the use of the fitted model in the disaggregation procedure 87 WRc plc 2009 STORMPAC 4 1 User Guide Quantile Plot for January Series c o E o na od ow N E LO o 9 o V 4 S ae e UU 05S 10 15 20 25 30 historical 5 min rainfall mm Figure C1 Comparison of Historical and Simulated 5 Minute Rainfall Distributions for Januar
30. 998 A Poisson Cluster Model of Rainfall Some High Order Moments and Extreme Values Proceedings of the Royal Society of London Series A Osborne M P 1993 A New Runoff Volume Model WaPUG User Note 28 Natural Environmental Research Council 1975 Flood Studies Report Volumes I and V Press W Flannery B Teukolsky S Vetterling W 1993 Numerical Recipes in C The Art of Scientific Computing Cambridge University Press Ormsbee L 1989 Rainfall disaggregation model for continuous hydrologic modelling Journal of Hydraulic Engineering 115 507 525 Rodriguez lturbe l Cox D R and Isham V 1987 Some models for rainfall based on stochastic point processes Proceedings of the Royal Society of London A 410 269 288 1987 Cowpertwait P S P Lockie T and Davies M D 2004 A stochastic spatial temporal disaggregation model for rainfall Research Letters in the Information and Mathematical Sciences 6 109 123 Glasby C Cooper G and McGechan M 1995 Disaggregation of daily rainfall by conditional simulation from a point process model Journal of Hydrology 165 1 9 Cowpertwait P S P 2005 A stochastic disaggregation procedure based on a Poisson rectangular pulses model Report for WRc January 2005 Institute of Hydrology 1999 Flood Estimation Handbook and CD ROM 73 WRc plc 2009 18 19 20 21 STORMPAC 4 1 User Guide J A Nelder and R Mead 1965 A simplex method for function min
31. AL ME AN Mix UCWI ANTEC ANT VALUES TIM hrs DEFTH INT HRLY DRY IN EVENT trite mm hr INT FED FER HOUR ra hr hra rear Tow 1 2000 1 i 5 2 4 5 1 2 124 1 35 5 dl wl a 2000 1 a 24 g 3 4 4 1 1 139 1 7 a 1 1 B 3 2000 1 7 1 10 5 8 6 1 6 126 1 94 U 4 B 4 2000 1 6 19 5 4 2 B zx 141 1 16 iu 7 1 1 5 2000 1 g 7 7 4 1 B 1 2 162 a 2 xu 25 ET amp 2000 1 9 18 7 5 1 1 2 1 2 176 1 4 1 0 1 3 1 4 7 2000 1 11 16 4 2 0 29 9 153 1 39 z 9 B a 2000 1 13 19 3 1 5 B 133 1 42 lt a 5 4 41 WRc plc 2009 STORMPAC 4 1 User Guide 5 2 Regionalisation and processing of SRG data 5 2 1 Simulated hourly rainfall data Location The SRG will output an hourly rainfall time series for any location within the UK Length of record The SRG can simulate up to 100 years of data It is recommended that at least 20 years of simulated data are used to ensure that the output provides a statistically representative series When working with historical data the maximum length of data series which can be imported is 50 years Format of data The output from the SRG populates the SRG Data table in the working project database The format is the same as that described for historical data in Section 5 1 5 5 2 2 Essential catchment location parameters These are required before the SRG can proceed STORMPAC will direct the user to the site parameters input page if it has not been already populated Once entered the SRG parameter data will be
32. Define Rainfall Events You are now in a position to define some events from your historical hourly database This is achieved by clicking on the Event Definition button on the Control Interface form The Storm Event Definition form shown in Figure 4 14 is displayed STORMPAC identifies each event within the hourly rainfall record that meets the criteria that the user defines on the Storm Event Definition form As described in Section 5 4 you can filter the events from your hourly historical record by choosing event depth mean rainfall intensity or maximum hourly intensity You must specify a ee ISS VSO CERES to identify individual storm events The minimum inter event dry period can range in value from 1 to the maximum number of hours in your hourly historical database Ensure the correct data type is selected at the top of the form Historical or Simulated in this exercise Historical should be selected For the purposes of this exercise you are required to put a value of 1 in the minimum inter box this will identify all storms in your historical database Also because you want to sort by depth you should select the Depth option in the Sort Criteria window and the Descending option which will sort the database by depth in descending order i e the event with the largest depth will be the first record in the database You should also select the checkbox since we require events in the bathing season only for this example 30 WRc plc 20
33. MD values As well as calculating UCWI values STORMPAC also calculates the API30 values required for the new UK Runoff Equation The initial thirty rainfall values are taken directly from the imported rainfall data The method used is to take 30 days of rainfall data from the month preceding the start month of your hourly rainfall series one year For example the program would select 30 days of rainfall data from December 2000 if your hourly rainfall series started on 1 1 00 The program gives the user the option to use evaporation estimates calculated during the SMD calibration procedure or default evaporation parameters of 1 mm day for Winter and 3 mm day for Summer Daily Annual SMD The choice of site to provide the daily rainfall or average annual rainfall and the SMD source records should be as local as possible to the site under study It is not possible to provide a general distance that can be considered local because of the variations in rainfall which can occur due to for example relief Variations may even occur depending on the season Each individual catchment has to be considered on its own To decide look at other available data for the area or consult the Met Office A best fit selection may have to be made and should err on the conservative wetter side 13 WRc plc 2008 STORMPAC 4 1 User Guide 2 4 2 The Disaggregation Model The Disaggregation model requires historical hourly or SRG hourly rainfall
34. Numberofyears 19 Mean hourly rainfall intensity ideis SEE Max hourly rainfall i intensity C Program Files WRc Stormpac 3 2 Minimum inter event f1 dry period hours 111 m Sort Criteria Sort by one of following C s C C C Li C C Date Depth Meanint Maxint Duration UCWI API30 Antec Dry Period Ascending Descending Selectonly in Bathing Season May to Sept Number of Filtered Events 3974 r Filter by date Filter by date Stat 03 12 2039 02 00 x End 31 10 2046 05 00 X Select criteria to sort storms Check the Filter by date button to filter down to a smaller time period using the calendar boxes to enter a start and end date This allows the user to filter down to one rainfall event This may be useful if a typical year of rainfall events has already been selected The events selected here will be available to output in the File Export form Filter Events By clicking this button you will generate an events database that has the storm identification and sort criteria specified on the storm event definition form Another window appears on top of the control interface form allowing the user to view the data Save Events i If you click this button the filtered events will be saved to a CSV file You must specify a file name in the events filename box on this form to which the filtered data will be saved Note you can filter the data as much as you require but the
35. SIMPOL model Data inputs required The data inputs required are as follows e 20 years daily rainfall data from a local site e Average SMD values for each month The source of these data can be found by looking at Section 5 of the main report Setup Create a new database by clicking the Create database button on the Control Interface form Save the database to a directory Import Daily Rainfall After creating the database you must import your daily rainfall Click the Daily Data button on the Control Interface form and open the relevant file A progress bar then appears showing the progress of the data import Once completed a message box appears stating that the daily data has been imported successfully Click OK and you are immediately taken to the Site Information form Site Information Click the Daily data option in the Site Data Type frame Fill in all the blank spaces in the Site Variables frame with the correct site information In the Simulation Data Information frame enter 10 in the Record Duration box and 2000 in the Starting Year box Click OK to calculate the site statistics and return to the Control Interface form SRG Data To simulate rainfall you must first fit the NSRP model parameters to the sample statistics 1 Click the Simulate SRG button to open the Parameter Fitting Interface form 2 Select Fit Parameters 3 Click the Calculate SRG Parameters button a progress bar and counter will appear The opti
36. SRG or process historical hourly rainfall data 2 13 Series length When importing a historical data series into Strompac the maximum length is 50 years However when generating a rainfall series without any historical data the maximum series length which can be created is 100 years 2 1 4 Storm Identification After processing the hourly synthetic or historical rainfall time series STORMPAC can produce a chronological event database Each event is identified and the event depth duration maximum and mean intensities and antecedent dry period are calculated UCWI values and API30 values are also written to this database An event is usually defined as having at least a one hour inter event dry period However STORMPAC will allow the user to specify a minimum inter event dry period of between one hour and the rainfall series length in hours The event database can also be saved to a Comma Separated Values CSV file for use with programs such as Microsoft MS Excel 2 1 5 Sorting The chronological event database can be sorted by STORMPAC in descending or ascending order on seven criteria e Time event number 5 WRc plc 2008 STORMPAC 4 1 User Guide e Total event depth e Mean rainfall intensity e Maximum hourly rainfall intensity e Duration e Dry period e UCWI e API30 In addition the output can be edited to select only events occurring during the months in the bathing season May to September or for a shorter
37. Wer STORMPAC Version 4 1 User Guide STORMPAC USER GUIDE VERSION 4 1 Copyright WRc plc 2009 The content of this manual and the accompanying software are the copyright of WRc plc and all rights are reserved No part of this manual or software may be reported stored in a retrieval system or transmitted in any form or by any means electronic mechanical photocopying recording or otherwise without the prior written consent of WRc plc This manual and the accompanying software are supplied in good faith While WRc has taken all reasonable care to ensure that the product is error free we accept no liability for any damage consequential or otherwise that may be caused by use of either this manual or the software WRc would like to acknowledge the assistance given during the software development by Dr P S P Cowpertwait IIMS Massey University Auckland New Zealand Enquiries to STORMPAC Support Services WRc plc Frankland Road Blagrove Swindon Wiltshire SN5 8YF Tel 44 0 1793 865185 Fax 44 0 1793 865001 Email stormpac wrcplc co uk TABLE OF CONTENTS 1 1 1 1 1 1 1 2 1 1 3 1 1 4 1 1 5 1 1 6 1 2 1 2 1 1 2 2 2 1 2 1 1 2 1 2 2 1 3 2 1 4 2 1 5 2 1 6 2 1 7 2 1 8 2 2 2 2 1 2 2 2 2 3 2 3 1 2 3 2 2 3 3 2 4 2 4 1 2 4 2 THE STORMPAC USER GUIDE STORMPAC Overview Who will use STORMPAC User inputs What is new in STORMPAC 4 1 What was new in STORMPAC 4 0
38. What was new in STORMPAC 3 2 User Guide Contents Objectives Organisation SIMULATION AND PROCESSING OF RAINFALL TIME SERIES What STORMPAC can do Main Purpose Hourly rain data Series length Storm Identification Sorting Selection Disaggregation InfoWorks HydroWorks Rwin SIMPOL When and how the SRG can be used SRG Testing Report Recommendations How the SRG and Disaggregation models work Background The SRG Model The Disaggregation Model Data Requirements The SRG The Disaggregation Model INSTALLATION OF STORMPAC ON YOUR PC Hardware Requirements Software Requirements Installing STORMPAC on a PC Running STORMPAC A SIMPLE WORKED EXAMPLE USING STORMPAC The task _ o oO D ND ON a a ar NN oo oo oo oo o o o wo C 5 3 2 5 3 3 5 3 4 5 3 5 5 3 6 5 3 7 5 4 5 4 1 5 4 2 5 4 3 5 4 4 5 4 5 5 5 5 5 1 5 5 2 Methodology Starting STORMPAC Description of database tables Step 1 Import Historical Hourly Data Step 2 Calculate Site Statistics Step 3 Calculate UCWI and API30 Values Step 4 Define Rainfall Events Step 5 Disaggregate Hourly Data Step 6 Create Output File USER INPUTS NECESSARY FOR STORMPAC Historical hourly rainfall data Source Selection Length of record Format of data Met Office Format of data STORMPAC Missing data Estimated rainfall Leap years Regionalisation and processing of SRG data Simulated hourly rainfall data
