District Surface Water Model (DSWM) Project Training Webinar SWFWMD December 19, 2013
Agenda Agenda Presentation Background DSWM development and calibration Select a watershed to go through Results for all watersheds HSPF Background/Theory Data Model development steps Interfaces Run procedures Recharge Rate Extraction Hands-on Session 212/19/2013
Objectives To develop and calibrate an HSPF*-based numerical model(s) that simulates surface water flow and groundwater recharge in the Northern District Model (NDM) domain area and the entire District. DSWM – District-wide Surface Water Model. To update the existing NDM (Version 3) with recharge and ET estimated by the surface water model. * HSPF – Hydrologic Simulation Program, Fortran 312/19/2013
DSWM Development and Calibration
Model Development 12/19/20135 Segmentation and Land Use Meteorological Data Other Input Data (springs, irrigation, etc.) Characterize Land and Stream Segments Observed Streamflow Data/Calibration Locations Calibration
Land Segmentation 12/19/20136 Basis for land segments is District’s “DrainageBasin” data derived from FDEP Basin-24 coverage Segment size similar to previous models: INTB and IBC Adjustments to boundaries to accommodate waterbody transition (lake vs river), tributary connections, and stream gage locations Non-contributing segments were “disconnected” and assigned parameters that resulted in all rainfall inputs that don’t evaporate going to recharge
Land Use 12/19/20137 Land Use Based on 2004 Florida Land Use from SWFWMD, SRWMD, SJRWMD, and SFWMD Used INTB categories and procedures Categories: Forest, Grass/Pasture, Agriculture/Irrigated, Mining/Other, Urban, Wetland, Urban Impervious, Water Aggregated to model categories using FLUCCS codes Impervious category Effective Impervious Area = directly connected impervious areas EIA computed by percentages of FLUCCS categories All impervious area combined into a single “Urban Impervious” category
Meteorological Data – Rainfall 12/19/20138 Rainfall data is NEXRAD (i.e., radar-derived) data at a 2x2 km grid resolution and 15 minute interval NEXRAD data provided by District Rainfall inputs for PERLNDs and IMPLNDs in each model segment are area-weighted averages of the rainfall values of the pixels overlying the segment Data are stored in WDM file A pilot study on one watershed determined that the NEXRAD data were sufficiently similar to the gage-derived rainfall database used in the INTB model to support a calibrated model Differences between NEXRAD and gage rainfall were observed
Meteorological Data – Potential Evapotranspiration (PET) 12/19/20139 PET data are computed (Priestly-Taylor method) at the same 2x2 km grid as used by NEXRAD PET data are 1-day totals; data are available for June December 2010 PET data developed by USGS and obtained from USGS PET inputs for PERLNDs and IMPLNDs in each model segment are area-weighted averages of the PET values of the pixels overlying the segment Daily PET data were disaggregated to 1 hour time step using a seasonally varying distribution function based on the pattern of daylight at the latitude of the watershed; data stored in WDM file The pilot study determined that the USGS GOES PET data were very similar to the gage-derived PET database used in the INTB model; differences had minimal impact on model results
Irrigation 12/19/ Irrigation is applied to the Agriculture/Irrigated PERLND in each model segment Amounts based on District’s monthly water use permit data for GW and surface water pumping Amounts separated into Spray and Drip categories Monthly totals disaggregated to daily using a “PET deficit” computed from the rainfall and PET data Data are stored in WDM files Daily Spray amount is applied as “rainfall” over three hours starting at 7 AM Daily Drip amount is applied as “surface inflow” (not subject to interception) over six hours starting at 7 AM
Surface Water Pumping 12/19/ Surface water pumping (diversions) computed from District’s monthly water use permit database Monthly surface water pumping totals for all permits in a subwatershed are summed, disaggregated to a constant daily rate, and stored in a WDM file Water is removed from the model reach in the subwatershed where pumping occurs
Springs 12/19/ Flow from springs are added to the model reach in the subwatershed where the springs are located Spring discharge timeseries obtained from USGS and WM Districts and from pre-existing models Springs: Crystal, Rainbow, Silver, Homosassa, Chassahowitzka, Weeki Wachee, Wekiva River watershed springs, Gourdneck, Harris,
