FyrisNP manual
Contents
1. 9 5 Calibration and parameter sensitivity uncertainty Good modelling practice requires the user to calibrate the model such that the computed output compares well to the measured data This is done by means of adjusting the model parameters and kvs The model provides the user with three different methods for calibration and or evaluation of sensitivity to individual parameter values It is possible to choose which sub catchments to include or exclude in the calibration procedure The time period can also be specified 1 The manual calibration allows the user to manually change both parameter values co and kvs after which the model performs one simulation over the selected time period using data from the selected measurement stations to calculate the model efficiency and the linear correlation coefficient The user can inspect the simulation by looking at time series graphs or graphs of simulated versus observed concentrations for the selected sub catchments The manual calibration is useful for getting acquainted with the model behaviour and provides graphical means of performing a more subjective eye pleasing calibration which is occasionally preferred 2 In the Monte Carlo simulation the user specifies a uniform distribution of values for both parameters and the number of individual simulations to carry out As was the case for the manual calibration the simulation covers the selected time period and the selected in stream conce2. 01 1985 12 31 128 99 129 99 127 87 149 02 1986 1 31 46 06 49 17 53 14 61 39 1986 2 28 17 43 17 62 18 97 19 82 1986 3 31 109 41 111 77 109 01 124 17 Figure 3 The Specific runoff worksheet 5 3 COBS The worksheet named COBS contains data on measured nutrient total nitrogen or total phosphorous concentrations Notice that only one nutrient at the time can be calculated for in the model Data is provided once a month in mg and the model only allows for one measurement station per sub catchment The station should also be situated in the outlet point of a sub catchment The first month must be assigned value 1 January The first row in this worksheet should contain headings Figure 4 From column C and further to the right the sub catchment ID corresponding to the measurement station should be provided Only include the sub catchments for which there are measured values If values are missing for a certain month type 99 in the cell This informs the model that there is a missing value for that month and sub catchment Notice that missing data may only be assigned the value 99 in this worksheet In the other worksheets 99 will be interpreted as data with value 99 12 aA Il Be Il e I gt I e T F e 1 Year Month 1367 1377 1386 1389 1394 281 1985 1 99 99 99 99 99 EN 1985 2 99 99 99 99 99 4 1985 3 99 99 99 99 99 9j 1985 4 99 1 16 1 71 3 09 1 71 BN 1985 5 99 99 99 99 99 eta 1985 6 99 9
3. catchments which for some reason deviate from the rest and is thus a useful tool in the process of understanding the system and identifying key processes Three types of data can be visualised in the catchment overview namely input data calibration data and output data The input data contains apart from areas information about the usage of the lake module and the deposition on lakes and clearcuts The calibration data contains the position of the observed chemical data and the calibration parameter values The output data contains the four columns from Catch contr 32 The colormaps are either defined over the range from minimum to maximum value of the data shown or by clicking the 0 100 check box over the range from 0 to 100 The latter is mainly useful the for percentage values of the areas r Catchment overview Click the listboxes to view data _ Scale 0 100 LakeArea km2 StreamArea km2 Mountain 2 DepLake kg month K DepClearcut kg mont Calibration data Observed station COE c0 Mean retention coeffic Contribution at source Contribution at outlet Delivered fraction Colormaps Standard Figure 28 The catchment overview graph showing delivered fractions of nitrogen to the catchment outlet 8 4 Figure options It is possible to zoom in any of the plots from the Calibration tab Results tab or Catchment overview plot by marking the area
4. flow Q gives the concentration The sum of the columns from Mountain Forest and onwards give the gross load from the sub catchment 7 Introduction to the various windows of the model 7 1 The project manager The project manager is where you organize the work you carry out using the FyrisNP model The project manager comprises three parts working directory the workspace panel and the projects panel Figure 14 The project manager is open by clicking on Menu Projectmanager in the main program window The default working directory is in the program file folder of FyrisNP but it can be changed to e g My Documents The workspace is essentially a directory on your computer found under the FyrisNP model directory that contains a number of projects A possible way of organizing your work is to give the workspace the same name as the catchment and nutrient N or P you currently study and then store all simulations related to that catchment as projects in that workspace If you click any workspace in the list the projects belonging to it will show up to the right Projectmanager working directory Workspace Projects Simulation Date Description Current working directory Demol 11 3 2011 11 52 58 AM demonstration data set from file demo xls C Documents and Settings Administrator Mina dokument FyrisNP simulations lt gt Figure 14 The project manager window 7 2 Workspace panel The New button a
5. means of Monte Carlo simulations where parameter values are generated randomly according to user specified uniform distributions The results can be visualised through scatter plots Figure 22 The scatter plots reveal if any parameter is redundant and will also give information about what parameter values will give the best fit to the measured data The parameter values co and kvs giving the best fit to measured data generated with Monte Carlo simulations are then typically used to run the model in manual calibration mode r arima Menu About Catchment overview General Data Q data Manual Calibration Monte Carlo Automatic Massflux efficiency N B Preliminary model implementation Values are not stored in a saved project Ef Subcatchments with observations the given properties affect the upstream subcatchments to the next one with measurements CatchmentlD IncludeObs v Uniform v CON Dio v N E SI IS S ST IST S ST ST S Figure 18 The calibration tab 7 7 1 Settings The calibration settings include specification of the time period you want to simulate The FyrisNP model does not use date information of any sort thus months are simply numbered from one and In addition to time period the user can specify which observation stations IncludeObs to include in the calibration Le the statistics will only be based on the
6. model changed name from Fyris to FyrisNP The major change from version 3 0 is that a new graphics library is used in FyrisNP The library ZedGraph http zedgraph org wiki is licensed under the GNU Lesser General Public License meaning that it can be included in FyrisNP for public use For more information consult the Zedgraph webpage at http zedgraph org wiki index php title LGPL_ License Zedgraph offers some new functionality including the ability to copy save or print what you see in the graph In addition it is possible to zoom and pan the graph or display individual point values by hovering over a specific data point The other important addition is a new plot of tree type that enables the user to view the structure of the catchment i e the sub catchments and the connections between them This is a great help to make sure data was supplied and imported as intended It is also a very effective technique to present data that is related to geography 2 3 Version 3 2 Three major changes have been made to FyrisNP e The first is the possibility of stepwise calibration with different parameter values in different sub catchments When implementing the stepwise calibration several changes were made to the calibration tab The plot was lift out to a separate window while the calibration settings were moved from their separate window into the calibration tab e The second is the possibility of calculating type specific concentrations for