39. a 3 23 10 20 5 10 3 a a le 196 51 25 23 33 15 Key Line 1 Site reference number year and 12 monthly rainfall totals in mm Jan Feb Line 2 January daily totals mm Line 3 February daily totals mm dhe dis until next year Line 14 Next year header information site reference etc Missing Trace 29 Feb data Missing data points represented by 999 99999 or 99999 Met Office standard formats will be replaced with zeros However any other representation of missing data will be read directly Therefore the user is advised to check the validity of a daily time series before attempting to import into STORMPAC In addition rainfall falling on 29 February will be retained in the data time series and used in the SRG calculations 46 WRc plc 2009 STORMPAC 4 1 User Guide 5 2 4 Average Annual Rainfall Need The Annually Regionalised version of the SRG requires an average annual rainfall value Source Standard average annual rainfall SAAR data can be obtained directly from the Met Office from annual published books by the Met Office from the Wallingford Procedure Volume 3 or from the Flood Estimation Handbook 5 3 SMD Data 5 3 1 Need SMD data are needed to calculate UCWI values 5 3 2 Source SMD data are available for many sites throughout the UK and may be obtained from the Met Office 5 3 3 Selection Selection of the most appropriate SMD data should be carried out in
40. ainfall total mm the last line in the file has END written in it Note if opening this CSV file in MS Excel you may not be able to open the full CSV file because Excel files are limited to 65536 lines of data 8 A CSV file with the suffix Threshold is created This file contains frequency analysis results of the number of days when the daily totals calculated from the hourly SRG data are above a certain threshold The thresholds are 10 12 14 16 to 40 mm A monthly summary of rainfall depths is also provided If you have imported historical daily data the program will also calculate threshold frequency analysis results and monthly depths from the Daily table in the database This is saved at the end of the Threshold file UCWI API30 Click the button to display the SMD form Select from the Import drop down box Select the relevant monthly SMD file and click to import the data Click on the message box that says SMD data successfully loaded A plot of the imported data black line will appear at the top of the form along with a plot of calculated data red line derived using default values Change the colours on the SMD graph by clicking on the chart labels Next calibrate the SMD model by changing parameters on the Seasonal Oscillation Runoff and Soil Store forms click these buttons to display each of the relevant forms It is recommended to start with the Seasonal Oscillation form Change the parameters on the form a
41. al rules are applied These include using overlapping 5 minute values from the previous disaggregated hour provided the total of these overlapping values do not exceed the 1 hour total to be disaggregated This allows for some influence due to overlapping cells from the previous hour In addition the disaggregated series are scaled to achieve an exact match to the hourly series although clearly this is only a very minor adjustment given they fall to within 5 of the hourly total The disaggregator has had a new option added Accelerated Poisson This option is only recommended if the standard Poisson method is found to be unacceptably slow The standard Poisson method breaks the hourly rainfall into 5 minute intervals by using the following approach do for each hour calculate 12 rainfall estimates from a Poisson distribution repeat until the sum of the 12 is within 5 of the measured SRG value Where the hourly rainfall is large it can take a long time to choose 12 random numbers that happen to be within 5 of that value Usually the Poisson method for typical UK rainfall will take a few minutes to disaggregate the rainfall WRc has had the Poisson method reported as being slow for areas of high rainfall such as the Lake District and areas of North Wales The Poisson method successfully disaggregated the rainfall but took in excess of eight hours to do so The Accelerated Poisson method has been added to Stormpac to cater for this Note that t
42. and Ormsbee Disaggregated 5 Minute Rainfall Distributions for January Series in the Farnborough Data Set C7 REFERENCES Cowpertwait P S P A Poisson cluster model of rainfall high order moments and extreme values Proceedings of the Royal Society of London A 454 885 898 1998 Cowpertwait P S P A continuous stochastic disaggregation model of rainfall for peak flow simulation in urban hydrologic systems Research Letters in the Information and Mathematical Sciences 2 81 88 2001 Cowpertwait P S P Metcalfe A V O Connell P E Mawdsley J A and Threlfall J L Stochastic generation of rainfall time series Foundation for Water Research Report F0217 1991 Cowpertwait P S P Kilsby C and O Connell P A space time Neyman Scott model of rainfall Empirical analysis of extremes Water Resources Research 38 8 1 14 2002 Cowpertwait P S P Lockie T and Davies M D A stochastic spatial temporal disaggregation model for rainfall Research Letters in the Information and Mathematical Sciences 6 109 123 2004 Durrans S Burian S Nix S and Hajji A Polynomial based disaggregation of hourly rainfall for continuous hydrologic simulation Journal of the American Water Resources Association 35 1213 1221 1999 Gao X and Sorooshian S A stochastic precipitation disaggregation scheme for GCM applications Journal of Climate 7 238 247 1994 94 WRc plc 2009 STORMPAC 4 1 User Guide Glasby C Cooper G an
43. are created 6 9 File Export Form The picture below shows the export form as it would look if HydroWorks was selected as the export file type There is the option to include UCWI and API30 values on the profiles and change local arial reduction factor The Export Multi event file option will produce a RED file with all selected events The Export Single Event Files option will produce a file for each event 60 WRc plc 2009 STORMPAC 4 1 User Guide File Export Export File Type Exportfilename Hydrowarks Infoworks red x c Program Files WRc Stormpac 3 2 Dem r Hydroworks Export Format Browse v Include UCWI values Save EH Exit v Export Multi Event File v Export Single Event Files 10 Max No of Single Events to be Exported leave blank for all Local Areal Reduction factor 0 1 1 0 Browse i Clicking the Browse button allows you to choose where to save the exported files to and a file name cave Clicking the Save button saves the data in the chosen format Files are written to the chosen directory In the Max No of Single Events to be Exported box fill in the number of files that you want to produce If this box is left blank all events will be exported as files this may number 1000 files or greater 6 10 Data views form This form has the same functionality as the charts and tables shown after the SRG simulation Summary statistics are generated from the imported daily data Daily
44. asting 1234 8 9 1234 Northing 1234 8 9 5432 Coastal 100 Distance km SAAR mm 100 Ei Figure 5 2 Site Parameters from STORMPAC 5 2 3 Historical daily rainfall data Daily rainfall Either historical daily rainfall data or the average annual rainfall values are necessary to aid regionalisation of the SRG Historical daily rainfall data are discussed here The standard average annual rainfall SAAR value is discussed in the next section 5 2 4 Source Digitised daily data are available for an extensive number of sites covering much of the UK These data are obtainable from the Met Office see Section 5 1 for the address although local sources such as Water Companies or Local Authorities may also have this information 44 WRc plc 2009 STORMPAC 4 1 User Guide Selection The following selection criteria should be considered e Gauge should be local to the study catchment e Length of record available e Periods of missing data e Format of the data Length of record The record length should be between 20 and 30 years to obtain an accurate representation of the local site conditions Format of data STORMPAC has the capability to input daily data in various formats These are XLS DAT and SCF files The current format supplied by the Met Office is an MS Excel XLS format The XLS worksheet has three columns for the date time information Day Month Year or sometimes Year Month Day and a
45. atchment Distance from coast Obtain this value from an OS map STORMPAC needs this value in rounded whole kilometres The range of distances that STORMPAC recognises is shown in Table 5 4 STORMPAC will generate a warning message to inform the user if the site under consideration is outside the allowable range Altitude The average altitude of the study catchment should be obtained from an OS map and should be recorded in metres The allowable range that STORMPAC operates within is given in Table 5 4 STORMPAC will generate a warning message to inform the user if the site under consideration is outside the allowable range It is likely that unreliable results will be produced using the Annually Regionalised SRG for catchments above this range Therefore it is recommended that historical rainfall should be used in this situation Table 5 5 Allowable Ranges for Site Parameters Site Parameter Minimum Value Maximum Value Easting Northing Distance from coast km Altitude m 0600 0642 0 0 6470 8462 145 381 43 WRc plc 2009 STORMPAC 4 1 User Guide Site Information Simulation Data Information Record Duration 33 years Starting year 2000 years Seed 0 to 1 5 Statistics Weight r Site Data Type r Site Variables Senate Record ees 33 Hourly Data tons Daily data Starting year 1970 years C Annually Regionalised As 100 m E
46. ative A or down positive A the y axis by the corresponding amount X shifts the values left negative X or right positive X on the time axis by the corresponding number of days Runoff is calculated and a simple water balance is carried out for each hour i of the rainfall in the historical hourly or SRG hourly database to determine the volume of water stored at the end of the day using the following calculations Runoff i MAX Rain i Rthres 0 Rfactor if Runoff i gt Rain i then Runoff i Rain i where Athres is the Rainfall Threshold mm Rfactor is the Runoff Factor Rain i is the rainfall depth value mm for the current hour 67 WRc plc 2009 STORMPAC 4 1 User Guide Volt i MIN Vmax Vol2 i 1 Rain i Runoff i Vol2 i MAX 0 Volt i MIN 1 Vol1 i Vthres PE i where Vmax is the full soil store mm Vol2 i 1 is the Volume in store 2 in the preceding hour Rain i is the rain depth mm in the current hour Runoff i is the Runoff mm in the current hour 68 WRc plc 2009 STORMPAC 4 1 User Guide 8 FREQUENTLY ASKED QUESTIONS 8 1 Do need Microsoft Access to run STORMPAC 4 1 No STORMPAC 4 1 needs only a reference to the JET Database Engine that is on your machine The necessary files are copied during installation if they are not already on your computer You will however need MS Access 97 or later if you want to look at the database If you are running later ve
47. button to click Section 5 describes all the different formats By clicking this button a progress bar is displayed which shows the progress of the rainfall import A message box appears at the end of the import specifying whether the import was successful or not Click OK to continue Calculate UCw1 8P130 Click this button to display the SMD calculations form See Sections 6 4 and 6 5 Simulate SRG Click this button to display the parameter fitting interface form See Section 6 6 Event Definition Click this button to display the storm event definition form See Section 6 7 Disaggregate Click this button to display the hourly rain disaggregator form See Section 6 8 Create Output File Click this button to display the file export form See Section 6 9 54 WRc plc 2009 STORMPAC 4 1 User Guide 6 3 Site Information Form Site Information C Hourly Date fo E n po l E po By clicking OK on this form statistics are calculated depending upon the data type you specified on the form Checks are also made on all the site information variables to see if they are numeric or within the allowable ranges specified in the program see Table 5 4 If your site is outside these ranges you can still generate rainfall but a process of extrapolation takes place and the program warns you of this 6 4 SRG Parameter Fitting Interface Form 55 WRc plc 2009 STORMPAC 4 1 User Guide
48. c data is recommended in these situations 99 rainfall sites were used in calibration of the annually regionalised model these are shown in Figure 2 1 Figure 2 1 Sites used in calibration of the regionalised model Daily Regionalised The Daily Regionalised version utilises historical daily rainfall data in its regionalisation process and is recommended if 20 or more years of local historical daily rainfall data are available This is particularly relevant if a microclimate is expected at the study site STORMPAC 4 0 uses the Nelder mead method of optimisation This method is often called the Simplex method and does not require derivatives of the optimisation function This has the advantage of making it robust as it will not be affected by any discontinuities in the optimisation function but it is slower than methods that use gradient information For more information see Nelder and Mead 1965 McKinnon 1999 and Mordecai 2003 8 WRc plc 2008 STORMPAC 4 1 User Guide STORMPAC 4 1 gives the option to view parameters relating to the SRG model The upper and lower bounds of parameters Beta Eta Nu Lambda and Shape detailed above have been added to the simulate SRG data screen Once parameters have been calculated against the default upper and lower bounds they will be shown in a table at the bottom of the screen see Figure 2 2 When rainfall has been simulated check the fit using the graphical views and compare