Observed Streamflow Data 12/19/ Observed daily streamflow is used to compare with simulated flow during calibration Calibration/comparison performed at 73 gage locations Data obtained from District database, USGS, and SJRWMD Data stored in WDM files
Data Used to Characterize Land Areas and Stream/Lake Reaches 12/19/ DEM – land area slope, stream slope, stream channel Soils – infiltration Land use – segment watershed by landuse/cover NHD – stream and lake locations, sizes, connectivity Existing models – conveyance system connectivity, lake and stream configuration, stream cross sections, HSPF FTABLEs USGS - stream cross sections and rating curves Lake surveys – lake FTABLEs Depth to Groundwater – infiltration, target ET
Calibration 12/19/ Calibration followed the standard hierarchical methodology Focusing first on the overall water balance using the LZSN (lower zone nominal storage), INFILT (infiltration index), DEEPFR (fraction of GW inflow that is lost to recharge), and major ET parameters (e.g., LZETP) Maintain reasonable differences in land-use sensitive parameters within a watershed Compare total actual ET with land-use specific target ET and adjust balance between ET and recharge as necessary Adjust low flow/high flow distribution with INFILT, AGWRC (GW recession), and BASET (baseflow ET)
Calibration (continued) 12/19/ Adjust storm shape using INFILT, UZSN (upper zone nominal storage), INTFW (interflow), and IRC (interflow recession) Comparisons between observed and simulated used in calibration Hydrographs Cumulative flow graphs Flow duration graphs Annual runoff totals Statistics (errors at various flow regimes, correlation coefficient, NS model fit efficiency)
Overview of DSWM Watersheds 12/19/ Area separated into 12 major watersheds to make model input and output manageable and reduce simulation times HAT – Hillsborough River, Alafia River, and other Tampa area watersheds CRY – Crystal River, Pithlachascotee River, Anclote River, Chassahowitzka River, Homosassa River, Weeki Wachi WIT – Withlacoochee River WAC – Waccasassa River OKL – Oklawaha River, including Orange Lake area
Watersheds (continued) 12/19/ KIS – Kissimmee River MAN - Manatee/Little Manatee MSR - Myakka and Sarasota Bay area PCH – Peace River and Charlotte Harbor area WOK - Western Okeechobee CAL - Caloosahatchee River EXT – Extended area - Etonia/Rice Creeks, Lake George area, Wekiva River
Summary of DSWM Development 12/19/ Principal Inputs: NEXRAD 15-minute rainfall and daily computed PET dataset, 2004 Florida Land Use (seven categories – same method as INTB); Segmentation: average size = 44 sq miles Irrigation Input: based on groundwater and surface water pumping data; same method as INTB Calibration: ~75 calibration locations ET Comparisons: within error bounds of target ET (+/- 10%) for much of the area Model Performance: overall "fair" calibration to daily streamflow, monthly average flows show best results
Watersheds and HSPF Models 12/19/201320
North Domain Model Expanded into St Johns River watershed 12/19/201321
Final Model Segmentation 12/19/201322
Watersheds in Northern Region
Hillsborough -Alafia- Tampa Watershed 12/19/201324
Crystal- Pithlachascotee and Withlacoochee Watersheds 12/19/201325
Oklawaha Watershed 12/19/201326
Extended Area Watersheds – Rice Creek, Etonia Creek, Lake George Area creeks, Wekiva River, and local drainage to St. Johns River 12/19/201327
Northern Region Calibration Results - Observed Annual Flows (inches) and Simulated Error Terms 12/19/ Annual Flow25% High Flow50% High Flow50% Low Flow25% Low Flow Site NameSite ID Obs (in) Rel Err (%) Obs (in) Rel Err (%) Obs (in) Obs (in) Rel Err (%) Obs (in) Rel Err (%) Obs (in) Hillsborough River Area Blackwater Ck nr Knights Hillsborough R nr Zephyrhills Hillsborough R at Morris Br Cypress Ck at Worthington Gardens low- Cypress Ck nr Sulphur Springs low- Sweetwater Ck nr Tampa Rocky Creek nr Sulphur Springs Brooker Ck nr Tarpon Springs low- Alafia River Area Bullfrog Creek nr Wimauma North Prong Alafia R at Keysville South Prong Alafia R nr Lithia Alafia R at Lithia Crystal-Pithlachascotee River Area Anclote R nr Elfers Pithlachascotee R nr New Port Richey Withlacoochee River Withlacoochee R nr Cumpressco low- Withlacoochee R at Trilby Little Withlacoochee R nr Tarrytown low- - Little Withlacoochee R at Rerdell low- Withlacoochee R at Croom Withlacoochee R nr Floral City Withlacoochee R nr Holder Waccasassa River Waccasassa R nr Gulf Hammock ( ) low- Oklawaha River Big Ck nr Clermont low- Palatlakaha R at Structure M low- Apopka-Beauclair Canal nr Astatula Haynes Ck at Lisbon Ocklawaha R at Moss Bluff Ocklawaha R nr Conner Ocklawaha R at Eureka Orange Ck at Orange Springs Ocklawaha R at Rodman Dam