7. not used in the calculations Column E is the Qfraction that decides how much of the stream flow is diverted from the stream into the wetland i e a value of 0 3 indicate that 30 of the stream water flow passes the wetland Finally column F contains the value of the degradation constant For details on the conceptual model and equations see the technical description of the FyrisNP Model ou D E F 6 Catch ID CW include 1 Area km2 Depth m Qfraction k degradation 1 1 30 1 1 0 0023 2 0 0 0 0 00023 3 0 0 0 0 00023 4 0 0 0 0 00023 5 1 50 2 03 0 0023 6 1 30 2 0 3 0 0023 7 0 20 2 03 00023 8 0 30 2 0 3 0 0023 9 0 0 0 0 00023 10 0 0 0 0 00023 Figure 12 The first rows and columns of the Constructed wetlands worksheet 18 5 12 Comments optional This is an optional worksheet giving the user the possibility to add information about the data constituting the input data file It will be seen in the Comments box in the General tab of the model The comments are written on row 1 8 in column B in the worksheet Figure 13 A B Comments row 1 Omr de Lagan M nad 2000 2010 Comments row 2 Skapad 2011 09 08 av Caroline f r FyrisNP 3 2 0 19 Comments row 3 I Settings skriv PLCS Type spec conc f r hygge org N skog Skogsregion Sv Region ppen Sv Other Utl Mountain Produktjor Comments row 4 Fl de Matdata Temp M tadata COBS M tdata Allt fr n v r databas ext
8. other minor point sources that should be included in the source apportionment calculations Even if a catchment has no minor point sources it must be included like catchment 7 in Figure 10 The catchments should be listed in the same order as in the Catchment worksheet A B 8 D Catchment ID Households kg month Minor 1 kg month Minor 2 kg month 11 315 21 065 14 83333333 7 511666667 29 71166667 5 636666667 0 1354166667 5 553333333 10 6 773333333 11 38 96333333 Figure 10 The first rows and columns of the Minor point sources worksheet o ee en s a ho Co mB Co 51 rem ho 5 10 External load optional Occasionally it is of interest to study a downstream part of a larger catchment in more detail In such situations it is unnecessary to model the upstream parts of the entire catchment if continuous measurements on a monthly basis of nutrient concentration and measurements or modelled data on water flow rate exist The measured nutrient input from upstream areas can be considered as an external load on the downstream system and may be added as an input to a specified sub catchment Notice that the actual external load will not end up in any results plot specified as external load However it will be implicitly visible since the sub catchment it is directed to will most likely have a quite high impact on downstream areas as compared to other
9. similar sub 17 catchments It will not influence the source apportionment calculations since it does not belong to any specific source The worksheet contains year in column A month in column B water flow rate in column C and mass flow rate of nutrients in column D Pay attention to the first line of the worksheet Figure 11 Looking at cell C1 Qe4 denotes external flow rate m month to sub catchment 4 Similarly for cell C4 Me4 denotes external mass inflow of nutrients kg to sub catchment 4 Month Qe4 2002 1 2 26765E 11 272119 2002 2 474422E 11 441213 2002 3 4 50913 11 392294 2002 4 1 82831 11 162720 2002 5 1 2528E 11 107741 2002 6 1 3017E 11 95912 2002 7 2 05226E 11 160076 2002 8 1 38993E 11 98695 2002 9 81597023914 57934 2002 10 74306581331 58702 2002 11 1 29254 11 94015 2002 12 1 07575E 11 91439 Figure 11 The first rows and columns of the External load worksheet 5 11 Constructed wetlands optional The Constructed wetlands worksheet contains technical information as well as data on the geometry and degradation characteristics for every wetland Figure 12 Looking at column B it contains a logical switch that determines whether the constructed wetland shall be taken into account for the specific sub catchment or not 0 not included 1 included Column C contains the surface area in km of the wetland Column D contains the mean depth in meters of the wetland The mean depth is currently
10. sub catchments based on the nitrogen deposition clearcuts in southern Sweden according to L fgren amp Westling 2002 or on altitude forest clearcuts and mires in northern Sweden according to Lofgren amp Brandt 2005 analog to the PLCS calcualtions Brandt et al 2008 e The third change is that retention is calculated also at negative temperatures Important minor changes 1 Added the possibility to set the working directory i e the folder where the simulation files are to be stored 2 Added program version to the program name such that old installations of FyrisNP will not be overwritten when the new program is installed 3 Automatic average calculation of total averages for each sub catchment as well as averages for each month Jan Feb etc and year Additionally a lot of meta data about the calculations added to Results out and Results averages out 4 Inthe catchment overview a possibility to plot the location of the nutrient measurement stations and the parameter values co and kvs have been included Additionally the land use areas are now shown in percentages of the sub catchment areas and a 0 100 scale can be chosen 5 A possibility to already in the input file write comments about the input data which are seen in the general tab of the model 6 Model efficiencies are now calculated for transports as well but these values are not saved Increased the number of possible Monte Carlo simulations to 10000 8 B
11. the Q stations data storage in model lakes and for all sub catchments area specific runoff flow rate and hydraulic load Figure 17 t FyrisNP Menu About Curent project N H0 Qmedel General Data Q data Calibration Scenario Result Qstations Storage Specific runoff Flowrate Hydraulic load Figure 17 The Q data tab contains data regarding runoff storage specific runoff flow rate and hydraulic load 23 7 7 The Calibration tab The basic idea of this tab window Figure 18 is to provide means of calibrating your model and evaluate how sensitive the chosen goodness of fit value is to changes in parameter values see section 9 5 Manual calibrations automatic calibrations and Monte Carlo simulations may be used to find out the optimum set of parameter values The user can specify calibration specific settings Table 3 Performing an ordinary simulation with FyrisNP is quite straight forward Select manual calibration enter calibration settings select parameter values for co and and push the button Parameter co is allowed to vary between 0 and 1 and parameter from 0 and up However kvs values above 30 are considered quite rare The simulated results are presented in graphs as time series for the catchments having nutrient measurement stations Figure 20 The influence of parameters co and kvs can be analysed by
12. the load from this sub catchment at the chosen outlet point i e after retention in every downstream sub catchment on the way to the outlet The column Mean retention coefficient contains the mean retention ratio in each sub catchment The column Delivered fraction contains the ratio of the Contribution at outlet to the Contribution at source Please note that the Mean retention coefficient and the Contribution at source are calculated for the sub catchment itself while the Delivered fraction and the Contribution ad outlet varies depending on which sub catchment is used in the Out textbox 23 Please note that the order of the sub catchments in the results from Catch contr button is different from the order of the input data The data can be viewed in the catchment overview graph Figure 28 provided that the results table contains the appropriate data 3l 7 14 Out button By clicking the Out button the outlet sub catchment will automatically be inserted in the textbox labelled Out Figure 24 7 15 The Catchment listbox Clicking any of the catchment ID numbers in this listbox will present the part of the complete results file relating to the chosen ID 7 16 Other features Other features of the Result tab are Tip copy the present contents of the data table to the clipboard by clicking the Copy button save the entire results file consisting of all catchment ID data tables to a semi colon se
13. to focus on with the mouse Right clicking will give several options to change the scale of the figure back Figure 29 Right clicking will also give the possibility to copy the figure to the clipboard and to highlight point value Figure 29 If Show Point Values is clicked point values will show up when moving the mouse over the figure 33 Save Image As Page Setup Print Show Point Values Un Zoom Undo All Zoom Pan Set Scale to Default Figure 29 Available options when right clicking in a figure 9 How to start using the FyrisNP model 9 1 System requirements In order for the FyrisNP Model to run properly on your computer it is required that you have Windows XP Vista or 7 including NET framework installed The NET framework should be provided with Windows Service Pack 2 for Windows XP but can otherwise be installed separately from Windows homepage Furthermore Office 2003 2007 or 2010 is required since the input data is imported using an Excel workbook 9 2 Installing the program on your computer A standard set up program comprising two files setup exe and Setup Fyris Model msi is provided for installation of the FyrisNP model on your computer The program will check your computer to make sure that the NET framework is previously installed and will encourage you to download it if it is missing During the installation you have the possibility to decide in what folder you want the program i