49. calculated data are plotted against the imported SMD data The goodness of fit between the two SMD distributions is also displayed Parameters used in the calculations can be adjusted and a new distribution is generated if the user is not satisfied with the calculated distribution 5 3 7 Format of data simulated Simulated SMD values for the calculation of UCWI values for simulated rainfall are held within STORMPAC during a project run However they are not saved within the project database 5 4 The Event File 5 4 1 Need An event file has several applications It holds summary information of all events from the considered rainfall series as defined by the user s specifications Events are specified in terms of an inter event dry period selected by the user Further filtering in terms of event total rainfall depth maximum intensity and mean intensity can also be performed Events are displayed in tabular form by STORMPAC This enables the user to view the results of their event specification and decide if further modifications are necessary before event data are exported for use in other applications such as HydroWorks In addition to filtering STORMPAC has the facility to sort data by any of the variables displayed in the Events table in either ascending or descending order and also if required exclude data that lies outside the typical bathing season May to September inclusive The Events form can be closed once the user is satisfi
50. ch is currently implemented in STORMPAC A number of approaches could be adopted For example a model based on random cascades similar to that studied by Olsson 1998 and Guntner et al 2001 could be fitted to hourly data and used to disaggregate to 5 minute series Such models incorporate scale invariance i e the statistical properties remain invariant at different time scales subject to a scaling exponent Clearly scale invariance can only apply to particular time scales as for example annual time series always recorded as non zero in the UK would exhibit very different statistical properties to daily series which contain many zeros Scale invariance is therefore at best a good approximation for some time scales Although there is scope for further research into using such models for disaggregation of hourly data in this work we adopt a model that has already been thoroughly researched This is the Poisson rectangular pulses model first studied in detail by Rodriguez lturbe et al 1987 and subsequently extended by Cowpertwait et al 2004 to disaggregated spatial hourly data The methodology we adopt is similar to that used by Glasby et al 1995 in that within storm rain cells have arrival times that occur in a Poisson process However these authors use a Bartlett Lewis process to disaggregate daily data to hourly data whilst our approach is to use a simple Poisson process to simulate fine resolution series directly
51. climate is suspected the annually regionalised data may not compare very well with historical data Of course you have no historical data to compare against We therefore recommend a sensitivity analysis to be carried out 8 4 Pm unsure if my SMD calibration is good enough STORMPAC 4 1 models the daily end of day SMD based upon a model of the potential evapotranspiration You must calculate the daily SMD even if you import daily SMD data When importing daily SMDs the long term daily averages are calculated and you calibrate the calculated daily SMD values to this curve To calibrate the model you must fit your calculated SMD to the imported monthly or daily SMD data A plot of the results is provided as a visual aid along with a fitness factor A good calibration is assumed if the fitness factor is greater than 0 9 However you must ensure that the shape of the calculated SMD curve follows the trend of the imported data for example low SMD values in winter and a peak value in summer 69 WRc plc 2009 STORMPAC 4 1 User Guide 8 5 Should daily or annually regionalise Typically if you have a sufficient length of daily rainfall data for your site 20 years or more you should daily regionalise your data However it may be wise to annually regionalise the data as well You can then compare the daily regionalised and annually regionalised rainfall series with the historical series and choose the most appropriate one Refer to Section 8 3
52. closely match the distributional properties of the 5 minute historical series The historical series are aggregated to 1 hour time steps and disaggregated by selecting matching 1 hour aggregated totals of 5 minute series generated from the PRP model The distribution of 5 minute rainfall given by the disaggregation procedure is compared to the historical distribution using high percentiles extreme values and quantile plots The same procedure is used to compare the Ormsbee method of disaggregation currently implemented in STORMPAC It is found that the new procedure significantly improves upon the Ormsbee method especially in the tail of the distribution C1 INTRODUCTION There are many stochastic models available for disaggregating rainfall For example there are those for downscaling output from global circulation models e g Skaugen 2002 Venugopal et al 1999 and those aimed at producing fine resolution data for urban catchment studies e g Hingray et al 2002 Cowpertwait 2001 Durrans et al 1999 Koutsoyiannis 1994 Ormsbee 1989 In addition there are models for disaggregating daily data to hourly data e g Glasby et al 1995 Koutsoyiannis and Onof 2001 Guntner et al 2001 and models for infilling and disaggregating spatial data e g Cowpertwait et al 2004 In this work we are interested in the disaggregation of hourly data to fine resolution data 5 minute time intervals and seek to improve upon the method by Ormsbee 1989 whi
53. cond iteration limit from 100 000 to 10 000 000 2 4 Data Requirements 2 4 1 The SRG Introduction The data inputs necessary for the SRG are required for the regionalisation of the model and are relatively easy to obtain Full details of the input information and the required formats are given in Section 5 and worked examples of how to run the SRG are in Appendix A Input Essential information for the SRG includes the following e Grid reference 4 figure Easting Northing e Altitude m e Distance from the nearest coast m e Standard Average Annual Rainfall SAAR or local daily rainfall series mm year SAAR The SAAR value can be obtained from the Wallingford Procedure the Flood Estimation Handbook or the Flood Studies Report Alternatively you can request the information directly from the Met Office The advantage of using an average annual rainfall value is that the information is readily available However the disadvantages are that this is an approximation and may not take into account microclimatic variations Also it should be noted that catchments with an average altitude above 381m will typically have a high SAAR value 12 WRc plc 2008 STORMPAC 4 1 User Guide and are outside of the current range of the Annually Regionalised SRG in STORMPAC See Section 5 2 page 36 Local historical daily rainfall data should be used in these situations Daily Rainfall If daily rainfall data are to be used at l
54. d McGechan M Disaggregation of daily rainfall by conditional simulation from a point process model Journal of Hydrology 165 1 9 1995 Guntner A Olsson J Calver A and Gannon B Cascade based disaggregation of continuous rainfall time series the influence of climate Hydrology and Earth System Sciences 5 145 164 2001 Hingray B Monbaron E Jarrar l Favre A and Musy A Stochastic generation and disaggregation of hourly rainfall series for continuous hydrological modelling and flood control reservoir design Water Science and Technology 45 113 119 2002 Koutsoyiannis D A stochastic disaggregation method for design storm and flood synthesis Journal of Hydrology 156 193 225 1994 Koutsoyiannis D and Onof C Rainfall disaggregation using adjusting procedures on a Poisson cluster model Journal of Hydrology 246 109 122 2001 Olsson J Evaluation of a scaling cascade model for temporal rainfall disaggregation Hydrology and Earth System Sciences 2 19 30 1998 Ormsbee L Rainfall disaggregation model for continuous hydrologic modelling Journal of Hyaraulic Engineering 115 507 525 1989 Rodriguez Iturbe Cox D R and Isham V Some models for rainfall based on stochastic point processes Proceedings of the Royal Society of London 410 269 288 1987 Skaugan T A spatial disaggregation procedure for precipitation Hydrological Sciences Journal 47 943 956 2002 Venugopal V Foufoula Georgio
55. daily rainfall data are used to help regionalise the model Availability of these data is relatively widespread and can be obtained from the Met Office At least 20 years of daily rainfall data should be obtained to ensure the data are representative of the site 3 The longer the series generated by the SRG the less likely extreme years will bias the results To some extent this is the same for historical data Twenty years are recommended but not less than ten should be considered for most applications STORMPAC 3 0 and hence versions 4 0 and 4 1 is an improvement on the previous STORMPAC 2 0 in that extremes in data as seen in historical rainfall data series are modelled with greater accuracy The recommendations given for STORMPAC 2 0 therefore hold for all later versions of STORMPAC 2 3 How the SRG and Disaggregation models work 2 3 1 Background The detailed background of the SRG and Disaggregation models incorporated in STORMPAC are not covered in this User Guide Users requiring this information should refer to FWR report FR0O217 Threlfall et al 1998 7 Cowpertwait 1998 and Cowpertwait 2000 A summary of the two models is given below 2 3 2 The SRG Model Regionalisation Parameters have to be regionalised within the model to make the synthetic hourly rainfall series representative of local conditions To achieve this the model requires some local site characteristics and rainfall statistics Further details on the
56. data to be imported into a STORMPAC 4 1 database Data do not have to be produced by STORMPAC to be used in the Disaggregation model 14 WRc plc 2008 STORMPAC 4 1 User Guide 3 INSTALLATION OF STORMPAC ON YOUR PC Introduction This section will help you install and run STORMPAC on your PC Previous versions of STORMPAC required a security dongle to run The current version runs a security check using CopyMinder over the internet You can launch STORMPAC for a limited number of times if not connected to the Internet 3 1 Hardware Requirements To run STORMPAC you will need a PC with the MS Windows operating system It is recommended to be run with an 800 by 600 screen resolution or higher During processing temporary files are sometimes created and deleted It is therefore advisable to have at least 20 MB of free disk space available at all times when running the software 3 2 Software Requirements To run STORMPAC you will need e STORMPAC distribution CD e MS Windows 95 98 NT 2000 XP or Vista e A text editor to create input files and view output files e Optional MS Excel and Access 97 to view files and databases For information on using later versions of Access 97 see Section 8 Frequently asked questions 3 3 Installing STORMPAC on a PC The installation of STORMPAC is easy using the setup routine supplied with the distribution CD Follow the steps listed below e Start Windows e Put STORMPAC dist
57. deviations of 0 129 historical 0 102 Ormsbee method and 0 149 PRP method C6 CONCLUSIONS AND RECOMMENDATIONS Some differences between the historical and PRP distributions were found around the middle part of the distribution tail These differences were of small magnitude and may not be of practical significance This could be verified through some flow simulation experiments e g using a drainage model of a simple sewer network Overall the PRP disaggregator showed a notable improvement upon the Ormsbee method especially in regard to the extreme values Therefore it is recommended as a suitable replacement to the Ormsbee disaggregator 92 WRc plc 2009 STORMPAC 4 1 User Guide Quantile Plot for Pooled Months disaggregated 5 min rainfall mm historical 5 min rainfall mm Figure C6 Comparison of Historical and Ormsbee Disaggregated 5 Minute Rainfall Distributions for all Months in the Farnborough Data Set Quantile Plot for January Series disaggregated 5 min rainfall mm T Fr T T 0 0 0 5 1 0 15 20 25 30 historical 5 min rainfall mm Figure C7 Comparison of Historical and Ormsbee Disaggregated 5 Minute Rainfall Distributions for January Series in the Farnborough Data Set 93 WRc plc 2009 STORMPAC 4 1 User Guide Quantile Plot for July Series disaggregated 5 min rainfall mm historical 5 min rainfall mm Figure C8 Comparison of Historical
58. e Daily table of the current project database The format as with the Hourly table are simply two fields date time information and corresponding rainfall depth data in mm 45 WRc plc 2009 STORMPAC 4 1 User Guide No blanks STORMPAC requires that all fields of each row and column have integer values Therefore do not leave blank fields A message box will be displayed if the daily time series is not continuous Table 5 4 Example of the Daily Rainfall Data Format Supplied by the Met Office 999991235456 1961 906 aa 3882 524 385 201 546 281 536 1479 1 99999 10 196 155 46 a 58 7i 18 2 53 94 54 3 59 25 76 35 3 16 3 56 i 4 25 61 130 20 a 20 15 5 56 45 5 55 25 15 41 a 35 79 36 5 amp 5 10 a 2a 20 51 T a a a7 23 a 5 46 445 5 195 5 51 23 114 406 3 56 3 E 74 51 15 46 5 a 25 61 10 51 41 33 76 23 55 23 20 a 11 35 53 a 20 a a 74 a a 3 36 le a a 36 208 25 76 102 51 102 76 999991235456 1962 1715 547 319 655 453 304 736 1254 1276 357 1 79 36 274 5 226 124 45 2 15 51 36 33 15 3 5 269 a 3 25 16 a a 36 36 3 4 43 130 79 36 a 54 10 a7 25 33 3 a 5 10 a 66 74 a 6 a a 23 10 7 213 a 69 312 a 5 3 15 a 10 liz 10 a 25 51 a6 35 a 35 35 173 41 290 a 10 51 3 11 41 45