Northern Region Calibration Results - Statistics of Daily and Monthly Flow Rates 12/19/ Site NameSite ID Daily FlowMonthly Flow Correl. Coeff. Coeff. of Deter. Mean Error (cfs) Mean Abs.Err. (cfs) RMS Error (cfs) Model Fit Effic. Correl. Coeff. Coeff. of Deter. Mean Error (cfs) Mean Abs.Err. (cfs) RMS Error (cfs) Model Fit Effic. Hillsborough River Area Blackwater Ck nr Knights Hillsborough R nr Zephyrhills Hillsborough R at Morris Br Cypress Ck at Worthington Gardens Cypress Ck nr Sulphur Springs Sweetwater Ck nr Tampa Rocky Creek nr Sulphur Springs Brooker Ck nr Tarpon Springs Alafia River Area Bullfrog Creek nr Wimauma North Prong Alafia R at Keysville South Prong Alafia R nr Lithia Alafia R at Lithia Crystal-Pithlachascotee River Area Anclote R nr Elfers Pithlachascotee R nr New Port Richey Withlacoochee River Withlacoochee R nr Cumpressco Withlacoochee R at Trilby Little Withlacoochee R nr Tarrytown Little Withlacoochee R at Rerdell Withlacoochee R at Croom Withlacoochee R nr Floral City Withlacoochee R nr Holder Waccasassa River Waccasassa R nr Gulf Hammock (99-06) Oklawaha River Big Ck nr Clermont Palatlakaha R at Structure M Apopka-Beauclair Canal nr Astatula Haynes Ck at Lisbon Ocklawaha R at Moss Bluff Ocklawaha R nr Conner Ocklawaha R at Eureka Orange Ck at Orange Springs Ocklawaha R at Rodman Dam
Examples Northern Watersheds
Withlacoochee River near Holder 12/19/ Add
Withlacoochee River near Holder 12/19/
Hillsborough River at Morris Bridge 12/19/
Hillsborough River at Morris Bridge 12/19/
Summary of Examples 12/19/ Both examples are good statistically except for low flows, which are too high in Hillsborough and too low in Withlacoochee Withlacoochee Basin was difficult to calibrate due to large surface storage and groundwater contributions Dry years are over-simulated and wet years are under-simulated, generally NEXRAD rainfall appears to be low in early years and higher than gage rainfall in later years Simulated Steamflow Volume Errors Withlacoochee River - HolderHillsborough River - Morris Br. Obs. Flow (in)Error (%)Obs. Flow (in)Error (%) Total Annual Flow % High Flow % High Flow % Low Flow % Low Flow Statistical Measures Withlacoochee River - HolderHillsborough River - Morris Br. DailyMonthlyDailyMonthly Correlation Coefficient Coefficient of Determination Mean Error (cfs) Mean Abs. Error (cfs) RMS Error (cfs) Model Fit Efficiency (NS)
Watersheds in Southern Region
Manatee and Myakka-Sarasota Watersheds 12/19/201337
Peace-Charlotte Harbor Watershed 12/19/201338
Kissimmee Watershed 12/19/201339
Western Okeechobee and Caloosahatchee Watersheds 12/19/201340
Peace-Charlotte Harbor Watershed 12/19/201341
12/19/ Annual Flow25% High Flow50% High Flow50% Low Flow25% Low Flow Site NameSite ID Obs (in) Rel Err (%) Obs (in) Rel Err (%) Obs (in) Rel Err (%) Obs (in) Rel Err (%) Obs (in) Rel Err (%) Manatee River Area Little Manatee R nr Ft. Lonesome S F Little Manatee R nr Wimauma ( ) Little Manatee R nr Wimauma Manatee R near Myakka Head Gamble Creek near Parrish ( ) Braden R at Lakewood Ranch nr Lorraine Ward Lake nr Bradenton low- Myakka River & Sarasota Area Myakka R at Myakka City Myakka R nr Sarasota Big Slough Canal nr Myakka City Big Slough at Tropicaire Blvd ( ) Walker Ck nr Sarasota Peace River & Charlotte Harbor Peace Ck Drainage Canal nr Wahneta Saddle Ck at St. Hwy 542 nr Lakeland Peace R at Bartow Peace R at Fort Meade Bowlegs Ck nr Fort Meade Peace R at Zolfo Springs Charlie Ck nr Gardner Peace R at Arcadia Horse Ck nr Arcadia Shell Ck nr Punta Gorda Kissimmee River Shingle Ck at Airport nr Kissimmee Reedy Ck nr Vineland Catfish Ck nr Lake Wales Boggy Ck nr Taft Tiger Ck nr Babson Park Livingston Ck nr Frostproof Carter Ck nr Sebring Arbuckle Ck nr De Soto City Josephine Ck nr De Soto City Western Okeechobee Fisheating Ck at Palmdale Fisheating Ck at Lakeport Caloosahatchee River Caloosahatchee R at S-79, nr. Olga Whiskey Ck at Ft. Myers Southern Region Calibration Results - Observed Annual Flows (inches) and Simulated Error Terms