14. 5 22696 0780720198 19708 5034381861 1025935 63048924 955605 801299128 45108 9189205572 45002 1628393904 83478 4350754792 82393 8505689664 86061 318079257 84425 8265142201 181662 384721949 170711 094076576 203371 170853669 191825 109771021 187685 821360688 162979 634122476 Figure 24 The Result tab provides the means to display copy and export output data as well as program procedures to e g perform source apportionment The Result tab lets you look at the entire results file in itself sorted according to sub catchment In addition it provides the user with the possibility to automatically perform various computations by pressing the buttons Internal load Sources Apportionment Catch contr Figure 24 These buttons are explained in the following sections 7 10 Internal load Pressing this button presents in the tabular area here referred to as the data grid view of the Results tab the internal gross load before retention in each sub catchment and the net contribution after retention of each sub catchment to the sub catchment itself summed over the entire simulation period The results can be viewed in the plot window Figure 25 Please note that while the gross contribution can be summed for all sub catchments it is not allowed to sum the net contribution as it will give a too high value To get the net contribution of each s
15. 9 99 99 99 om 1985 T 99 1 32 0 93 0 85 0 89 BN 1985 8 99 99 99 99 99 BOE 1985 9 99 99 99 99 99 1985 10 99 2 04 1 02 0 94 1 04 1985 11 99 99 99 99 99 13 1985 12 99 99 99 99 99 Figure 4 A typical COBS worksheet containing measured and missing concentration values of the studied nutrient 5 4 Catchment This worksheet contains characteristics on the different sub catchments that is needed for the quantification of the nutrient transport This includes information on hydrological network of sub catchments and sub catchment specific data on land use deposition type specific concentration in runoff from arable land and pasture and data on included model lakes F DownstreamlD Area km2 LakeArea km2 StreamLength m 1348 108004 1350 46 507 0 16892 1360 108004 1367 24 185 0 007 15783 1364 108004 1360 20 853 n 8506 1367 108004 1397 125 418 20 933 26148 1377 108003 1391 46 887 0 844 16529 1383 108003 1389 69 316 1 084 34153 1386 108005 1412 69 856 1 086 36019 1389 108003 1391 4 672 0 3392 1391 108003 1386 10 473 0 4136 1393 108003 1383 37 246 0 9436 1394 108003 1389 48 064 0 253 21290 1397 108004 1407 15 453 0 323 5535 1402 108005 1412 5 55 0 812 3043 1403 108003 1383 7 228 0 5001 Figure 5 first rows and columns of an example Catchment worksheet The position for each column must not be changed The variable name in the header row can however be changed possible to include translation of v
16. Catchment worksheet in section 4 1 If a modelled lake is judged not to have any significant changes in storage fill in zeros 0 for this lake The number of columns equals the number of Model lakes and dams in the model and the number of rows equals the number of months Include a header row where the column headers are the catchment IDs of the sub catchment the model lake belongs to The catchment IDs should increase from left to right In the example in Figure 9 there are two model lakes one in sub catchment 1367 and one in 1407 Only 16 the lake in 1407 has significant changes in storage over the year Figure 9 Every row represents one month If there are no Model lakes included in the model you can leave the worksheet blank ao e 1367 1407 23 37 10 97 26 13 29 33 33 97 9 27 5 87 283 10 53 OOoOooOooOooocoo o Figure 9 The Storage worksheet 5 9 Minor point sources This worksheet contains data from small point sources such as scattered households with autonomous sewage treatment system The data should be organized according to Figure 10 i e the first row should contain headings for every column It is of no importance what is actually written just make sure to write something on line one and start providing data on row two Column A should contain the catchment ID column B the calculated load per month kg month from households In column C and D you have the possibility to insert data on
17. JJ SLU Sveriges lantbruksuniversitet The FyrisNP model Version 3 2 A tool for catchment scale modelling of source apportioned gross and net transport of nitrogen and phosphorus in rivers A user s manual Institutionen f r vatten och milj SLU Box 7050 750 07 Uppsala Rapport 2012 8 Sveriges The FyrisNP model Version 3 2 A tool for catchment scale modelling of source apportioned gross and net transport of nitrogen and phosphorus in rivers A user s manual Elin Wid n Nilsson Klas Hansson Mats Wallin Faruk Djodjic amp Caroline Orback Institutionen f r vatten och milj SLU Box 7050 750 07 Uppsala Rapport 2012 8 Institutionen f r vatten och milj SLU Box 7050 750 07 Uppsala Tel 018 67 31 10 http www slu se vatten miljo www slu se vattennav Contents A MODEL HISTORY 7 VERSION HISTORY aep T NERSION RE Sa 7 2 2 VERSION 3 1 eee aene ee 8 2 9 VBRSION S rs tete he ete oe Shean ie ets De eae A 8 MODEL DESCRIPTION vb ae ao UE EE YN aeuo ra oe e sa ve e be roo Nea eg 9 USE OF OTHER MODELS TO CALCULATE INPUT DATA creen eee ee ene een to ee sese esee esten ae eS 10 INPUT DATA BILE oseoees sesto ca obe cod aao Fo Fe ee Ue Ne eU Eve ep sua A 11 SSE TEMPERATURE washed deiner dicere iden
18. Lindstr m 2010 the HBV model Bergstr m 1995 the Q model Kvarn s 2000 or FyrisQ the windows version of WASMOD Xu 2002 Atmospheric deposition of nitrogen on water is calculated by the MATCH model www smhi se 5 Input data file In order to perform simulations with the FyrisNP model an Excel file containing all input data is required Any name may bee chosen for the Excel file as long as it has the x s extension Input files made in Excel 2007 or Excel 2010 should be saved as in Excel 97 2003 file format to get the x s extension and to be able to open it in FyrisNP The Excel data file contains between eight and twelve different worksheets depending on what features are used Figure 2 The inclusion of the External load Constructed wetlands Settings and Comments worksheets is optional but the Storage worksheet must always be supplied it can be empty if no storage is considered The order of the worksheets in the workbook is of no importance however the names and content of each worksheet must obey the guidelines in this manual Make sure the name of the worksheet is spelled according to Figure 2 and that the upper lower case is correct Remember to include N or P in the file name N Catchment COBS Major point sources Minor point sources Temperature Type spec conc Specific runoff Storage Figure 2 The worksheets required in the Excel workbook Optional worksheets are External load Constructed wetlands
19. Settings and Comments The contents and layout of the worksheets included in the input Excel file will be covered in the following paragraphs First the worksheets containing the driving variables i e temperature and runoff will be described The temperature worksheet is governing the setup of the internal data tables and arrays of the program in other words the program will set up everything to fit the time period specified in this file Make sure that the temperature worksheet and the runoff worksheet cover the same time period and that values are specified for every month or week Additionally observed concentrations major point sources and external load must match this time period and resolution The model requires input data time series to start in January It is later possible to restrict the calculation to any time period in the Calibration tab Additional worksheets are allowed although not recommended They will give a critical model error message but are thereafter ignored by the model 5 1 Temperature This worksheet includes data on measured water temperature monthly mean from a station in the modelled catchment or from a nearby station outside the catchment If such data are not available data on measured air temperature monthly mean is an acceptable approximation of water temperatures The columns from left in the temperature worksheet are Year Month 1 12 Number of days per month 28 31 and Temperature F
20. an 2 0 0415 1 300 0 765 0415 3 0 0 415 1 300 0 765 0 415 0 4 0 0 425 1 330 0 805 0 425 0 5 0 0 425 1 330 0 805 0 425 0 6 0 0 425 1 330 0 805 0 425 0 ri 0 0 620 1 550 1 220 0 620 0 8 0 0 620 1 550 1 220 0 620 0 9 0 0 620 1 550 1 220 0 620 0 10 0 0 530 1 450 1 030 0 530 0 11 0 0 530 1 450 1 030 0 530 0 12 0 0 530 1 450 1 030 0 530 0 13 0 0 415 1 300 0 765 0 415 0 Figure 8 specific nutrient concentrations mg l in the runoff from different land uses 5 7 2 PLC5 calculations With the selection of PLCS calculations for nitrogen in the Settings tab the model first checks if the altitude is above 0 in the sub catchments Altitudes 0 is used in FyrisNP for the southern Sweden forest regions according to the division in PLCS Brandt et al 2008 In this case the clear cut column in the Type spec conc tab should be set to the organic nitrogen leakage instead of the total leakage These values can be found in L fgren amp Westling 2002 Altitudes gt 0 is used for the northern Swedish forest region In this case all columns in the Type spec conc tab should be set to a monthly factor around 1 These values can be found in L fgren amp Brandt 2005 For further details about these calculations it is referred to the technical description of FyrisNP version 3 2 Wid n Nilsson et al 2012 5 8 Storage This worksheet contains changes in water storage m 10 in lakes and dams included as Model lakes in the model see