59. e checked against historical data if longer series are generated In the absence of historical data it is recommended not to produce series with lengths greater than 30 years This approach should be suitable for the majority of applications 9 WRc plc 2008 STORMPAC 4 1 User Guide 2 3 3 The Disaggregation Model General The disaggregation model within STORMPAC 3 2 has been upgraded from the Ormsbee method Ormsbee 1989 used in previous versions of the software A Poisson rectangular pulses model was first employed within STORMPAC 3 2 This model was first studied in detail by Rodriguez Iturbe et al 1987 and subsequently extended by Cowpertwait et al 2004 to disaggregate spatial hourly data The methodology adopted is similar to that used by Glasby et al 1995 in that within storm rain cells have arrival times that occur in a Poisson process However these authors used a Bartlett Lewis process to disaggregate daily data to hourly data whilst the approach used here is a simple Poisson process to simulate fine resolution series directly In the Poisson rectangular pulses model rain cells have arrival times that occur in a Poisson process Each rain cell has a random lifetime which is distributed as an independent exponential random variable The intensity of each rain cell remains constant throughout the cell lifetime and has been taken to be an independent Weibull random variable The total rain intensity at any point i
60. e gauge site from the catchment e Local rainfall variation consult rainfall maps in Volume 3 of the Wallingford Procedure 5 or the Flood Estimation Handbook e Length of record available see below e Periods of missing data e Format of the data see below e Cost of the data 5 1 3 Length of record The accuracy of the rainfall event statistics will be affected by the record length used This is particularly important when considering extreme events The recommended record length is 10 to 20 years However some design criteria may require longer records to include more extreme events 5 1 4 Format of data Met Office Potentially historical hourly rainfall data may be provided in several different formats depending on whether they are from the Met Office and the age of the data A full explanation of the format is provided with any data supplied by the Met Office 5 1 5 Format of data STORMPAC STORMPAC has the capability to import a variety of hourly file types These are TAB DAT SIM SCF HIS and STM Files are imported into the working project database in the Hourly table This consists of 2 fields the date time data and the rainfall depth data The following describes each of the import file formats 5 1 5 1 TAB and DAT tab de limited files This is a file format for hourly rainfall data used by WRc an example file is included with the installation The files are tab de limited file with 5 columns that co
61. e software will not bring up this window again and instead will go straight to the main menu as shown in Figure 4 4 It should be noted that country mdb is very important to the calculations performed by STORMPAC Therefore take care not to delete it otherwise it will be necessary to reinstall the file from the installation CD 18 WRc plc 2008 STORMPAC 4 1 User Guide Control Interface United Kingdom ee wem D STORMPACKT ests T est mdb Figure 4 3 Control Interface Form At this stage of the exercise the control interface will look like the following form Control Interface United Kingdom v E Sense supe po Figure 4 4 Control Interface Form During Runtime You must now decide to open an existing STORMPAC database or create a new one You will have to create a database during the first use of the software Click the 19 WRc plc 2008 STORMPAC 4 1 User Guide button This will display a Save As dialog box and you can select a directory to save your database For this exercise select the Create database button 4 4 Description of database tables The STORMPAC 4 1 database contains the following tables Daily Disagg EventParam Events This table contains a DateTime field days and a Depth field mm A chronological file of daily data is stored in this table if daily data have been imported This table contains a DateTime field 5 minutes and an Intensity field mm hr
62. east 20 years of data for an adjacent location should be obtained The model calculates relevant statistics from these data for use in its regionalisation process The availability of these data is relatively widespread and the Met Office will be able to inform you of the most suitable sites and the data that are available More details including the required format are given in Section 5 2 3 STORMPAC 3 2 4 0 and 4 1 can import up to 50 years of data SMD UCWI API30 If a calculated UCWI value is required for each event in the output then monthly or daily SMD data can be used STORMPAC calculates SMD in the way described in the Wallingford Procedure Chapter 7 9 An average monthly SMD value is needed 12 values in all for SRG rainfall The average is obtainable from the Met Office see Section 5 3 The actual monthly or daily SMD values should be obtained for the corresponding years and months of the record that is one value for each month or day of the rainfall record if historical hourly rainfall data are being processed If a synthetic rainfall series is used it is important that the UCWI values relate to the pattern of rainfall within this synthetic series Representative synthetic UCWI data can then be calculated However the UCWI data should also be physically realistic and representative of the study catchment SMD data can be calculated using a suitable SMD model that can be calibrated against the 12 long term average historical S
63. eate database button on the Control Interface form Save the database to a directory Site Information Click the button on the Control Interface form to display the Site Information form Select the option in the Site Data Type frame then fill in all the blank spaces in the Site Variables frame with the correct site information Put 20 in the box and 2000 in the Starting year within the Simulation Data Information frame Click OK to calculate the site statistics and return to the Control Interface form SRG Data To simulate rainfall you must first fit the NSRP model parameters to the sample statistics 1 Click the Simulate SRG button to open the Parameter Fitting Interface form 75 WRc plc 2009 STORMPAC 4 1 User Guide 2 Select Fit Parameters 3 Click the Calculate SRG Parameters button a progress bar and counter will appear The optimisation is achieved by the SIMPLEX method see Press et al 4 Click the Simulate Rainfall button to generate 20 years of rainfall The rainfall is generated and then written to the SRGData table in your STORMPAC 4 1 database The number of years generated was specified on the Site Information form 5 Click the Select button and choose a directory and filename for the CSV files 6 Click the Save Rainfall and Thresholds button to create the CSV files T Afile of the generated hourly rainfall values is created Each row in this file has the date and time and a value of hourly r
64. ed with the events defined and filtered from the data STORMPAC automatically copies the event information to a CSV file comma separated variables when the user exits from the Events form The name of this events file is specified by the user on the Events form It should be noted that the Events table in the project database contains all events as defined by the inter event dry period and ignores any filtering performed by the user hence this would have to be repeated if the project is reopened The filtered events also act to define which locations to extract data from the disaggregated time series when writing export files The layout of the Events form is shown in Figure 5 3 48 WRc plc 2009 STORMPAC 4 1 User Guide Storm Event Definition r Data Type m Storm Criteria mm 4 C Historic Simulated v Rainfall depth for event Number of years Mean hourly rainfall intensity Events filename Max hourly rainfall csv file __Select_ z intensity D Program Files Stormpac 3 2 Dem Minimum inter event dry period hours Filter Save Events Events Close m Sort Criteria Sort by one of following C e C C C C C C Date Depth Meanint Maxint Duration UCWI API30 Antec Dry Period Ascending Descending M Select only in Bathing Season May to Sept Number of Filtered Events 144 Filter by date Filter by date Stat 53 12 2033 02 00 End
65. ents Filter Events Disaggregate Save CSV file Create Output File D GS BSE G5 m5 Export To HydroWorks file Figure 4 1 Main steps in worked example of STORMPAC 17 WRc plc 2008 STORMPAC 4 1 User Guide 4 3 Starting STORMPAC Select Start then Programs STORMPAC and STORMPAC 4 1 to run the program Select the splash screen or rainfall icon on the main tool bar as shown in Figure 4 2 This will then take you to the Control Interface form T STORMPAC 0 v6 Click on splash screen or raincloud icon STORMPAC 4 1 ub WRe Internal Serial 0001 ee Wa Frankland Road Blagrove Swindon Wiltshire SN3BYF UK Tel 44 0 1793 865000 Fax 44 0 17938 01 www wrcplc co uk Figure 4 2 Splash Screen The Control Interface form looks like the picture shown in Figure 4 3 although it may not have all the buttons displayed since the number of buttons that are displayed is a function of what data are present in your STORMPAC 4 1 database The Country drop down box is displayed first and this currently defaults to the UK When STORMPAC is run for the first time after installation the initial window will ask you to specify the location of the country mdb file This database contains rainfall parameters specific to the UK It will have been downloaded into the STORMPAC directory during installation STORMPAC will store this location once specified Future runs of th
66. es Figures C7 and C8 Comparing the plots Figures C3 5 for the PRP method with the equivalent obtained from the Ormsbee method Figures C6 8 respectively it is clear that the PRP method improves upon the Ormsbee method It should be noted that this is a fair comparison because the Ormsbee model was also calibrated to the Farnborough data set Cowpertwait et al 1991 91 WRc plc 2009 STORMPAC 4 1 User Guide C5 2 Extreme Values A more detailed analysis of the extremes which are likely to cause overflow problems can be obtained by estimating the high quantiles of the distributions These are shown in Table C2 The PRP method represents the extreme values well overall providing a better fit to the historical series than the Ormsbee method Table C2 The exceptions are the 95 th and 99 th percentiles which are over estimated by the PRP method However the actual differences between the values are approximately 0 04 mm for the 95 th percentile and 0 17 mm for the 99 th percentile which may not be of practical significance Table C2 Extreme values Upper Tail Quantiles Quantile Historical mm Ormsbee Method mm PRP Method mm 0 90000 0 120 0 138 0 120 0 95000 0 200 0 223 0 243 0 99000 0 510 0 469 0 679 0 99900 1 550 1 002 1 697 0 99990 3 486 1 937 3 363 0 99999 6 266 3 292 5 713 1 00000 8 270 4 811 9 686 All three distributions have exactly the same mean 0 0464 mm to three significant figures and have standard
67. events database will not be over written You must change the minimum inter event dry period value which is the underlying value defining an event to overwrite the events database Close Click the Close button to close this form 59 WRc plc 2009 STORMPAC 4 1 User Guide 6 8 Hourly Rain Disaggregator Form File Type Model Enter Random Fos Historic C rmsbee Number Seed 0 1 Poisson C Accelerated Poisson Accelerated Poisson parameter Switch to relaxed rainfall tolerance after 30000 iterations Relaxed relative tolerance i 5 Force normalisation after iteratiot F Use defect carry over rather than normalisation x oa Choose the appropriate data type Historical or SRG Simulated Then make sure the Poisson is selected If this process takes too long several hours select the accelerated Poisson disaggregator and alter the parameters as required For example the Switch to relaxed rainfall tolerance after 30 000 iterations or Force normalisation after 5 000 000 iterations can be reduced to speed up the process See section 2 3 3 for details Click OK to start the disaggregation calculations The process may take some time if there is a long time series to disaggregate The data are then written to the Disaggregator table in the STORMPAC database Note Zeros are excluded from the Disagg table but will be put back into the disaggregated data when export files