Southern Region Calibration Results - Statistics of Daily and Monthly Flow Rates 12/19/ Site NameSite ID Daily FlowMonthly Flow Correl. Coeff. Coeff. of Deter. Mean Error (cfs) Mean Abs.Err. (cfs) RMS Error (cfs) Model Fit Effic. Correl. Coeff. Coeff. of Deter. Mean Error (cfs) Mean Abs.Err. (cfs) RMS Error (cfs) Model Fit Effic. Manatee River Area Little Manatee R nr Ft. Lonesome S F Little Manatee R nr Wimauma Little Manatee R nr Wimauma Manatee R near Myakka Head Gamble Creek near Parrish ( ) Braden R at Lakewood Ranch nr Lorraine Ward Lake nr Bradenton Myakka River & Sarasota Area Myakka R at Myakka City Myakka R nr Sarasota Big Slough Canal nr Myakka City Big Slough at Tropicaire Blvd Walker Ck nr Sarasota Peace River & Charlotte Harbor Peace Ck Drainage Canal nr Wahneta Saddle Ck at St. Hwy 542 nr Lakeland Peace R at Bartow Peace R at Fort Meade Bowlegs Ck nr Fort Meade Peace R at Zolfo Springs Charlie Ck nr Gardner Peace R at Arcadia Horse Ck nr Arcadia Shell Ck nr Punta Gorda Kissimmee River Shingle Ck at Airport nr Kissimmee Reedy Ck nr Vineland Catfish Ck nr Lake Wales Boggy Ck nr Taft Tiger Ck nr Babson Park Livingston Ck nr Frostproof Carter Ck nr Sebring Arbuckle Ck nr De Soto City Josephine Ck nr De Soto City Western Okeechobee Fisheating Ck at Palmdale Fisheating Ck at Lakeport Caloosahatchee River Caloosahatchee R at S-79, nr. Olga Whiskey Ck at Ft. Myers
Examples Southern Watersheds
Peace River at Arcadia 12/19/201345
Peace River at Arcadia 12/19/
Braden River near Lorraine (Manatee) 12/19/201347
Braden River near Lorraine (Manatee) 12/19/201348
Summary of examples 12/19/ Peace River is good statistically except for low flows, especially in two dry years (2000 and 2006) Braden River (Manatee tributary) was difficult to calibrate; possibly because watershed/reach storage is under estimated, since peaks are early Braden River low flows are over-simulated Simulated Steamflow Volume Errors Peace River - ArcadiaBraden River - Lorraine Obs. Flow (in)ErrorObs. Flow (in)Error Total Annual Flow % % 25% High Flow % % 50% High Flow % % 50% Low Flow % % 25% Low Flow % % Statistical Measures Peace River - ArcadiaBraden River - Lorraine DailyMonthlyDailyMonthly Correlation Coefficient Coefficient of Determination Mean Error (cfs) Mean Absolute Error (cfs) RMS Error (cfs) Model Fit Efficiency (NS)
Sample – Little Manatee River near Wimauma 12/19/201350
Little Manatee River near Wimauma 12/19/201351
Little Manatee River near Wimauma 12/19/201352
Little Manatee River near Wimauma 12/19/201353
Little Manatee River near Wimauma 12/19/201354
Little Manatee River near Wimauma 12/19/201355
Little Manatee River near Wimauma 12/19/201356
Little Manatee River near Wimauma 12/19/201357
Little Manatee River near Wimauma 12/19/201358
Little Manatee River near Wimauma 12/19/201359
Little Manatee River near Wimauma 12/19/201360
Little Manatee River near Wimauma 12/19/201361
Little Manatee River near Wimauma 12/19/201362
Little Manatee River near Wimauma 12/19/201363
Little Manatee River near Wimauma 12/19/201364
Other Calibration Metrics 12/19/ Water Balance Target ET Baseflow Separation
Water Balance (inches) Hillsborough River at Morris Bridge near Thonotosassa 12/19/ ComponentForest Grass/ Pasture Agri- culture/ Mining/ Other UrbanWetland Imper- vious Water Weighted Average Influx Rainfall Irrigation Overflow from Withlacoochee R Point Source Flow from Crystal Springs Runoff Surface Interflow Base flow Total GW Inflow Deep Active Evaporation Potential Interception Storage Upper Zone Lower Zone Ground Water Base flow Total Target
Area-Weighted Target ET (inches/year) and Simulated Errors for Selected Watersheds 12/19/ Site NameSite IDForest Error (%) Grass/ Pasture Error (%) Agric./ Irrigated Error (%) Mining/ Other Error (%) Urban Error (%) Wetland Error (%) Imperv Error (%) Hillsborough River Area Hillsborough R at Morris Br %32.912%41.00%42.0-7%35.33% % 15.9% Cypress Ck nr Sulphur Springs %31.913%42.5-6%42.0-9%35.20% % 15.7% Alafia River Area Bullfrog Creek nr Wimauma %32.13% % % %47.7-7% 15.11% Alafia R at Lithia %31.97% %42.0-5%36.0-8% % 15.7% Crystal-Pithlachascotee River Area Anclote R nr Elfers %32.6-4%41.78%42.016%36.35%48.318% 15.25% Withlacoochee River Withlacoochee R nr Cumpressco %32.910%44.31% %37.1-7% % 15.11% Little Withlacoochee R at Rerdell %33.613%42.64%42.0-9% % % 15.47% Withlacoochee R nr Holder %29.231%38.217%42.0-8%31.711% % 15.41% Waccasassa River Waccasassa R nr Gulf Hammock %26.543%43.51% %30.323%42.3-6% 15.10% Oklawaha River Big Ck nr Clermont %26.045% % %30.025%40.0-6% 15.33% Apopka-Beauclair Canal nr Astatula %26.037%40.5-5% %30.119% % 15.30% Haynes Ck at Lisbon %26.041%38.71% %30.022% % 15.28% Ocklawaha R nr Conner %26.145%40.2-3% %30.125% % 15.30% Orange Ck at Orange Springs %35.26%26.150% %30.127%40.6-7% 15.27% Manatee River Area Little Manatee R nr Wimauma %33.77%44.1-9% %37.6-4% % 15.14% Manatee R near Myakka Head %30.613% % %34.5-2%45.9-1% 15.17% Myakka River & Sarasota Area Myakka R nr Sarasota %30.819%43.8-5% %34.29% % 15.17% Peace River & Charlotte Harbor Saddle Ck at St. Hwy 542 nr Lakeland %32.716%40.219% %36.24% % 15.32% Bowlegs Ck nr Fort Meade %3122%44.80%n/a 35.36% % 15.27% Charlie Ck nr Gardner %29.922%42.82% %33.85% % 15.33% Peace R at Arcadia %30.521%43.90% %34.96% % 15.45% Horse Ck nr Arcadia %30.325%44.12% %34.111%45.5-5% 15.35% Kissimmee River Shingle Ck at Airport nr Kissimmee %30.02% % % % % 15.20% Catfish Ck nr Lake Wales %28.127% % %31.712%41.40% 15.32% Tiger Ck nr Babson Park %28.426% % %31.713%40.66% 15.33% Carter Ck nr Sebring %29.424%44-9%n/a 33.58%43.52% 15.47% Josephine Ck nr De Soto City %29.121% % %32.78%43.3-5% 15.33% Western Okeechobee Fisheating Ck at Lakeport %30.211% % %33.8-2% % 15.14% Caloosahatchee River Caloosahatchee R at S-79, nr. Olga % %45.0-1%42.011%37.0-2%47.515% 15.18%