21. and you can assume that the loading is evenly distributed through the year you may divide the yearly load with 12 to get the monthly load While the model lakes are assumed to be at the sub catchment outlet the major point sources are assumed to be situated far upstream in the sub catchment No missing values are allowed so for months with no data you must insert an approximation of the load e g the load for the same month previous year or a mean value for e g the previous three months If your simulated catchment has no major point sources you must put a fictive major point source in this sheet with load 0 kg month A B EEG E J Catchment ID Facility Year Month Load kgimonth EN 1345 Odensberg WWTP 1985 1 85 0 1345 Odensberg WWTP 1985 2 85 0 mac 1345 Odensberg WWTP 1985 3 85 0 5 1345 Odensberg WWTP 1985 4 85 0 ies 1345 Odensberg WTP 1985 5 85 0 EA 1345 Odensberg WWTP 1985 B 85 0 g 1345 Odensberg WWTP 1985 7 85 0 ES 1345 Odensberg WWTP 1985 8 85 0 10 1345 Odensberg WWTP 1985 9 85 0 1345 Odensberg WWTP 1985 10 85 0 12 1345 Odensberg WTP 1985 11 85 0 Figure 6 The first rows and columns of the Major point sources worksheet 5 6 Settings optional This is an optioal worksheet if the standard type of calculations is chosen If PLCS calculations are selected cell B2 Figure 7 should contain the text PLCS or plc5 plc 5 or PLC 5 and cell A2 must contain information of the substance Nitrogen N n nitroge
22. ariable and more detailed description The variable names units and description of the variables are given in Table 1 Table 1 The variables included in the Catchment worksheet Variable name Unit Description Catchment ID Sub catchment ID number Station ID The ID number of the nearest downstream flow measuring station Downstream ID Downstream sub catchment Area km Total area of sub catchment Lake Area km Lake area Stream Length m Stream length 13 Stream Area km Stream area Mountain km Mountain area above tree line Forest km Forested area Clearcuts km Clear cuts not older than 5 years in S Sweden and 10 years in N Sweden Mire km Mire Wetland Arable km Arable land Pasture km Pasture Open land km Other open land Settlements km Settlements Urban km Urban areas cArable mg l Type specific concentration from arable land cPasture mg l Type specific concentration from pasture Altitude m Altitude above sea level Used to identify Northern Sweden in the PLCS type calculations Set to 0 for Southern Sweden Lake Model 1 Yes 0 No Dep Lake kg month km Nitrogen deposition on lakes zero for P kg week km monthly or weekly values Dep Clearcut kg month km Nitrogen deposition on clearcuts zero for P kg week km monthly or weekly values Model Lake Name Model Lake Area km Model lake area M
23. berecceeeves 19 INTRODUCTION TO THE VARIOUS WINDOWS OF THE MODEL eee ee eee eee enne eene eee tnueee 20 11 CPHEPROJECEMANNXGER mete beni b r a ven penderet e ke 20 7 2 WORKSPACE PANEL edet edendis iere 20 Usd PROJECTS PANEG S ven ii 21 THE GENERAL TAB rites Iii eU 21 d 5 2PHEDATAXTABAZ SR ES ee iSt 22 ZDHEQ DATATABZz Aran enemies dum 23 JJ THE CALIBRATION eeeEeUT dumb eni ugue D denbi He 24 FAD uSetingssssi aan due p eb eno n E E 24 7 7 2 Results and plots in the calibration esee enne nennen rennen 26 773 Distributed parameter values so here Hte ere ee Rete HR RR eh 28 8 SDHESCENARIO eene eran sne sa dumis eic nuni menitue stet iatis eed 29 QE HEAR SUIT cte chiari eaten eth ica tester D ben vat 29 7 10 INTERNAT LOAD i MN re ae 29 7 11 SOURCES io toe LIE tort ctetu Das eo Robb Ate cost oid te cae E Dri ben ved ta 30 7 12 APPORTIONMENT cs omncm eine TA ELI etes b ia I Mm E eie 31 7 13 CATCH CONTE 5 nievet a Or VH UHR REN REOR RR eee 31 7 14 OUT BUTTON AIR RERBA QE 32 7 15 THECATCHMENT LIS TBOX 32 7 16 OTHER FEATURES b en MIRO ERR EVI RH EN UB Ba 32 SPECTAL FRA TURES rc 32 8 1 REMOVAL OF EMPTY ROWS IN THE INPUT DATA TABLESG ccccccccecececececece
24. cececesececesscecesssesesesesesssessssceseeesseseeeeess 32 9 2 COLUMN MODE SWITCH tee ret C EO E Ree e ANN 32 8 3 CATCHMENT OVERVIEW GRAPH cccccccsssssssssscccscececscecscscscscscecscscscesscssscscesssscscecssecscesscesesssecssscssesscecssesseseusecess 32 8 4 IPIGURE OPTIONS 33 9 HOW TO START USING THE FYRISNP MODEL e eene eee eee oeste seta ases seen sese sees eee 34 9 1 SYSTEM REQUIREMENTS sosseosseseereressosersrereressererrerersererreresrererr enes res state dadas asas 34 9 2 INSTALLING THE PROGRAM ON YOUR COMPUTER sssssoreerereseseresereseresererererrrerererrrrrrrrrrr rnnt rr rr rr rr rr rr rr rr br rr br sr rr er rr en 34 9 3 IMPORTIDATA inc det deaucoancedac 34 9 4 STATISTICS USED FOR CALIBRATION 35 9 5 CALIBRATION AND PARAMETER SENSITIVITY UNCERTAINTY 35 UG ZSCENARIOS M eT A des 36 10 REFERENCES E 36 1 Model history The dynamic Fyris model was originally developed by Hans Kvarn s at the Dept of Aquatic Sceince and Assessment at SLU for calculating
25. checked stations in this list 24 Table 3 The calibration settings in FyrisNP explained for the three modes of calibration tpn E The manual calibration settings are the c0 and kvs 05 values pH _ kys 2 Eff 10 4877 0 7870 The Monte Carlo settings include number of Monte Carlo simulations to perform Currently the maximum number Number of simulations 300 is limited to 10 000 Parameters Distribution Lower Upper mean std Monte Carlo settings When carrying out a Monte Carlo simulation the a 0 05 parameters are randomly selected from an interval Only I I5 30 uniform distributions are used in FyrisNP In the example to the left the parameter values are thus randomly selected 300 times from the specified intervals 0 to 0 6 for co and 5 to 30 for kvs Parameter may vary between 0 and 1 and parameter kvs from 0 upwards The settings for automatic calibration contain two criteria for judging when the calibration process should be terminated The first is simply the maximum allowed Iteration ends when best worst lt 0 001 number of iterations the second is a condition related to Parameter limits Lower Upper the method used The given number is associated with the 0 difference model efficiency between the worst and best 3 simulations of the current simplex Put in other words for the specific example it would be very surprising if an unk
26. ciety for Water Conservation Report no 46 Kvarn s 2000 The model A Simple Conceptual Model for Runoff Simulation in Catchment Areas Dept of Environmental Assessment SLU Report 2000 15 Larsson M H Persson K Ul n B Lindsj A and N J Jarvis 2007 A dual porosity model to quantify phosphorus losses from macroporous soils Ecological Modelling 205 123 134 Lindstr m G Pers C P Rosberg R Str mqvist J Arheimer B 2010 Development and test of the HYPE Hydrological Predictions for the Environment model A water quality model for different spatial scales Hydrology Research 41 3 4 295 319 L fgren S amp M Brandt 2005 Kv ve och fosfor i skogsmark fj ll och myr i norra Sverige Rapportserie SMED och SMED amp SLU Nr 14 2005 ISSN 1652 4179 Nitrogen and phosphorus in forests mountains and mires in northern Sweden Report series SMED and SMED amp SLU 14 2005 36 L fgren S amp O Westling 2002 Modell f r att ber kna kv veforluster fr n v xande skog och hyggen i Sydsverige Institutionen f r milj analys SLU rapport 2002 1 Model for estimating nitrogen losses from growing forests and clear felled areas in southern Sweden Dept of Environmental Assessment SLU Report 2002 1 23 pp Nash J E and J V Sutcliffe 1970 River flow forecasting through conceptual models part I A discussion of principle Journal of Hydrology 10 3 282 290 Son
27. d uie 11 53 2 SPECIFIC RUNOEE e te eeieheete dee ebeetdpeete eei a a eeu e 12 mo UCOBS I 12 FA CATCHMENT E Ne 13 3 9 MAJOR POINT SOURCES 5 cccsnivsscosdendudadaoubesadauajavacdsdoaburdenospangeupdearsdaodbendeceetledeqoopenteancbursdes dberseapetueduaepargvandessedee 15 5 6 SETTINGS OPTIONAL eie audeat ie DE uae eiiis 15 9 7 UE 16 5 7 1 Standard calculatioris snl ie ede SA EEES E eR 16 5 7 2 cdlculations nte opido 16 SE ENNIO P CLR m 16 3 9 MINOR POINT SOURCES d sette o ce Peek eco et eee EE e Pese ied sae deve t E Pe ee eei eve 17 5 10 EXTERNAL LOAD OPTIONAL aede d cf recte ce ee du ete Eee a es Rea e E ve E Ee tee 17 5 11 CONSTRUCTED WETLANDS OPTIONAL asessesssssrsrsssrsrserrrrsrrressrnrrrrrrnsrrrrr sr rss sr rr rr rr ser seen rene enne enne 18 5 12 COMMENTS OPTIONAL 5 ee ici enes tentan re e o eed te e i Re t eire ee 19 OUTPUT DATA CETT w 19 61 RESULETS XML zusaeueNSNUNDeuIeveus ege aden sede a beaded Nis edn sat ele eed 19 0 2 MONTECARL Oi OUT ss dra deras rss SVs Ses aed nee eae ade ee ee Dd 19 6 3 RESULTS OUT RESULTS AVERAGES OUT ccce ccessccccsscesescvennccetegueeuscaucetuecdeedeessptiecessovsvsdentendeovessecctcateaseess