68. ey are all in the same location Use Clear to remove all entries The form can be populated and edited directly or by copying data to MS Excel to make changes and then paste back to the form Take care to paste back the header line as well MultiRED files Select STORMPAC databases Define event criteria Select events Create output fles Close Add files Clear For catchments with more than one Stormpac database set up at many rain gauge sites multi profile RED files can be produced for use in Infoworks catchment modelling Events are defined using either historic or simulated hourly data and RED files created from the disaggregated 5 minute data amp set of Stormpac databases can be selected as long as they each have hourly data covering the same period UCIW API30 have been calculated and a disaggregation has been carried out Create Create Output name UCwl API30 profile profile Database name If a chosen database has no SRG data and no hourly data or of course is not a Stormpac database then it will not be added to the list Each database added will automatically default to creating an UCWI and an API30 profile The Soil index value will be set to the value stored within the Stormpac file The user should confirm that the soil index shown is the value used to calculate the API30 data they have already generated A default short name will be taken from the file name 63 WRc plc 20
69. f date and time field 5 Spaces to end of timestep field 5 Spaces to end of number of profiles 20 spaces to end of global UCWI 10 Spaces to end of global ANTEC 5 Spaces to end of Global Evaporation 5 Spaces to end of Wetness Index number not used 10 spaces to Local UCWI 10 Spaces to Local ANTEC 10 Spaces to Local Areal Reduction factor 5 Spaces to Local Evaporation 5 Spaces to Wetness Index Model not used Then from this point on is text 98 WRc plc 2009
70. fourth for the daily rainfall value An example of the correct format is supplied with STORMPAC as Daily xls On some occasions a quality code may also be supplied in an extra column to the right of the rainfall value This code represents the quality of the rainfall value recorded A key for the code values is usually supplied at the end of the rainfall data These quality data are not used by STORMPAC It should be noted the header information on the rainfall files and the different available formats do not present a problem to STORMPAC with respect to successfully reading in the daily rainfall data Another file format that STORMPAC can import that has been used by the Met Office for daily rainfall data has the extension DAT The first line of data shows a reference number of the site the year and then 12 monthly totals of rainfall Line 2 starting with 1 shows the daily rainfall totals in mm for January line 3 starting with 2 daily totals for February etc Line 14 starts with the reference number of the site for the next year An example of this format is shown in Table 5 5 The format of files with the extension SCF has already been described in Section 5 1 5 3 with respect to hourly data However this format may also be used to represent daily time series as the STORMPAC daily import routine also has the facility to read this file type An example of the SCF file format is shown in Table 5 2 STORMPAC stores the imported data in th
71. given and references are provided for further reading 1 2 2 Organisation Sections 1 to 4 of this User Guide provide a general background to STORMPAC including installation instructions and a worked example Sections 5 6 7 and 8 detail the user inputs and formats necessary for all the functions of STORMPAC describing all the button functions relating to each form in the software the SMD model and listing some frequently asked questions Section 9 includes references for further reading on both the background to the models and the UPM programme The Appendices provide worked examples of using the SRG and a flow diagram suggesting a working methodology 3 WRc plc 2008 STORMPAC 4 1 User Guide WRc plc 2008 STORMPAC 4 1 User Guide 2 SIMULATION AND PROCESSING OF RAINFALL TIME SERIES Introduction This section introduces in more detail what the STORMPAC package can do and what it can be used for It also tells you more about how the SRG and Disaggregation models work and the data inputs that are required 2 1 What STORMPAC can do 2 1 1 Main Purpose STORMPAC has been developed specifically for producing suitable rainfall data for input to other models as part of the UPM programme It achieves this by utilising the SRG Disaggregation model and processing software which analyses samples and reformats rainfall data 2 1 2 Hourly rain data STORMPAC can simulate hourly rainfall time series for a site using the
72. he default values can be adjusted The values were selected by comparing a series of accelerated disaggregations of a 25 year dataset to the Poisson disaggregator results These disaggregations each took less than an hour to run rather than running overnight The procedure for the accelerated Poisson method is 1 Use the standard Poisson method up to a first iteration limit The default is 30 000 2 If that limit is reached without the Poisson method succeeding relax the rainfall tolerance from 5 to a higher level 3 Repeat the Poisson method with the relaxed tolerance up to a second iteration limit The default is 5 000 000 11 WRc plc 2008 STORMPAC 4 1 User Guide 4 Atthat point there are then two options e The default is to stop and normalise the rainfall estimate each 5 minute rainfall estimate is multiplied by the ratio of the required hourly value to the summed 5 minute values e An alternative is to continue with the Poisson method but to accept the first value where the sum of the 5 minute estimates exceeds the hourly value The difference between the two is carried over to the next hour as this difference will be negative and will reduce the required total for that following hour Tests showed that some runs with carry over crashed and the one run that completed had very similar results to the run with normalisation Suggested ranges to adjust the limits are e first iteration limit from 10 000 to 100 000 e se
73. hich records the imported SMD data for the catchment This table contains the values used in SMD calibration as well as values used in UCWI and API30 calculations This table contains a DateTime field Hours and a Depth field mm A chronological file of SRG hourly data is stored in this table if hourly SRG data have been generated This table contains a SRGParameter field to store the calculated SRG parameters These parameters are required if you wish to generate hourly SRG data This table contains DateTime Hours UCWI and API30 fields UCWI and API30 values are stored here after they have been calculated The evaporation value at each hourly step is also recorded here 21 WRc plc 2008 STORMPAC 4 1 User Guide Figure 4 5 shows what the database looks like when opened in MS Access 97 EZ Microsoft Access File Edit View Insert Tools Window Help D a l amp 3 X Ra 5 s o Wa 55 i i o 8 Q i Testi Database Gfjopen BE Design ane X 2a EE HE Objects i Tables Queries E Forms Reports Pages Macros Modules Ea Favorites EJ EJ E 6 EJ EJ 6 8 E EL 7 N EN Figure 4 5 STORMPAC 4 1 Database After saving your database you will be presented with the button and the import options of Daily Data Hourly Data and buttons as shown in Figure 4 6 22 WRc plc 2008 STORMPAC 4 1 User Guide control interiace Country ju nited Kingdom x i Create new Open
74. if you have no site rainfall data to compare to your generated series 8 6 What return periods is STORMPAC 4 1 valid for The software is capable of generating return periods equivalent to those seen in the historical data set However accuracy of predicted rainfall compared to historical rainfall is better up to a 1 in 20 return period While the model follows the Gumbel distribution for extreme events beyond a 1 in 20 year return period observed rainfall events tend not to follow a Gumbel distribution 8 7 Developments in STORMPAC 3 2 compared to STORMPAC 2 0 STORMPAC 3 0 saw the enhancement of the rainfall generator used in version 2 0 This is more able to generate extreme events that are observed in historical data Accuracy of the generated rainfall is improved up to 1 in 20 year return period equivalent storms although the generator is capable of generating storms with higher return periods The software was also updated to a more supportable format for users and version 3 1 built on this making the software more user friendly through updates to the interface and allowing users to store and record the parameters and processes used in generating their rainfall and selecting their events The latest version has a new built in Poisson rectangular pulses disaggregator This has been built in as an improvement on the Ormsbee method used in previous versions and allows five minutely data to be generated more accurately particularly with more e
75. imization Computer Journal vol 7 pp 308 313 K I M McKinnon 1999 Convergence of the Nelder Mead simplex method to a non stationary point SIAM J Optimization vol 9 pp148 158 Avriel Mordecai 2003 Nonlinear Programming Analysis and Methods Dover Publishing Cowpertwait P S P 2000 An updated regionalised stochastic rainfall generator for the UK Report for WRc October 2000 74 WRc plc 2009 STORMPAC 4 1 User Guide APPENDIX A WORKED EXAMPLES USING STOCHASTIC RAINFALL GENERATOR Introduction This Appendix provides a step by step guide on two possible applications of STORMPAC The methodology follows that shown in the flow diagram given in Appendix B Section A 3 provides the user with information regarding multiple runs of STORMPAC with different seed values A1 WORKED EXAMPLE 1 Task Generate 20 years of data using average annual rainfall published statistics and produce the largest 50 rainfall events in HydroWorks RED format The largest events will be found by looking at mean storm intensity values There is no local hourly or daily rainfall information Data inputs required The data inputs required are as follows e Average annual rainfall e Grid reference for catchment e Altitude of catchment e Distance from the nearest coast e Average SMD values for each month The source of these data can be found by looking at Section 5 of the main report Setup Create a new database by clicking the Cr
76. in more detail and exported by clicking on the Views button Select To export the simulated rainfall click the Select button to choose a directory and filename for the CSV files Save Rainfall and Thresholds 56 WRc plc 2009 STORMPAC 4 1 User Guide The Save Rainfall and Thresholds button allows you to save your generated rainfall to a CSV file The output file contains hourly generated rainfall totals with a date time stamp One row is equal to one hour A second file will be generated with the suffix Threshold in its filename This file records a summary of daily threshold totals of the generated rainfall A summary of historic hourly rainfall will also be recorded if you have loaded hourly rainfall in your STORMPAC database 6 5 SMD Form Import SMD data Imported SMD Timestep E Runoff Soil Store Select the imported SMD time step from the drop down box and then browse to the location of the SMD file and choose Open A message box appears informing the user whether the import was successful or not After import a plot of calculated and imported SMD values is displayed on the top of the SMD form Click on the chart labels to change the colours if required Seasonal Oscillation Runoff Soil Store Click these buttons to display the seasonal oscillation runoff or soil store forms Each of these forms contains parameters pertaining to the SMD model By selecting different parameters you can calib
77. ing run time Now calibrate the SMD model by changing the model parameters until a good fit is achieved between the calculated and imported SMD data To assess the goodness of fit you need to select the Compare check box This will display the goodness of fit factor in the Fit box Section 7 describes the SMD model in more detail and describes which parameters should be changed With this exercise and future runs of STORMPAC it is recommended to click the Seasonal button first and change the F A or x parameters described in Section 7 since these have the most significant impact on the shape of the SMD curve By clicking on the button the form shown in Figure 4 10 will be displayed Seasonal oscillation data The seasonal oscillation is Em by the equation 2n sin x days o Apply Exit Figure 4 10 Seasonal Oscillation Data Form 26 WRc plc 2008 STORMPAC 4 1 User Guide Try selecting a new set of different values for these parameters and the shape of the plot will change after you click Exit on the Seasonal Oscillation form Click the button again to try different parameters Clicking Exit will change the plot again If the check box is selected the goodness of fit factor will also change with each plot change For the purposes of this exercise input the following parameters F 1 9 A 0 8 X 0 This will give a goodness of fit value of 0 9902 This indicates a very good fit Therefore there is