Baseflow Separation - Mean Annual Baseflow (Observed vs. Simulated) for Selected Gages 12/19/ Site NameSite IDObservedSimulatedDifference Percent Difference Hillsborough R at Morris Br % Alafia R at Lithia % Withlacoochee R nr Holder % Little Manatee R nr Wimauma % Manatee R near Myakka Head % Myakka R nr Sarasota % Big Slough Canal nr Myakka City % Peace R at Bartow % Peace R at Arcadia %
Model Performance Summary 12/19/ Statistical measures of calibration indicate that many of the main river basins are fairly well calibrated at the outlets; correlation coefficients for daily flows at these gages range from 0.75 (Poor) for the Peace River to 0.91 (Good) for the Withlacoochee River, based on typical hydrology calibration criteria. Average daily correlation coefficient at all gages: The monthly flow correlation coefficients at the major gages are generally better, and are all above 0.90 (Good). Results at the smaller gages are variable, with some poor statistical comparisons and some good. Average: Errors in total flow and 25%/50% high flows are small, and flow frequency comparisons are good at higher flows Some of the calibration error is attributable to apparent errors in the principal hydrological input (rainfall). The NEXRAD rainfall dataset appears to be lower than gage-based rainfall in the earlier years of the model calibration period (WY ), and higher in the later years, specifically it is higher in , and lower prior to 2004 Low flow years are over-simulated and high flow years are under-simulated in many watersheds
HSPF: HYDROLOGIC SIMULATION PROGRAM - FORTRAN Continuous simulation model Natural and developed watersheds and water systems Land surface and subsurface hydrology and quality processes Stream/lake hydraulics and water quality processes Time series data management and storage Core watershed model in EPA BASINS Development and maintenance activities sponsored by U.S. EPA and U.S. Geological Survey
CONTINUOUS SIMULATION Represent hydrologic processes, storages, and pathways for a watershed, continuously for many days to multiple years, with time steps of one day or less, usually in the range of minutes to hours
72 of 32 RESULTS FROM CONTINUOUS SIMULATION Daily FlowFlow Duration/Frequency Storm Hydrographs
HSPF APPLICATION & UTILITY MODULES PERLND IMPLNDRCHRES Snow Hydraulics Water Conservative SedimentSolidsTemperature Quality Sediment PesticideNonconservative NitrogenBOD/DO PhosphorusNitrogen TracerPhosphorus Carbon Plankton BMP Flow Any constituent simulated in PERLND, IMPLND or RCHRES APPLICATION MODULES UTILITY MODULES COPY, MUTSIN, PLTGEN, DURANL, GENER, DISPLY, REPORT
HSPF - STRENGTHS Comprehensive representation of watershed land and stream processes Comprehensive representation of watershed pollutant sources, including nonpoint sources (by multiple land uses), point sources, atmospheric, etc. Flexibility and adaptability to a wide range of watershed conditions Well-designed code modularity and structure Companion database and support programs to assist model users (e.g., BASINS, WDMUtil, WinHSPF, GenScn, HSPEXP+) Development and support by U.S. EPA and USGS
HSPF - LIMITATIONS AND WEAKNESSES Extensive data requirements User training normally required Lack of comprehensive parameter guidance Large number of parameters Limited spatial definition (i.e., lumped parameter approach) Hydraulics limited to non-tidal freshwater systems and unidirectional flow Simplified representation of urban drainage systems
HSPF APPLICATION PROCESS Study definition Development of modeling strategy Learn how to use HSPF Development of time series data Parameter development Calibration/validation Analysis of alternate scenarios
HSPF HYDROLOGY PROCESSES PARAMETERS CALIBRATION
78 of 35 THE HYDROLOGIC CYCLE
HYDROLOGIC COMPONENTS Hydrologic Components: Rainfall Interception Depression storage Evapotranspiration Infiltration Surface storage Runoff Interflow Groundwater flow Evapotranspiration Interceptio n Depression storage Ground surface Capillary rise Precipitation Infiltration Soil moisture Percolation Groundwater storage Underground flow into or out of the area Ground water flow Streamflow Interflow Surface runoff Channel pptn.