28. e PLC data are derived using different models These models as well as some complementary models are summarised below For several years SLU researchers have been developing robust calculation models that can be used to estimate leaching of both nitrogen and phosphorus from Swedish arable land and to see how leaching is affected by various measures The NLeCCS Nutrient Leaching Coefficient Calculation System modelling system comprises the SOILNDB and ICECREAMDB models The dynamic SOILNDB model Johnsson 2002 is used for calculating type specific concentration of nitrogen in leaching from agricultural land For calculating the type specific concentration of phosphorus in run off from agricultural land the dynamic ICECREAMDB model Larsson et al 2007 is used The NLeCCS modelling system generate the type specific concentrations for a given nutrient N or P as an annual average concentration normalized for climatic conditions during a longer time period and typical for a combination of crop and soil types for phosphorus also for P content in the soil and slope More details about the calculations of type specific concentrations in runoff from arable land 1s described in Johnsson 2008 For calculating the type specific concentration of nitrogen and phosphorus in run off from forested areas a regression model is used L fgren amp Westling 2002 Several options are available for calculating runoff and water discharge Examples are the HYPE model
29. ernt och SRK Comments row 5 PLCS data Delaro Area markanv o jordbruk Deposition Dagvatten Typhalter Comments row 6 PLC6 data Enskilda avlopp Comments row 7 F rb ttrningar M tdata stora punktk llor 75 percentil skogstyphalt L fgren amp Westling 2002 inre Gr dmix dvs nya jordbruksl ckage Comments row 8 Extra flikar External Load till delaro 43 00 4 oD ane Figure 13 The Comments worksheet 6 Output data The program produces two major output files montecarlo out and results xml montecarlo out contains information regarding the Monte Carlo simulation of the project provided such a simulation has been performed while results xml contain all other results Optionally by clicking the write file button in the Results tab it can also write the files results out and results averages out 6 1 Results xml The content of this file can be viewed in the graphical user interface of the FyrisNP model see chapter 6 of this manual but may also be opened in external programs for data analysis It is however recommended that the user instead export data using the write file button in the Results tab window to obtain two semi colon separated text file named Results out and Results averages out These file can be opened in Excel and contains the information included in the Result tab described in section 6 3 6 2 montecarlo out This file contains the parameter values and model efficie
30. ers To activate this feature unclick the Uniform checkbox More columns will then be seen in the in the tabular area here referred to as the data grid view of the calibration tab Clicking IncludeSub means that that sub catchments including its all upstream sub catchments will be included in the calculation IncludeObs is similar to the calculation with uniform parameter values i e it tells if the concentration measurements of that sub catchment will be used in the model performance statistics or not If OwnValues is checked the parameter values of that sub catchment will be fixed to the ones listed in the table and will not be varied in a Monte Carlo or automatic calibration 28 7 8 The Scenario tab This tab is currently not in use In order to run different scenarios you have to make the changes in the input file and upload it as a new project Then run the model using the original calibration settings and without any data on measured nutrient concentrations 7 9 The Result tab td FyrisNP Menu About Current project Demo1 General Data Q data Calibration Scenario Result 139293 051136159 137409 096615681 112225 777128255 104055 797156628 177847 287533861 170819 205994847 287356 800027855 236692 428605082 81636 5305348438 72123 9800860578 1269719 32461446 1113781 44532034 648 118671028888 647 16744112528 27761 0064395196 27325 720889619
31. ess tillrinningsomrade L nsstyrelsen i G vleborgs l n Rapport 19998 13 ISSN 0204 5954 Modelling nutrients in Lake Storsj n and its catchment County Administration of G vleborg Report 1998 13 Johnsson H Larsson M H M rtensson K and M Hoffmann 2002 SOILNDB a decision support tool for assessing nitrogen leaching losses from arable land Pages 505 517 Environmental Modelling amp Software 17 505 517 Johnsson H Larsson M Lindsj A M rtensson K Persson and G Torstensson 2008 L ckage av n rings mnen fr n svensk kermark Ber kningar av normall ckage av kv ve och fosfor f r 1995 och 2005 Naturv rdsverket Rapport 5823 Leaching of nutrients from Swedish arable land Estimates of normal leaching of nitrogen and phosphorus for 1995 and 2005 Swedish EPA Report 5823 Kvarn s H 1996 Modellering av n rings mnen I Fyris ns avrinningsomr de K llf rdelning och retention Rapport fr n Fyris ns vattenf rbund 1996 31 sid Modelling nutrient transport in the Fyris River catchment Source apportionment and retention Report from River Fyris Water Conservation Board Uppsala Sweden Kvarn s H 1997 Modellering av n rings mnen i V tterns tillrinningsomr de K llf rdelning och retention V tterv rdsf rbunder rapport nr 46 ISSN 1102 3791 Modelling nutrient transport in the catchment of Lake V ttern Source apportionment and retention Lake V ttern So
32. esten L Wallin M amp Kvarn s 2004 Kv ve och fosfor till V nern och V sterhavet Transporter retention och tg rdsscenariern inom G ta lvs avrinningsomr de L nsstyrelsen V stra G talands l n rapport nr 2004 33 ISSN 1403 168X Nitrogen and phosphorus loading on Lake V nern and V sterhavet Sea Transport retention and nutrient reduction measures within the River G ta catchment County Administration of V stra G taland report no 2004 33 ISSN 1403 168X Sorooshian S and V K Gupta 1995 Model calibration In V P Singh ed Computer models of watershed hydrology chapter 2 Wallin M stlund M amp Kvarniis 2000 N ringsbelastning p V nerns vikar inom Karlstad kommun K llf rdelning retention m l och rg rder Institutionen f r milj analys SLU rapport 2000 6 ISSN 1403 977X Nutrient loading in coastal areas of Lake V nern within Karlstad Township Source apportionment retention environmental objectives and measures Swedish University of Agricultural Sciences SLU Dept of Environmental Assessment Report 2000 6 ISSN 1403 977X 113 p Wid n Nilsson Hansson K Wallin M amp G Lindgren 2008 The Fyris model Version 3 2 Technical description Swedish University of Agricultural Sciences Dept of Environmental Assessment Report 2012 9 ISSN 1403 977X Xu C Y 2002 WASMOD The Water And Snow balance MODelling system Chapter 17 of the Book Mathemat
33. esult tab Figure 24 Le it is possible to make an upstream source apportionment for every sub catchment outlet in the system if wanted The upper row of the presented data contains the actual net load in kilograms per month that each source contributed with at the chosen outlet point The lower row comprises the corresponding fractions of the load The results can be viewed in the plot window Figure 27 Results plot Source apportionment for catchment 1474 Mountain 0 0 96 Major point sources 26 3 96 Forest 11 8 96 Clearcut 1 8 96 Households 4 3 96 Mire 0 8 96 Arable 36 7 96 Pasture 2 6 96 3 Open 1 5 96 Urban 2 4 96 Lake deposition Built 0 0 96 Urban 2 4 96 Major point sources 26 3 96 Pasture 2 6 96 Households 4 3 96 pen Minor1 0 0 96 Minor2 0 0 96 Lake deposition 11 8 96 Forest 11 8 96 Arable 36 7 96 Figure 27 The source apportionment presented graphically by means of a pie chart 7 13 Catch contr Clicking this button will start the calculation of the nutrient mass contribution in kilograms per month from each sub catchment to the chosen outlet As mentioned in paragraph 7 12 the outlet catchment is selected by inserting its ID number in the textbox labelled Out Figure 24 The column labelled Contribution at source contains values that are identical to the values presented as gross contribution when pressing the internal load button The column labelled Contribution at outlet contains