78. ion calculation is proceeding Once complete the user is returned to the Control Interface form You have now disaggregated the data to five minute values You are now in a position to export the disaggregated data to either HydroWorks or RWIN format files Export 77 WRc plc 2009 STORMPAC 4 1 User Guide To export data to a HydroWorks RED file click the Create Output File button on the Control Interface form You can save data to either a HydroWorks single events file or a HydroWorks multiple events file by selecting the appropriate boxes For this exercise we require the top 50 storms sorted by mean hourly intensity We have already sorted the storms Therefore to select the top 50 storms sorted by mean hourly intensity enter 50 in the Max No of Single Events to be exported box Click Browse to select where the exported files will be saved and to name the files Enter the name Storm in the file name box and click Save Click Save to save the files to the specified directory next click Exit to exit the File Export form and then click STOP on the main menu bar to exit STORMPAC 78 WRc plc 2009 STORMPAC 4 1 User Guide A2 WORKED EXAMPLE 2 Task Generate 10 years of data 20 years of daily data are available for a local site and produce a chronological event file suitable for input to a SIMPOL model Note this worked example is intended to describe the use of STORMPAC and not necessarily producing a suitable input to a
79. ither a HydroWorks single events file or a HydroWorks multiple events file by selecting the appropriate boxes Also you can save the API30 values in the same RED files as the UCWI values if you select the include API30 values checkbox For the purposes of this exercise we require the top 10 storms sorted by depth We have already sorted the storms in descending order To select the top ten you have to enter 10 in the Max No of Single Events to be exported box and select the check box File Export Export File Type Exportfilename Hydroworks Intoworks red x c Program Files WRc Stormpac 3 2 Dem r Hydroworks Export Format Browse v Include UCWI values Save Exit Bul v Export Multi Event File v Export Single Event Files 10 Max No of Single Events to be Exported leave blank for all Local Areal Reduction factor 0 1 10 Figure 4 16 File Export Form Click Browse to select where the exported files will be saved and to name the files The export filenames that are created will contain the event number as well as the filename you specified The filename defaults to the same name as that specified on the Storm Event Definition form For example choosing the filename Storm results in the following filenames for our 10 storms Storm001 RED Storm002 RED Storm010 RED Click Save to save the files to the specified directory next click Exit to exit the File Export form and then click STOP on the main menu ba
80. ity and Min Hourly Intensity are used Minimum hourly intensity mm h 2 S Once the eventis started Starting Inter event dry period h 12 Average Intensity is used to test ifthe following hours are dry Start year 1998 The eventis ended when the length of dry Ed 1898 period is greater or equal to Inter Event C Use SRG data ES Use historic data v Output hourly averages Two parameters are used together to indicate that an event has begun First the average rainfall intensity across all sites must exceed the specified and secondly the minimum rainfall for at least one site must exceed the specified intensity mm hr Once an event has started the value is used to test if the following hours are dry The event is ended when the length of the dry period is greater than or equal to the specified In addition the events can be created using either SRG or hourly data If there is no data for your chosen option of SRG or hourly you will be informed that you should change the choice SELECTING EVENTS Proceeding to select events there will be a pause while the events are being identified Following identification the events will be displayed on two grids The default is for all events to be selected but any event can be excluded if required The screen displays two grids The upper grid displays summary parameters for the event and allows you to include or exclude events by clicking on the Include event tick box
81. misation is achieved by the SIMPLEX method see Press et al 4 Click the Simulate Rainfall button to generate 10 years of rainfall The rainfall is generated and then written to the SRGData table in your STORMPAC 4 1 database The number of years generated was specified on the Site Information form 79 WRc plc 2009 STORMPAC 4 1 User Guide 5 Click the Select button and choose a directory and filename for the CSV files 6 Click the Save Rainfall and Thresholds button to create the CSV files 7 A file of the generated hourly rainfall values is created Each row in this file has the date and time and a value of hourly rainfall total mm the last line in the file has END written in it Note if opening this CSV file in MS Excel you may not be able to open the full CSV file because Excel files are limited to 65536 lines of data 8 A CSV file with the suffix Threshold is created This file contains frequency analysis results of the number of days when the daily totals calculated from the hourly SRG data are above a certain threshold The thresholds are 10 12 14 16 to 40 mm A monthly summary of rainfall depths is also provided If you have imported historical daily data the program will also calculate threshold frequency analysis results and monthly depths from the Daily table in the database This is saved at the end of the Threshold file UCWI API30 Follow the advice given in the UCWI and API30 section in Appe
82. n time is the sum of the intensities of all cells alive at that point The model has been fitted using a minimisation procedure which matches model properties to their equivalent within the sample data and is based on a squared differences methodology as devised by Cowpertwait et al 1991 2004 14 Model Validity Using the fitted model and these model properties estimates of the variables above were made for each calendar month from over thirty years of UK five 5 minute data Following Cowpertwait et al 2004 only wet hourly sequences are included in the estimation of these sample properties Fifty years of 5 minute data were simulated for each month using the fitted model The historical and simulated distributions of 5 minute rainfall were compared on quantile plots for each month separately and showed good correlation Cowpertwait 2005 9 Cowpertwait et al 2004 had found that the spatial Poisson rectangular pulses model performed well when disaggregating hourly data for input in urban catchment models justifying its use in STORMPAC for temporal disaggregation as this is just a special case of the spatial Poisson rectangular pulses model The good correlation reached between historic and simulated 5 minute data Cowpertwait 2005 9 adds credence to the use of the fitted model in the disaggregation procedure Overall the fitted Poisson rectangular pulses disaggregator also showed a notable improvement upon the Ormsbee method
83. nd click Apply to see the changes take effect Select the Compare box to determine what the goodness of fit is A value greater than 0 9 is sufficient If a value of 0 9 or greater is not achievable by changing the Seasonal Oscillation parameters it is recommended to try more than one change of the parameters then click the Runoff button on the SMD form and change the parameters on this form Again clicking Apply will allow the changes to take effect Finally click the Soil Store button on the SMD form if you have still not managed to achieve a suitable calibration Note a suitable fitness factor is usually achieved after changing the Seasonal Oscillation parameters and the Runoff parameters To calculate UCWI and API30 values click the Calculate UCWI APIS0 button on the SMD form Select the soil class from the drop down box and choose the relevant evaporation checkbox Click to calculate the values Select the Yes button in the message box Do you wish to continue 76 WRc plc 2009 STORMPAC 4 1 User Guide Select OK in the message box stating that UCWI API30 calculations are complete then press the Exit button to close the form Events You are now in a position to define some events from your simulated rainfall data This is achieved by clicking on the Event Definition button on the Control Interface form 1 For the purposes of this exercise put a value of 1 in the Minimum inter event period box to identify all storms in yo
84. ndix A Worked Example 1 Events You are now in a position to define some events from your simulated rainfall data This is achieved by clicking on the Event Definition button on the Control Interface form 1 For the purposes of this exercise put a value of 1 in the Minimum inter event period box to identify all storms in your historical database 2 Select the Date button in the Sort Criteria window to sort the database by date 3 Select the Ascending option in the Sort Criteria window which will sort the database in ascending order 4 Select a directory and enter a file name by clicking the Select button The generated events will be saved to this file 5 Select the Simulated data type option at the top of the form This ensures that the program will read data from the SRG data table in the database 6 Click Filter Events and a progress bar and event counter will appear 7 Click OK to the message box Events table successfully generated 8 The events as they are stored in the database will be displayed on the Control Interface form Click on the button 9 Click Save Events to save the events as a CSV file 10 Click Close to return to the Control Interface form Disaggregation To disaggregate the data you have to click the Disaggregate button on the Control Interface form This displays the Hourly Rain Disaggregator form 80 WRc plc 2009 STORMPAC 4 1 User Guide Select the SRG option to make sure the SRG hourly
85. ng a working methodology for using STORMPAC Databases with tables Hourly SRGData Disagg UCWI and API30 Storm Criteria Calculate Site Statistics select Yes data available s hourly historic data Is result satisfactory Daily Data vailable No Calculate Site 5 Statistics select Simulate SRG Annually Regionalised Compare historic and simulated No Simulate SRG 4 data Yes Change SEED value A r v Multi RED analysis nent Calculate UCWI series APG Storm Filter Storm Criteria Events i Disaggregate to 5 RED files with many profiles 5 minute minute data data series Hydroworks RED files RWIN files Expoitdatg SIMPOL 2 0 formatted files WRc plc 2009 84 STORMPAC 4 1 User Guide WRc plc 2009 STORMPAC 4 1 User Guide APPENDIX C REPORT FOR WRc ON A STOCHASTIC DISAGGREGATION PROCEDURE BASED ONA POISSON RECTANGULAR PULSES MODEL Paul S P Cowpertwait January 2005 SUMMARY A Poisson rectangular pulses PRP model is fitted to wet sequences of 5 minute data extracted from a 31 year rainfall record taken from a gauge in Farnborough UK A comparison of simulated and historical quantiles verifies that the model is able to
86. no need to change any of the other parameters However click on the Runoff button and change the parameters if you wish this form is shown in Figure 4 11 Runoff occurs if the rainfall is above the given threshold The runoff factor is 4 simple multiplier Rainfall Threshold mm 5 Runoff factor 0 3 Apply Exit Figure 4 11 Runoff Data Form Again try selecting different values for the Rainfall Threshold and Runoff Factor and clicking Exit will cause the plot and goodness of fit to change The final way to change the calibration is to click the Soil Store button This will display the form shown in Figure 4 12 27 WRc plc 2008 STORMPAC 4 1 User Guide Soil store data Full soil store mm CS Soil store threshold mm 5 00 Initial SMD mm NN Initial depth in soil store 2 Cam fmm Apply Exit Figure 4 12 Soil Store Data Form It is usual to achieve a good calibration by changing the parameters on the Seasonal Oscillation form only However you can fine tune the calibration with the parameters on the Soil Store data form Typically achieving a goodness of fit value more than 0 9 signifies a good calibration However the sensitivity of the calculated UCWIs and API30s to the fit of the SMD curve is not that critical therefore do not worry if your fitness value is less than 0 9 To calculate UCWI and API30 values click the Calculate UCWI API30 button on the SMD Calculations form and
87. ntain the following data Column 1 Year Column 2 Month Column 3 Day 36 WRc plc 2008 STORMPAC 4 1 User Guide Column 4 Hour Column 5 Rainfall intensity mm hr 5 1 5 2 SIM files This is the common format for hourly rainfall files used in STORMPAC see Table 5 1 This format has one row per 24 hours containing the year month and day followed by two more rows containing the hourly rainfall depths 12 hours on each row It is essential that the historical hourly data are input exactly in the following format for STORMPAC to read these data correctly Line 1 YEAR MONTH DAY Format 10 spaces four digits for the YEAR e g 1999 5 spaces 2 digits for the MONTH e g 08 5 spaces 2 digits for the DAY e g 04 Line 2 HOUR 1 HOUR 2 HOUR 3 HOUR 12 Line 3 HOUR 13 HOUR 24 Format x e real number with one decimal place dp 3 spaces real number with p Each real number must take up 4 spaces e g 12 4 02 4 but the leading zero may be omitted e g lt space gt 2 4 These files may represent either historical data or data from the SRG in STORMPAC 5 1 5 3 SCF files The first three rows give information about the gauging station name and location Row 4 contains column headings Rainfall data follows with one data point per row At the end of the file definitions of Quality values given An example of this format can be seen in Table 5 2 Line 1 Gauge reference Line 2 Met Office Gauge