WATER BALANCE Water balance equation R = P - ET - IG - Δ S where:P=Precipitation R=Runoff ET=Evapotranspiration IG=Deep/inactive groundwater Δ S=Change in soil storage Inter-relationships between components Variation of components with time consideration of soil condition, cover, antecedent conditions, land practices
STANFORD WATERSHED MODEL To Stream Actual ET Potential ET Precipitation Temperature Radiation Wind, Dewpoint Snowmelt Interception Storage Lower Zone Storage Groundwater Storage Interflow Upper Zone Storage Overland Flow Deep or Inactive Groundwater CEPSC* BASETP* AGWETP* DEEPFR* LZSN* INFILT* INTFW* UZSN* AGWRC* NSUR* SLSUR* LSUR* IRC* Delayed Infiltration Direct Infiltration PERC 1 ET LZETP* * Parameters Output Process Input Storage ET -Evapotranspiration n Order taken to meet ET demand Decision
PWATER PARAMETERS CEPSC - Interception storage capacity NSUR - Manning’s N for overland flow plane LSUR - Length of assumed overland flow plane SLSUR - Slope of assumed overland flow plane INFILT - Index to infiltration capacity of the soil UZSN - Upper zone nominal soil moisture storage LZSN - Lower zone nominal soil moisture storage LZETP - Lower zone ET parameter; index to density of deep-rooted vegetation
PWATER PARAMETERS INTFW - Interflow inflow parameter IRC - Interflow recession parameter (the ratio of interflow outflow rate today / rate yesterday) KVARY - Variable groundwater recession parameter AGWRC - Basic groundwater recession rate (when KVARY is 0) DEEPFR - Fraction of groundwater inflow which will enter deep (inactive) groundwater BASETP - Fraction of remaining potential ET which can be satisfied from baseflow AGWETP - Fraction of remaining potential ET which can be satisfied from active groundwater storage
PWATER PARAMETERS INTFW - Interflow inflow parameter IRC - Interflow recession parameter, i.e., the ratio of interflow outflow rate today / rate yesterday LZETP - Lower zone ET parameter; an index to the density of deep-rooted vegetation
Mean runoff volume for simulation period (inches) Annual and monthly runoff volume (inches) Daily flow timeseries (cfs) observed and simulated daily flow scatter plots Flow frequency (flow duration) curves (cfs) Storm hydrographs, hourly or less, (cfs) CALIBRATION/VALIDATION COMPARISONS “Weight-of-Evidence” Approach
Precipitation Total Runoff (sum of following components) Overland flow Interflow Baseflow Total Actual Evapotranspiration (ET) (sum of components): Interception ET Upper Zone ET Lower Zone ET Baseflow ET Active Groundwater ET Deep Groundwater Recharge/Losses CALIBRATION/VALIDATION COMPARISONS Water Balance Components
Graphical Comparisons: Timeseries plots of observed and simulated values for fluxes (e.g., flow) or state variables (e.g., stage, sediment concentration, biomass concentration) Observed and simulated scatter plots, with 45 o linear regression line displayed, for fluxes or state variables Cumulative frequency distributions of observed and simulated fluxes or state variable (e.g., flow duration curves) Statistical Tests: Error statistics, e.g., mean error, absolute mean error, relative error, relative bias, standard error of estimate, etc. Correlation tests, e.g., correlation coefficient, coefficient of model- fit efficiency, etc. Cumulative Distribution tests, e.g., Kolmogorov-Smirnov (KS) test CALIBRATION/VALIDATION COMPARISONS Graphical/Statistical Procedures & Tests
R & R 2 VALUE RANGES FOR MODEL PERFORMANCE Criteria PoorFairGoodVery Good PoorFairGoodVery Good R 2 Daily Flows Monthly Flows R
HYDROLOGIC (PWATER) CALIBRATION Annual Water Balance - Runoff = Prec. - Actual ET - Deep Perc. - Δ Storage Key Parameters:Repre. Precipitation (MFACT) LZSN LZETP INFILT DEEPFR Groundwater (Baseflow) Volume and Recession - Runoff = Surface Runoff + Interflow + Baseflow Key Parameters:INFILT AGWRC/KVARY DEEPFR BASETP/AGWETP Surface Runoff + Interflow (Hydrograph Shape) - Key Parameters:UZSN INTFW IRC LSUR, NSUR, SLSUR
IMPERVIOUS AREA - IWATER IWATER Simulate water budget for impervious land segment Simulate moisture retention Determine how much of the moisture supply runs off Evaporate from retention storage