34. f the calculations are specified to start at a later time step The Model lakes are assumed to be situated close to the outlet of the sub catchment and hence receive the total load from sources after retention in the sub catchment Furthermore there can only be one Model lake per sub catchment Information on water storage in Model lakes is inserted in a separate worksheet see section 5 7 Please note that even if you do not use the Model lakes the columns concerning model lakes cannot be empty The Model Lake Name can e g be set to 14 5 5 Major point sources This worksheet contains data from large point sources such as wastewater treatment plants and industries The data should be organized according to Figure 6 i e the first row should contain headings for every column It is of no importance what is actually written just make sure to write something on line one and start providing data on row two Column A should contain the catchment ID numbers column B the name of the facility column C the year for which the data applies column D the month for which the nutrient load is measured calculated numbered from 1 to 12 and column E the load per month kg month There can be more than one facility in a certain catchment area The data for the different facilities are inserted below each other Catchments that have no major point sources do not need to be included If only data on yearly loads are available
35. for more information textboxes displaying 27 parameter values and statistical measures and buttons for running the model and plotting simulations and data Menu About Current project 2011 08 16 General Data G data Calibration Scenario Result Manual Calibration Time period Manual calibration Monte Carlo settings Monte Carlo From cl Number of simulations 0 2742 Parameters Distribution Lower Upper mean std Multiple series o Composite a 2 Uniform o 10 Subcatchments with observations the given properties affekt the upstream subcatchments to the next one with measuremeyl CatchmentlD IncludeObs Uniform 1 KJ RIS IncludeObs Figure 23 The resulting view after a Monte Carlo simulation have been performed The arrow shows the copying of the best Monte Carlo or automatic parameter values to the settings for a manual simulation Tip pushing the Best MC Best Automatic button will copy the parameter values to the corresponding textboxes used in manual simulations Figure 23 You must always end with a manual calibration run in order to see the results 7 7 3 Distributed parameter values FyrisNP has a possibility to use different parameter values for different subcatchments This gives the possibility to do a step wise calibration of co and kvs starting in the headwaters and continuing downstream This function is only recommended for advanced us
36. hments and time steps months or weeks number of measurement stations for water discharge Q stations and nutrient concentration c stations and number of lake models included If information is given in the Comments or Settings tab it will also be shown here 7 5 The Data tab Under the data tab one can find data tables corresponding to the worksheets of the input data Excel file Figure 16 by clicking the corresponding button on the left These data tables are useful for checking that the import of data worked well As of version 3 1 of FyrisNP the data can also be visualised in a tree plot by pressing the Catchment view button in the upper right corner 22 3 FyrisNP Menu About Current project test2 General Data Q data Calibration Scenario Result 14 954723 28 887464 0 669911004036 72 243167 0 10 330802 0 001980523849 58 278824 0 114872600173 3 636629 0 333188413290 4 675659 0 12 923981 0 002717009884 31 63073 3 426793075502 13 085085 3 162783402920 76 436122 2 786612055847 24 921583 63 934125 33 627515 12 400127 Figure l6 The Data tab is useful for browsing the input data tables making sure that all information was imported correctly Buttons for Constructed wetlands and External load only show up if such worksheets are present in the input file 7 6 The Q data tab The data under the Q data tab shows
37. ical Models of Small Watershed Hydrology and Applications Edited by V P Singh and D K Frevert Water Resources Publications LLC Chelsea Michigan USA 37
38. igure 2 These data are used for all sub catchments in the model and you have to fill in a row for each time step of the model This means that if the model includes data for 5 years 60 months you have to fill in 60 rows The model requires input data time series to start in January Make sure to include a header row since data is imported from the second row down 11 A B D _1_ Year Month Days of month Temperature 28 1985 1 31 0 3 EA 1985 2 28 12 4 1985 3 31 05 1985 4 30 1 6 1985 5 31 13 3 Hirn 1985 6 30 154 Figure 2 snapshot from a typical Temperature worksheet 5 2 Specific runoff This worksheet contains information on the measured or modelled area specific runoff mm month The first row contains headings and Q station numbers ID The Q station numbers should be in increasing order but need not be consecutive numbers The following rows contain the actual data Figure 3 The Q stations should be situated in the outlet point of a sub catchment A B D EET Year Month Days of month 108003 108004 108005 108007 1985 1 31 16 25 15 88 16 94 17 44 1985 2 28 8 3 10 2 10 7 15 34 1985 3 31 37 9 55 61 51 96 61 17 1985 4 30 131 85 108 3 105 06 101 96 1985 5 31 33 11 23 78 25 84 23 83 1985 6 30 19 85 154 16 84 15 96 1985 7 31 16 86 11 94 14 19 13 51 1985 8 31 21 98 18 95 19 59 292 1985 9 30 40 66 41 78 41 17 70 27 1985 10 31 37 57 35 53 36 15 44 8 1985 11 30 39 43 37 69 37 87 41
39. llows the user to create a new workspace and the delete button allows the user to delete a workspace but ONLY if it is empty Thus the user must delete all projects within the 20 workspace first and then delete the workspace itself The rationale behind this procedure is to reduce the risk of undesired deletions of work projects 7 3 Projects panel The New button creates a new project by prompting the user to import an Excel file The user should browse to the desired Excel file and open it The data in the Excel file will now be imported if the contents fulfil the requirements on it The Delete button is used to delete a project Only one project at a time can be deleted The Copy button is used to copy an entire project to a new project for which the user will be prompted to provide a name settings and results of the old project will be copied to the new one except the text written in the comments box of the General tab The Open button opens the selected project The same action will be taken by double clicking a project in the list ee 99 The project file name may not contain any point characters Several instances of FyrisNP can be open allowing the user to work simultaneously with multiple simulations 7 4 The General tab The general tab provides the user with some significant numbers concerning the project at hand Figure 15 These numbers can be used for a quick check that data was loaded pr
40. n Phosphorus p or phosphorus If the standard type of calculations is chosen one can also write Standard or standard in cell B2 The phosphorus calculations will be similar for both the PLCS and the standard alternative The selection of PLCS will affect how the type specific concentration is written A B Substance Type spec conc type 2 Nitrogen PLCS Figure 7 The Settings worksheet for PLCS type calculations of nitrogen 15 5 7 Type spec conc 5 7 1 Standard calculations This worksheet includes type specific concentrations mg l in runoff from different land use Type specific concentrations in runoff from arable land and pasture are however given in the Catchment worksheet thus allowing different concentrations to be used for different sub catchments The values in the Type spec conc worksheet can vary with season requiring one value for each month of the year 12 rows The land use classes are from the left mountain areas above tree line forests clear cuts mire wetlands open land other open land which is not arable land or pasture and urban areas Figure 8 Suitable values to be used for different land use for Swedish catchments can be found in Brandt et al 2008 The values in Figure 8were used for a model application in the southern part of Sweden Make sure to include a header row since data is imported from the second row down A E F 1 Mountain Forest Clear cuts Mires Openland Urb
41. ncies as well as the correlation coefficients for all Monte Carlo simulations performed within the project table 2 Table2 The content and organization of data in the montecarlo out ASCII file Column 1 2 3 4 5 cr 20 kvs E R empirical calibra upper limit in C empirical calibra model efficiency the linear tion parameter for fortemperature tion parameter for according to Nash correlation temperature dependency hydraulic amp Sutcliffe 1970 coefficient dependency dependency A more detailed description of the parameters in the montecarlo out file is given in the technical description of FyrisNP version 3 2 Wid n Nilsson et al 2012 6 3 Results out Results averages out The results files contain time series with the following columns Catchment ID Year Month Retention Mass flow rate Concentration Station ID Lake model Q Area Mountain Forest 19 Clearcut Mire Farmland Pasture Open Land Built Urban Major point sources Households Minor 1 Minor 2 Lake deposition The retention is the retention occurring in the corresponding sub catchment acting on both the load from the sub catchment itself as well as on the contribution from upstream sub catchments The mass flow rate is the outflow in kg of nitrogen or phosphorous from the sub catchment The concentration is the corresponding sub catchment outflow concentration in mg l The mass flow rate divided by the water