88. nuous basis and not per event This is achieved by calibrating a SMD model with the long term average SMD data obtained for your study site Click the Calculate UCWVAPI30 button on the Control Interface form to display the SMD Calculations form The form is shown in Figure 4 8 SMD Calculations Click the drop down box Import SMD data Imported SMD Timestep jig Seasonal Oscillation Daily oie vi Calculate Monthl UCWIZAPI30 Runoff Exit Soil Store Figure 4 8 SMD Calculations Form First determine the time step of your SMD data by selecting 4 or Monthly from the Import SMD data drop down box For this exercise select Find the example SMD XLS file using the Open Import File dialog box select the file and press Open The data will then be imported Click OK on the message box that appears at the end of the import The monthly long term average SMD data are then displayed on the form along with the SMD values calculated using default parameters The SMD Calculations form should now look like Figure 4 9 Change the colours on the SMD graph by clicking on the chart labels 25 WRc plc 2008 STORMPAC 4 1 User Guide SMD Calculations SMD mm Imported SMD mm 96 Calculated SMD mm Jan Feb Mar Apr May Jun Jul Aug Sep Oct NovDec ers 0 25272 Seasonal Oscillation Calculate UCWI API30 Runoff Import SMD data Monthly La Exit Soil Store Figure 4 9 SMD Calculations Form dur
89. o achieve an exact match to the hourly series although clearly this is only a very minor adjustment given they fall to within 0 05 mm of the hourly total It should be noted that the disaggregation procedure is not accounted for when fitting the model it is a separate procedure that could be applied to any model capable of simulating 5 minute series Therefore it is likely that some probably small bias would be introduced into the distributional properties of the resultant disaggregated series C4 2 Tests The Farnborough data were aggregated to hourly values and then disaggregated using the above procedure The distribution of the resultant disaggregated series was compared to the historical 5 minute distribution using quantile plots the results are shown in Figures C3 5 Figure C3 gives the overall result where the data for all months have been pooled From this figure it can be seen that the distribution of the disaggregated series compares favourably to the distribution of the historical 5 minute series The exception is a slight curvature in the mid to lower part of the plot which would translate into a slight over estimation in part of the distribution tail The plots for January and July show a similar pattern i e a slight over estimation for some of the distribution tail Figures C4 and C5 89 WRc plc 2009 STORMPAC 4 1 User Guide Quantile Plot for Pooled Months disaggregated 5 min rainfall mm his
90. r to exit STORMPAC 33 WRc plc 2008 34 STORMPAC 4 1 User Guide WRc plc 2008 STORMPAC 4 1 User Guide 5 USER INPUTS NECESSARY FOR STORMPAC Introduction This section describes all of the different types of input necessary for STORMPAC The data inputs are divided into the following separate sub sections 5 1 Historical hourly rainfall processing 5 2 Regionalisation and processing of SRG data 5 3 Soil moisture deficit data 5 4 The event file 5 5 WASSP PCD files The likely source of these data and the formats necessary for STORMPAC to import them are described 5 1 Historical hourly rainfall data 5 1 1 Source Digitised hourly data are available for approximately 200 UK sites and may be obtained on disk from The Meteorological Office FitzRoy Road Exeter Devon EX1 3PB United Kingdom Tel 0870 900 0100 or 44 0 1392 885680 from outside the UK Fax 0870 900 5050 or 44 0 1392 885681 from outside the UK website www metoffice gov uk e mail enquiries metoffice gov uk Other local sources for data include for example Water Utilities Local Authorities The Environment Agency in England and Wales SEPA in Scotland and EHS in Northern Ireland 35 WRc plc 2008 STORMPAC 4 1 User Guide 5 1 2 Selection Selection of the most appropriate gauge site should if appropriate be carried out in consultation with the Met Office The following factors should be borne in mind e Distance of th
91. rate the model to fit the imported SMD data Change the colours on the SMD graph by clicking on the chart labels 57 WRc plc 2009 STORMPAC 4 1 User Guide Calculate UCWIAAPI3O Click this button to display the UCWI API30 calculation form See Section 6 5 You should only click this button if you have achieved a good SMD calibration i e goodness of fit greater than 0 9 6 6 UCWI API30 Calculation Form UCWI API30 calculation data Rain mm UCWI 4 R4 mm APLS 3 R3 2 R2 1 R1 Day Number UCWI input data API30 input data Rainfall depths mm Initial Day 1 o Soil Class W sDey2 p k mudey x Initial Day 3 fo p C Calculated EVAP Initial Day 4 n Default EVAP Winter Imm day Initial Day 5 pooo Summer 3mm day SMD atend of day mm Calculate UCWI API30 Calculate UCWIAAPI3O Click this button after you have filled in the necessary information on the UCWI form see Section 4 7 and Figure 4 13 Then the program calculates UCWIs and plots rainfall depth and UCWI values This plot is a visual aid to allow the user to determine if the first few days of UCWI calculations are acceptable For example that they contain no negative values 58 WRc plc 2009 STORMPAC 4 1 User Guide 6 7 Storm Event Definition Form Storm Event Definition Data Type Storm Criteria mm C Historic Simulated v Rainfall depth 1 for event
92. ribution CD into CD drive e Browse to select your CD Drive and select setup then choose Open Clicking OK opens the STORMPAC installation e Follow the simple instructions to complete the setup Note The default drive and directory is C Program Files WRc STORMPAC You must change this if you do not want the install files to be put in this default directory 15 WRc plc 2008 STORMPAC 4 1 User Guide e At the end of the Installation you will be prompted to Restart your computer You should do this if you wish to use STORMPAC immediately 3 4 Running STORMPAC Once the software has been installed run STORMPAC by selecting Start then Programs STORMPAC and STORMPAC 4 1 16 WRc plc 2008 STORMPAC 4 1 User Guide 4 ASIMPLE WORKED EXAMPLE USING STORMPAC Introduction Once you have installed STORMPAC onto your machine you will want to see how it works This section takes you through a relatively straightforward example and in so doing will also introduce you to many of the features in the package Further examples are given in Appendix A 4 1 The task To process one year of hourly historical rainfall data and obtain the largest ten rainfall events by depth in the bathing season in a format suitable for input to HydroWorks 4 2 Methodology The main steps to be taken are shown and summarised below q Historical Hourly Rainfall Import into Database Calculate Site Statistics Calculate UCWI API30 Define Ev
93. rsions than Access 97 you will be asked if you want to update the database Reply no as you will not be able to write to the database and you will not be able to run STORMPAC with that database Useful files are also saved to disk at various stages of STORMPAC giving the user the option to read these using a spreadsheet such as MS Excel or a text editor such as Wordpad 8 2 What is the minimum length of daily record should import It is recommended that a minimum of 20 years daily data should be imported However weighted averages of the regression estimates and site estimates taken from the site data can be calculated if your historical series is relatively short say less than 10 years Remember to select the checkbox on the site information The weight checkbox is for users who want to combine historical data with the regionalised model For example if a user has five years of daily data these could be combined with the regionalised model to give weighted estimates of the statistics needed to fit the Neyman Scott Rectangular pulses NSRP model which underpins the SRG algorithm This is a half way house between those who want to fit the model using a good record of daily data and those who have no daily data 8 3 What should do if have no data at my site to compare values In these circumstances the normal procedure is to annually regionalise the data based on site variables However in mountainous areas or areas where a micro
94. s identified within those procedures However the long rainfall time series which are produced can be used for other applications 1 1 3 User inputs The SRG only needs a few inputs from the user to produce hourly rainfall data Essential inputs are grid reference distance from coast altitude and some local rainfall statistics calculated from actual daily rainfall data or from published tables and are obtainable from the Meteorological Office Met Office Contact Wallingford Software for details of the Hydroworks and Infoworks products Stormpac generates Rainfall RED files that are compatible with these applications 1 WRc plc 2008 STORMPAC 4 1 User Guide Historical hourly data when available are usually obtained from the Met Office however the data may need to be reformatted so that STORMPAC can read it Soil Moisture Deficit SMD data monthly average or daily values are necessary if representative UCWI values are required API30 values are also calculated and both UCWI and API30 are output to the same HydroWorks files The remaining inputs to the various sections of STORMPAC are all created internally by the software However files created externally to STORMPAC can be read by the package if they are in the correct format 1 1 4 What is new in STORMPAC 4 1 1 Rainfall series of up to 100 years can now be generated Calculated SRG parameters and upper and lower parameter limits can be viewed Graphical view
95. s automatically displayed after importing the historical hourly rainfall data You must fill all the boxes with the correct site information This is important for the regionalisation element of the software Select the correct Site Data Type to run the correct regionalisation routine i e Annual or Daily or it could be that you wish to import some historical hourly data For this exercise select Hourly Data and fill in the form as illustrated in Figure 4 7 Then click OK to start the site statistics calculations and click OK on the message box that appears at the end of the calculations Clicking the Site Statistics button at any time during a STORMPAC session displays the Site Information form Site Information Site Data Type Site Variables r Simulation Data Information Record Duration 33 AUS Record cuen 33 Hourly Data pee years Starting year 2000 Daily data Starting year 1970 ican years Seed Oto 1 C Annually Regionalised Altitude 100 ud 5 fm r Statistics 3j Easting 1234 e g 1234 Weight Northing 1234 e g 5432 Coastal 100 Distance km SAAR mm 100 Exit Figure 4 7 Site Information Form 24 WRc plc 2008 STORMPAC 4 1 User Guide 4 7 Step 3 Calculate UCWI and API30 Values To calculate UCWI and API30 values in STORMPAC it is first necessary to calculate the SMD The UCWI values are calculated on a conti
96. s button to run the program Alternatively click the splash screen This will display the Control Interface form a Click the Stop button to terminate the program and unload it from memory Click the question mark button to display the HTML help system Create NEw i database Click the Create database button to create a new STORMPAC 4 1 database and save it by selecting a directory and entering a filename in the Save As dialog box Click this button to open an existing STORMPAC 4 1 database Browse through directories until you select the correct one On opening a database a message box will appear reminding the user of what tables have been created in the database 53 WRc plc 2009 STORMPAC 4 1 User Guide Site Statics You must click this button at some point during the STORMPAC run even if you open an existing database that already contains site information data By clicking the button the site information form is displayed You must fill in the relevant information making sure to select the correct data type Hourly Daily Annually Different statistics are calculated for each data type Also remember to change the record duration for the simulated data The default is 20 but you may wish to put in a different value see Section 6 3 Disaggregated Daily Data ps Hourly Data Click one of these buttons to import rainfall data The format and time step of your rainfall data should determine which
97. s comparing simulated results with historical data are produced Updates to Data Views function Create output file function for multi site RED generation speeded up 1 5 What was new in STORMPAC 4 0 RED file output includes evaporation on the first profile used in Wallingford runoff model Options added to speed up Poisson disaggregator Chart display to compare summary statistics for imported daily rainfall and generated rainfall Time filter on Event definition form Multi site RED generator available this allows the user to generate RED files with many rain profiles 1 6 What was new in STORMPAC 3 2 e Imports up to 50 years of rainfall data Output files are exported to the directory and folder selected by the user e Updates to the user interface The STORMPAC database will record Criteria used to select rainfall events 2 WRc plc 2008 STORMPAC 4 1 User Guide Parameters used in the SMD calibration and UCWI calculations Imported SMD data e A new Poisson rectangular pulses dissaggregator this is an improvement on the previous Ormsbee method disaggregator particularly for extreme rainfall values see Section 2 3 3 for further details 1 2 User Guide Contents 1 2 1 Objectives This guide is designed to tell you what STORMPAC can do and how to install and use the software It is aimed specifically at the first time or the occasional user Brief details about the SRG and the Disaggregation models are