IWATER CALIBRATION Impervious area process IWATER parameter Interception RETSC – retention storage Overland flow/runoff LSUR, NSUR, SLSUR Evaporation (no parameter, occurs at PET)
HSPF INTERFACES WinHSPF –Interactive interface to HSPF –Access to all HSPF Features –Scenario development WinHSPFLt –Interface to HSPF (batch) HSPEXP+ –Hydrologic calibration support
HSPF SUPPORTING PROGRAMS BASINS –Build/maintain WDM time series file and meteorologic data for HSPF and other models –Meteorologic data generation and fill-in –Graphical and tabular display of time series data –Etc. WDMUtil (becoming obsolete) –Build/maintain WDM time series file and meteorologic data for BASINS –Meteorologic data generation and fill-in –Graphical and tabular display of time series data GenScn (becoming obsolete) –Graphical and tabular display of time series data
FILES UTILIZED BY HSPF UCI (User’s Control Input) - contains all input except time series data Run Interpreter Output (ECH) - output summary of user’s input Operation Module Output - state variables and fluxes at user-selected intervals for PERLND/IMPLND/RCHRES WDM - time series data input and output (binary format) PLTGEN/MUTSIN/SEQ - Time series data input and output (text format) HSPF Binary Output - Operation module output in binary format Basin Specifications (EXS) - contains information for the HSPF hydrology calibration support program HSPEXP+
Recharge and ET Estimation
STANFORD WATERSHED MODEL To Stream Actual ET Potential ET Precipitation Temperature Radiation Wind,Dewpoint Snowmelt Interception Storage Lower Zone Storage Groundwater Storage Interflow Upper Zone Storage Overland Flow Deep or Inactive Groundwater CEPSC* BASETP* AGWETP* DEEPFR* LZSN* INFILT* INTFW* UZSN* AGWRC* NSUR* SLSUR* LSUR* IRC* Delayed Infiltration Direct Infiltration PERC 1 ET LZETP* * Parameters Output Process Input Storage ET -Evapotranspiration n Order taken to meet ET demand Decision ETVadose ETSat IGWI AGWI AGWET BASET 9612/19/2013
97 AGWI IGWIBASET AGWET
Infiltration, Effective Recharge and ET in MODFLOW I = AGWI + IGWI Vadose Zone R’ I ET Sat Groundwater I = Precip Eff - ET Vadose 9812/19/2013 ET sat = AGWET + BASET =
12/19/ Extinction Depth ET at depth d below ET surface ET Surface DEDE d ET From Groundwater ET Max
12/19/ BASINS 4.1 *_REC.WDM File Subbasin AGWI, IAGWI, AGWET, BASET Arc/MAP Shapefiles Landuse, Grid Grid AGWI, IAGWI, AGWET, BASET GW Model Files RCH, EVT
12/19/ BASINS 4.1 *_REC.WDM File Subbasin AGWI, IAGWI, AGWET, BASET Arc/MAP Shapefiles Landuse, Grid Grid AGWI, IAGWI, AGWET, BASET GW Model Files RCH, EVT
Processing 12/19/ Model_REC.WDM Text_Grid_Format.txt Subbasin AGWI, IAGWI, AGWET, BASET
Step 1: Open BASINS 12/19/
Step 2: Select File Type 12/19/ File>Open Data Select a File Type>WDM Time Series> OK
Step 3: Select File 12/19/ Select File> Open
Step 4: Open Data File 12/19/ Select by clicking on Filename Analysis>List File>Manage Data
Step 5: Select Data Sets 12/19/ Yes W A I T…………
Step 6: Select Data to List 12/19/ OK
Time Series List 12/19/
Adjust Formats 12/19/ Time Series List View> Date and Value Formats
Selected Formats 12/19/
Select Attributes to be Included 12/19/ Time Series List File>Select Attributes
Save Data File as a Text File 12/19/ File>Save Grid as Text
Save Data File 12/19/
Saving Data File 12/19/
12/19/ BASINS 4.1 *_REC.WDM File Subbasin AGWI, IAGWI, AGWET, BASET Arc/MAP Shapefiles Landuse, Grid Grid AGWI, IAGWI, AGWET, BASET GW Model Files RCH, EVT
Transfer of Arrays to Grid Cells using ArcMAP 12/19/ Area-Weighted Average of Infil and ET Apply intersection of Polygons to determine contributions of respective units Upscale from LU SubBasin Basin Cell SUBASIN i in BASIN j SUBBASIN m in BASIN n LU1 LU2 GRID CELL BASIN n BASIN j
12/19/
Plots 12/19/ Annual 1995, 2000, 2004 (Net Recharge) Annual 1995, 2000, 2004 (I 2) Annual 1995, 2000, 2004 (ET 2)
1995 Net Infiltration
2000 Net Infiltration
2004 Net Infiltration
1995 Total Infiltration
2000 Total Infiltration
2004 Total Infiltration
1995 ET
2000 ET
2004 ET
NonNorthern District Model 12/19/
12/19/ Northern Basin Eastern Basin Central Basin
12/19/
12/19/
12/19/
12/19/ BASINS 4.1 *_REC.WDM File Subbasin AGWI, IAGWI, AGWET, BASET Arc/MAP Shapefiles Landuse, Grid Grid AGWI, IAGWI, AGWET, BASET GW Model Files RCH, EVT