42. nd Q data tabs respectively If there is more than one empty line beneath your imported data remove it paragraph 8 1 If you get error messages when importing the input data file check that all time series data nutrient conc runoff temp point source discharges cover the same time period that the sub catchment network is correct downstream ID s in the Catchment worksheet and that the excel file contains no equations or functions in the imported cells 9 4 Statistics used for calibration In order to evaluate the fit of simulated to measured values two statistical measures are used in the FyrisNP Model the model efficiency Eff and the linear correlation coefficient r Eff 1 implies that the measured and modelled series are identical and 0 indicates that the simulation is no better than a straight line representing the average value of the observations The FyrisNP Model supports two modes of statistical measures calculation 1 One series E and r are calculated based on all selected pairs of observed and simulated concentrations i e all value pairs are lumped before calculation of E and r This is the default setting This calculation gives usually high values of E since the comparison is made against a mean value of the observed values that is far from the individual observation points 2 Multiple series E and r are calculated separately for each selected sub catchment and then the arithmetic average is taken
43. nown parameter combination exists that would result in a model efficiency exceeding the best known simulation with more than 0 01 In general the model efficiency surface is quite smooth Criteria for exiting the Simplex algorithm Maximum number of iterations 15 kys The algorithm searches within the limits specified by the user The automatic calibration makes use of the Simplex algorithm Sorooshian amp Gupta 1995 to find the parameter combination giving the best model efficiency within the user specified parameter interval This automatic calibration algorithm seems very efficient and reliable for the simple model structure in FyrisNP having only two calibration parameters 25 7 7 2 Results and plots in the calibration tab Efficiency and r value for both concentrations and transports are shown after a manual calculation Figure 19 When opening an old project the efficiency and r value of the concentration calculations can be seen but transport efficiency and r value needs to be recalculated Manual calibration kvs Massflux efficiency I N B Preliminary model implementation Values are not stored in a saved project Eff 0 4977 Eff 9 9025 M 0 7070 09349 Figure 19 Efficiency values of concentrations and transports from a manual simulation Results of a manual simulation can be plotted as time series of concentrations or transports in each sub catchment with measureme
44. nstalled and the set up program will create a desktop shortcut as well as a start menu shortcut If your computer is connected to a network it is likely that administrator rights are needed for the installation From version 3 2 0 6 the program will be installed to a folder named with the version number It is thus it is possible to have several program versions installed Installation of new program versions will not overwrite the older program versions The FyrisNP icon in the Windows Start menu points to the most recent program version 9 3 Import data The first thing do to when you want to start working with the FyrisNP model is to import the data needed to run the model The requirements on the input data have been mentioned in chapter 4 of this document A short walkthrough is provided here Open the project manager using the Menu If needed create a new workspace which will contain your projects by clicking the new button in the workspace panel Create a new project by clicking the new button in the projects panel A dialog will emerge and show Excel files in the directory last used on your computer Browse to the Excel file you want to use as data source and open it 34 Even if no error messages appeared during the opening of the Excel file it might be a good idea to look through the imported data to make sure it was imported as expected This can be done by browsing the data tables in the data grid view under the Data a
45. ntrations are used to calculate model efficiency and the correlation coefficient The outcome may be analyzed graphically in the model by means of scatter plots 35 3 The automatic calibration option uses the Simplex algorithm Sorooshian amp Gupta 1995 to find the optimal parameter values within user specified parameter intervals The optimal parameter values are considered to be the ones that provide the highest E value for the chosen calibration set up The result is presented by means of parameter values plus E and r values calculated in accordance with the selected statistical mode as presented in the previous paragraph When you feel confident with the results it is time to start with scenario modelling as described below 9 6 Scenarios The scenarios need to be created by the user by means of altering the input data and then re importing the data to the model to see the changes in output Remember to keep the same parameter values as was found during the calibration process 10 References Brandt M Ejhed amp L Rapp 2008 Naringsbelastningen p stersj n och V sterhavet 2006 Sveriges underlag till HELCOM s femte Pollution Load Compilation Naturv rdsverket Rapport 5815 2008 Nutrient loads to the Swedish marine environment in 2006 Sweden s Report for HELCOM s Fifth Pollution Load Compilation Swedish EPA Report 5815 Johansson J amp Kvarnis 1998 Modellering av n rings mnen i Storsj n och d
46. nts Figure 20 The simulated concentrations or transports can also be plotted versus the corresponding measured values Figure 21 Model efficiency and r values for different c0 and values of the Monte Carlo simulations are plotted when clicking the Plot button when Monte Carlo is selected Figure 22 Calibration plot Catchment 3 Sim VS Meas Transport Time series Sim VS Meas Concentrations eo E e 2 i S o 5 o 36 48 60 72 84 Month number Figure 20 The plot shows a manual time series simulation of nitrogen concentration in catchment 3 26 Calibration plot Catchment 1288 Concentrations Time series Sim VS Meas Transport Time series Sim VS Meas e lt eo x amp 2 a o 2 Nn 200 300 Measured transport kg month 10 3 Figure 21 The plot shows a simulated nitrogen transports versus measured nitrogen transports in a catchment amp Calibration plot Model efficiency Figure 22 The plot shows an example Monte Carlo simulation consisting of 1000 individual simulations Every single simulation is plotted as a dot where the model efficiency is a function of the value of the calibration parameter kvs The Calibration tab includes in addition to the calibration options discussed above radio buttons for selection of statistic approach see paragraph 9 4
47. o 0 5 Due to some changes in the internal model data tables when loading older projects the user should expect to see messages about changing the old project file The model performs these changes automatically However if the original Excel file did not contain Minor Point Sources information for all sub catchments the user must change the input file manually and then start a new project 6 The model can now simulate constructed wetlands former Department of Environmental Assessment 7 The model can also make use of a feature called External Load meaning that in cases where only the lower part of a river system is of interest the upper part can be added as external mass flows if the proper measurements exist This is similar to a Major Point Source except that both water mass flow rate as well as nutrient mass flow rate is added to the designated sub catchment 8 The Q data tab now also provides values of specific runoff flow rate and hydraulic load for every sub catchment 9 An automatic calibration algorithm has been added 10 It is possible to decide which measurement stations to include in the calibrations 11 It is possible to choose between two different ways of computing the model efficiency as well as a statistical measure involving the combined results for linear correlation the model efficiency and the variance In addition to these changes many minor alterations have been undertaken 2 2 Version 3 1 The