98. scrutinise the file at the point where a missing data value occurs and then decide if the insertion of a 0 0 will affect the event selection For example if a series of missing data values occur because of a gauge failure lasting for several weeks a series of zero values will only distort the total number of events and not any individual storm However if a missing data value occurs in the middle of a storm and a zero value is inserted STORMPAC will interpret this as two events if the minimum inter event dry period is set at 1 In this situation it may be necessary to omit the whole event i e edit the rainfall to zeros Clearly the final decision will depend on what the data are to be used for However you should guard against producing a rainfall file which could be misinterpreted by future users 5 1 7 Estimated rainfall The Met Office show estimated rainfall as negative values These should be edited to show either true rainfall or zero rainfall Again careful scrutiny of the data should be made before editing is carried out 5 1 8 Leap years STORMPAC does not require any editing of rainfall data with respect to leap years 38 WRc plc 2008 STORMPAC 4 1 User Guide Table 5 1 Example of SIM and HIS Formats for Hourly Rainfall Data 2000 1 1 0 0 00 02 12 23 22 17 08 01 0 0 0 0 0 0 0 0 0 0 00 00 00 00 00 00 00 00 0 0 0 0 2000 1 2 0 0 00 00 00 00 00 00 00 00 00 0 0 0 0 0 0 00 00 00 00 00 00 00 00 00 02 2 0 2000 1 3 2 0
99. ta to populate the Hourly table in the project database This facility enables STORMPAC to reanalyse disaggregated data and redefine storm events The hourly time series obtained from this import routine has to assume zero rainfall for all times outside the events described in the PCD file This is necessary to ensure that a continuous time series of rainfall is generated However the user should be aware that this may not be as good a representation for a region as would be obtained from reanalysing the initial historical hourly or daily rainfall data 5 5 2 Format An example of the WASSP PCD format is shown in Table 5 7 The first three lines in the file contain the header information about the site where the data originates Each event is then described separately For each event the first 3 rows represent summary information about the storm The disaggregated rainfall data are then displayed in rows of 10 equally spaced data entries The PCD file must be in chronological order 50 WRc plc 2009 STORMPAC 4 1 User Guide Table 5 6 WASSP Data Format WRe WASSP Data Date 12 16 99 Time 13 34 09 Storm Location leixlip Total Number of Events 300 300 highest depth storms from 20 year SRG at Leixlip Ireland 1 EVENT 1 23 00 2 1 0 96 VALUES UCWI 126 0 30 126 1 96 300 720 600 0 512 0 512 0 000 o 000 0 206 0 206 0 256 o 000 o 000 0 000 0 000 0 285 0 564 1 725 1 440 1 152 1 725 2 016 2 526 2 616 3 6850 2 016 2 304 0 564 1 154
100. the filtered and sorted events by clicking the Select button and if required enter a new file name in the Save Event Data As dialog box To produce the CSV file click If you wish to change the number of events saved in the database you have to select a different minimum inter event dry period You are warned before you do this so that you cannot overwrite the full list of events in error Click Close to return to the Control Interface We have now reached step 5 4 9 Step 5 Disaggregate Hourly Data To disaggregate the data you have to click the Disaggregate button on the Control Interface form The Hourly Rain Disaggregator form as shown in Figure 4 15 will be displayed 31 WRc plc 2008 STORMPAC 4 1 User Guide Hourly Rain Disaggregator File Type Model Enter Random 0 5 Historic C Ormsbee Number Seed 0 1 SRG Poisson Accelerated Poisson Accelerated Poisson parameters Switch to relaxed rainfall tolerance after nnn iterations Relaxed relative tolerance 5 Force normalisation after 5000000 iterations J Use defect carry over rather than normalisation x oa Figure 4 15 Hourly Rain Disaggregator Form Select the Historical option to make sure the hourly historical rainfall is disaggregated If simulated rainfall data was to be disaggregated the SRG option should be selected Also select the Poisson option to use the new Poisson rectangular pulses dissaggregator This should be
101. the UCWI API30 calculation data form as shown in Figure 4 13 is displayed 28 WRc plc 2008 STORMPAC 4 1 User Guide Rain mm UCWI 4 R4 mm APO 3 R3 2 R2 1 R1 pis A Drop down box for choosing soil class UCW input data API30 input data Rainfall depths mm Initial Day 1 n Soil Class Initial Day 2 o LK 0 1 day SMD at end of day mm Initial Day 3 fo CE C Calculated EVAP Initial Day 4 n Default EVAP Winter Imm day Initial Day 5 o Summer 3mm day Calculate Exit UCWIAPBO i Figure 4 13 UCWI API30 Calculation Data Form STORMPAC calculates UCWI as described in the Wallingford Procedure Chapter 7 9 9 UCWI 125 8 API5 SMD Therefore as API5 depends on the previous 5 days rainfall you are required to input the 5 daily rainfall values prior to the start time of your hourly rainfall data in the boxes marked Initial to This represents the first day to the fifth day before the start of your hourly rainfall data respectively You are also required to enter the SMD at the end of the day in the box We suggest that you use the following values INITIAL DAY 1 0 INITIAL DAY 2 0 INITIAL DAY 3 0 INITIAL DAY 4 0 INITIAL DAY 5 0 SMD AT END OF DAY 0 API30 values are calculated using the method described in WaPUG User Note 28 You must choose the soil class from the drop down box and then decide whether you want to use default evaporation parameters of
102. torical 5 min rainfall mm Figure C3 Comparison of Historical and PRP Disaggregated 5 Minute Rainfall Distributions for all Months in the Farnborough Data Set Quantile Plot for January Series disaggregated 5 min rainfall mm T T T T T T T 0 0 0 5 1 0 15 20 25 3 0 historical 5 min rainfall mm Figure C4 Comparison of Historical and PRP Disaggregated 5 Minute Rainfall Distributions for January Series in the Farnborough Data Set 90 WRc plc 2009 STORMPAC 4 1 User Guide Quantile Plot for July Series disaggregated 5 min rainfall mm historical 5 min rainfall mm Figure C5 Comparison of Historical and PRP Disaggregated 5 minute Rainfall Distributions for July Series in the Farnborough Data Set C5 COMPARISON WITH THE ORMSBEE DISAGGREGATION METHOD C5 1 Quantile plots The aggregated 1 hour Farnborough data were disaggregated using the Ormsbee method currently implemented in STORMPAC The distribution of the resultant disaggregated series was compared to the historical 5 minute distribution using quantile plots Figures C6 8 Figure C6 gives the overall result where the data for all months have been pooled From this figure it can be seen that the Ormsbee method produces a consistent under estimation in the distribution tail i e the Ormsbee method fails to produce sufficient extreme values The plots for January and July show a similar pattern i e a consistent under estimation of extreme valu
103. u E and Sapozhnikov V A space time downscaling model for rainfall Journal of Geophysical Research 104 D4 19705 19721 1999 95 WRc plc 2009 96 STORMPAC 4 1 User Guide WRc plc 2009 STORMPAC 4 1 User Guide APPENDIX D RED FILES FORMAT IMPORTANT RED files are right justified and the spacing between numbers is vital For example if a local UCWI value is 99 it would start 8 spaces from the left of the profile properties line However it is 250 then it would start 7 spaces from the left of the same line If there are missing spaces etc then this will cause errors when importing to InfoWorks Therefore the format shown on next page must be maintained Location 1 ddmmyyyyhhmmss sss p GU GA GE W LU LA LR LE W EVENT 138 17 00 26 6 85 36 Sss timestep in seconds P no of profile GU Global UCWI GA Global ANTEC GE Global Evaporation W Wetness Index LU Local UCWI LA Local ANTEC LR Local Areal Reduction factor LE Local Evaporation NB In InfoWorks If global value 0 then local value is used Event description is limited to 40 letters 97 WRc plc 2009 STORMPAC 4 1 User Guide Location 1 24062000160000 300 2 5 990 1 0 0 5 99 0 1 00 1 0 ONewUK API 1 99 99 0 1 00 0 0 OWallingford UCWI 2 Location m ddmmyyyyhhmmss sss p GU GA GE 5 Spaces to end of first integer 15 Spaces to end o
104. ur historical database 2 Select the Mean hourly rainfall intensity option in the Storm Criteria window 3 Select the Depth button in the Sort Criteria window to sort the database by maximum intensity 4 Select the Descending option in the Sort Criteria window which will sort the database in descending order i e the event with the largest depth will be the first record in the database 5 Select a directory and choose a file name and directory by clicking the Select button The generated events will be saved to this file 6 Select the Simulated data type option at the top of the form This ensures that the program will read data from the SRGdata table in the database 7 Click Filter Events and a progress bar and event counter will appear 8 Click OK to the message box Events table successfully generated 9 The events will be displayed on the Control Interface form as they are stored in the database Click on the Close button 10 Click Save Events to save the events as a CSV file 11 Click Close to return to the Control Interface form Disaggregation To disaggregate the data click the Disaggregate button on the Control Interface form This displays the Hourly Rain Disaggregator form Select the SRG option to make sure the SRG hourly rainfall is disaggregated and make sure the Poisson option is selected to use the Poisson disaggregator Then click OK A progress bar will appear informing the user that the disaggregat
105. xtreme values New functions added to the software over this time include e Calculation of API30 and as well as UCWI values e Greater than 9 hours inter event dry period Systems with long drain down times can now be studied e Continuous disaggregation e More than 300 storms can be analysed per run e Can be used on more up to date operating platforms e 50 years rainfall data can be imported e Several updates to the user interface e All output files can be exported to the directory and folder selected by the user e Imported SMD data is now stored within the database 70 WRc plc 2009 STORMPAC 4 1 User Guide e Parameters used in the Soil Moisture Deficit SMD calibration and UCWI calculations are now recorded within the database tables and the software will default to these values on reopening of the database e Similarly values used to filter and sort events are also recorded within the database and will be defaulted to on reuse of the database 8 8 What was new in version 4 0 RED file output includes evaporation on the first profile used in Wallingford runoff model Options added to speed up Poisson disaggregator Chart display to compare summary statistics for imported daily rainfall and generated rainfall Time filter on Event definition form e Multi site RED generator available this allows the user to generate RED files with many rain profiles Option to change SEED value in SRG removed 8
106. y Quantile Plot for July Series simulated 5 min rainfall mm 6 0 2 ai 6 8 10 12 historical 5 min rainfall mm Figure C2 Comparison of Historical and Simulated 5 minute Rainfall Distributions for July 88 WRc plc 2009 STORMPAC 4 1 User Guide C4 DISAGGREGATION PROCEDURE C4 1 Summary of the Algorithm The disaggregation procedure can be summarised as follows A 1 hour rainfall depth is read in from a file which contains the hourly series to be disaggregated A 5 minute series is simulated using the fitted model The 5 minute simulated series is summed and the 1 hour total of the simulated series compared to the 1 hour total that was read in The simulated series is discarded if the absolute difference between the simulated 1 hour total and the total read in exceeds 0 05 mm The process is repeated until the totals are in agreement to within 0 05 mm The simulated series then represents a possible realisation of 5 minute data representative of the 1 hour value that was read in Following this procedure a record of 1 hour rainfall depths can be disaggregated In the above procedure some additional rules are applied These include using overlapping 5 minute values from the previous disaggregated hour provided the total of these overlapping values do not exceed the 1 hour total to be disaggregated This allows for some influence due to overlapping cells from the previous hour In addition the disaggregated series are scaled t

Download Pdf Manuals

Related Search

Related Contents

Manual de instrucciones bilingüe (Español e Inglés)  INSTALLATION MANUAL - Discovery Resources  Sinistar Instruction Manual  取扱説明書 バタフライバルブ 56・75 型 バタフライバルブ 56D・75D 型  Leia - AGR - Agricultura de Precisão  SEZ220 - Siemens  Samsung XE500T1C User Manual (Windows 8)  MT-14/MT-17 - Marmon-Herrington Aftermarket  Cypress Perform STK12C68 User's Manual  

Copyright © All rights reserved. Failed to retrieve file