48. odel Lake Depth m Model lake mean depth Model Lake Volume m 10 Model lake volume Initial Lake Concentration mg l Initial concentration for model lakes Parameter currently not in use in the model Settlements and urban areas do have the same type specific concentration see 5 7 The outflow ID in the Catchment worksheet contains information on how the sub catchments are connected to each other If for instance sub catchment 1360 is immediately downstream of sub catchment 1348 1360 should be typed into column C Downstream ID column on the row containing the sub catchment 1348 information see example in Figure 5 The outlet area for the whole model catchment should have the downstream ID 1 At present the catchment ID must be in a form that can be converted to an integer i e preferably it should be a number from the start ID s such as 1234 5678 or Catchment1 do not work Large lakes with water turnover time significantly higher than the time step of the model 1 month may be included as Model lakes in the Catchment worksheet Tab 1 For these lakes additional information is needed area volume and initial concentration A yearly mean concentration from the beginning of the calibration period can be used as initial concentration The purpose of the initial concentration is for the model to find a stable modelled concentration as quickly as possible Please note that the initial lake concentration is used even i
49. operly You can also store comments about your project in the large textbox The first lines of information contained herein will be visible in the project manager 21 JF FyrisNP Menu X About Current project Demol Catchment overview General Data Q data Calibration Scenario Result This project uses Monthly data Weekly data subcatchments 16 timesteps 84 c stations spec conc type lake models 0 Q stations 5 Substance PLC5 Comments upper box from Excel text in the lower box can be modifed Omr de Sv rta M nad 2004 2010 Skapad 2011 10 21 av Elin for FyrisNP 3 2 0 19 och 3 2 0 20 Datak lla Samlingsfil Swv rta n 2011 xls Settings skriv PLCS Type spec conc f r hygge org N skog Skogsregion S0 Region ppen SoL Produktionsomrade 60 Fl de S Hype i 5 delaro Temp M tadata COBS M tdata Allt fr n v r databas externt och SAK PLC5 data Delaro Area markany o jordbruk Deposition Dagvatten Typhalter PLCE data Enskilda avlopp F rb ttringar Stora punktkallor manadsvarden 2004 2005 r samma som 2006 ett nytt delaro precis i utloppet Extra flikar 4 demonstration data set from file demo A Dept of Aquatic Sciences SLU andAssessment SLU Figure 15 The General tab is the first thing the user sees of the actual model It contains general information of the projects such as number of sub catc
50. parated text file called results out by clicking the Write file button A second file results averages out with averages for each year and subcatchment each month and subcatchment as well as a total average for each subcatchemt is also written These two files do also contain meta data about the simulation such as input data file name parameter settings and calibration results 8 Special features 8 1 Removal of empty rows in the input data tables If the program imported more rows from the Excel worksheet than are filled with data you can double click on a row header in any data table shown under the data tab to remove all rows where the first cell is empty The model will not run if there is more than one empty row after the last data filled row 8 2 Column mode switch Under the data tab you can change column mode by clicking the salmon coloured little square in the bottom right corner 8 3 Catchment overview graph The catchment overview graph depicts the catchment structure 1 e linking sub catchments and thus showing the water flow directions The sub catchments are visualised as boxes corresponding to the model concept of the FyrisNP model It can be used at anytime by pressing the Catchment overview button in the upper right corner of the program window The boxes can be coloured where the colours represent a user selected variable of interest Figure 28 The catchment overview graph enables immediate identification of sub
51. source apportioned nitrogen and phosphorus transport in the River Fyris catchment in central Sweden Kvarn s 1996 After this first application the model has been further developed in applications for the Lake V ttern catchment Kvarn s 1997 the Lake Storsj n catchment Johansson amp 1998 catchments of coastal areas in Lake V nern Wallin et al 2000 and the River G ta catchment Sonesten et al 2004 During 2005 2006 the platform for the Fyris model was changed from LabView http www ni com labview to Visual Studio and Net Framework http msdn microsoft com netframework This user manual describes the new version of the model released in spring 2012 2 Version history 2 1 Version 3 0 A number of changes and additions have been introduced since version 2 of the Fyris Model Some examples are listed here 1 The model can be used with weekly resolution in time in addition to the standard one month resolution 2 It should no longer be necessary to provide input data using decimal point The conversion from national to U S settings should be taken care of by the model The reason why U S settings are used in the model is that the automatic conversion only works for this setting 3 Column names changed in the Catchment data table such that units are included 4 In the Minor Point Sources Excel worksheet the user must specify a value for all sub catchments even if there is no source if so set equal t
52. sphorus concentrations by adjusting two parameters Root zone leaching from 1 Stream length and width arable and pasture land Z Ci Z clearcuts and wetlands A Initial lake concentration 4 E Observed N and P conc d 1 Temperature 1 Point source discharges Point source discharges Forested land Sub catchment calculations ita em 1 hn us 1 m 1 Net and gross transport 1 Source apportionment 1 Mire wetlands Arable land Urban areas 1 A 1 1 Athmospheric deposition Figure 1 The general structure of inputs and outputs to the FyrisNP model Data used for calibrating and running the model can be divided into time dependent data e g time series on observed nitrogen and phosphorus concentration water temperature runoff and point source discharges and time independent data e g land use information lake area and stream length and width see Figure 1 4 Use of other models to calculate input data In Swedish applications with FyrisNP a major part of the input data is derived from the Swedish Pollution Load Compilation PLC reporting to HELCOM The fifth PLC PLCS is the latest report Brandt et al 2008 Depending of the modelling scale PLC data are usually complemented with local and regional data with higher spatial and temporal resolution Some of th
53. ub catchment to the downstream sub catchment see the column Contribution at outlet of the downmost sub catchment in the Catch contr button 29 amp Results plot Gross and net contribution at sub catchment outlet for entire time period lt o x 2 a z o o v c v L4 o o 9 Catchment ID Figure 25 The bar chart presents the nutrient contribution per sub catchment before gross and after net retention within the same sub catchment 7 11 Sources After the computation the data grid view will contain the gross contribution before retention of the different sources in each sub catchment summed over the entire simulation period The results can be viewed in the plot window Figure 26 amp Results plot Gross contribution per source and sub catchment for entire time period Mountain Forest Clearcut Mire Arable Pasture Open Built Urban Major point sources Households Minor1 Minor2 Lake deposition Gross contribution kg 10 6 Catchment ID Figure 26 The bar chart presents the source composition for every sub catchment 30 7 12 Apportionment Clicking this button will start the computation of the source apportionment of the total upstream net load after retention calculated for the outlet of the sub catchment that is chosen by inserting its catchment ID in the textbox labelled Out in the left margin of the R
54. ug fixes e g a Solved several bugs in the constructed wetlands module b Solved bugs occurring when calculating on a shorter timeperiod than the input data and the shorter timeperiod was implemented more consequent in the different calculations c Deletion of the currently open project is now blocked user have to open another project first there is no option to close the open project d New projects can be opened after another within the same FyrisNP session Previously FyrisNP had to be closed down before opening a second new project One additional major change 15 possibility to do Monte Carlo runs also with the type specific concentrations but this option is currently hidden in the standard versions of the program 3 Model description The dynamic FyrisNP model calculates source apportioned gross and net transport of nitrogen and phosphorus in rivers and lakes The main scope of the model is to assess the effects of different nutrient reduction measures on the catchment scale The time step for the model is in the majority of applications one month and the spatial resolution is on the sub catchment level Retention i e losses of nutrients in rivers and lakes through sedimentation up take by plants and denitrification is calculated as a function of water temperature nutrients concentrations water flow lake surface area and stream surface area The model is calibrated against time series of measured nitrogen or pho
Download Pdf Manuals
Related Search