Hands-on Session 12/19/ Install software BASINS HSPEXP+ Model files Learn Concepts and How to Make Changes Programs Model files Outputs generated Make a run WinHspfLt or HSPEXP+ Review Output Graphs Statistics and Errors Water Balance and other Summaries
Install Software 12/19/ Download BASINS package Change file extension to exe Uninstall any previous versions of BASINS Run the file as an administrator to install Download HSPEXP+ Unzip the file and run setup.exe to install Download Model files Unzip the file Copy the three separate directories (HAT, EXT, OKL) to the C:\BASINS41\modelout directory
Learn Concepts 12/19/ Watershed Model Files (XXX = three letter code for watershed) XXX.uci – primary input file for watershed Contains all parameters, option flags, land use, watershed connectivity, and instructions for reading input and writing output Text file; column-specific inputs in all active lines; three * indicate a comment line Experience with HSPF and access to User’s Manual are “necessary” XXX.wdm – time series input and output data for watershed Binary file; access/maintain with BASINS or WDMUtil Multiple datasets XXX_MET.wdm - WDM file containing meteorological time series input data for watershed XXX_IRR.wdm – WDM file containing irrigation and surface water pumping data for watershed
Watershed Model Files (continued) 12/19/ XXX_REC.wdm – WDM file containing recharge-related output for watershed XXX.exs – one or more “basin description” files for defining information used by the HSPEXP+ program to generate graphical and statistical output at specific “outlet” locations in the watershed XXX.out – PERLND, IMPLND, RCHRES output summaries at user- defined intervals (text file) XXX.hbn - PERLND, IMPLND, RCHRES output summaries at user- defined intervals (binary file) for use by post-processing programs XXX.ech – output file; echo/summary of all info in the UCI file; contains error/warning messages
Programs (review from earlier) 12/19/ BASINS – display time series data graphically and listing; build/maintain WDM files and datasets; generate meteorological data; much more! HSPEXP+ - runs simulation; generates statistical and graphical output at locations in watershed; provides hydrology calibration advice WDMUtil – build/maintain WDM files and datasets; generate meteorological data WinHspf – run HSPF models; edit UCI file; implement scenarios; set up output locations in model WinHspfLt – run HSPF model (batch)
Outputs Generated by HSPF 12/19/ XXX.out – textfile summaries of state variables and fluxes at user- defined intervals for HSPF operations (PERLND, RCHRES, IMPLND) XXX.hbn – binary file version of XXX.out (used by HSPEXP or BASINS to generate statistics and other outputs) XXX.ech – echo of UCI file information; error/warning messages XXX.wdm – HSPF writes timeseries output to the primary WDM file (flow rates and other information specific to subwatershed outlets) XXX_REC.wdm – recharge-related timeseries output (AGWET, BASET, AGWI, IGWI)
Outputs Generated by HSPEXP+ 12/19/ Look in subdirectory with name: Reports_ Hydrographs Full period, Annual, Log, Monthly Flow Duration graphs Cumulative Flow graphs Statistical Summaries - all are text files DailyMonthlyFlowStats - *.txt AnnualFlowStats - *.txt MonthlyAverageFlowStats - *.txt Expert System Statistics - *.sts Expert System Calibration Advice - *.adv
Examples - graphs 12/19/
Examples – Daily Flow Stats 12/19/ RCH418: Daily Simulated Observed Count 4,110 4,110 Mean Geometric Mean Note: TS, Time Series Correlation Coefficient Coefficient of Determination Mean Error Mean Absolute Error RMS Error Model Fit Efficiency Table 1 TS 1 - EXT RCH418 SIMQ TS 2 - WEKIVA RIVER NEAR SANFORD, FL
Examples – Expert System Statistics 12/19/ Expert System Statistics for EXT.uci Simulation Period: 3 Water years from 2003/10/01 to 2006/09/30 Site: RCH507 Total (3 year run) Observed Simulated Simulated Simulated Total Runoff Total Runoff Surface Runoff Interflow total (inches) = 10% high (inches) = 25% high (inches) = 50% high (inches) = 50% low (inches) = 25% low (inches) = 10% low (inches) = storm volume (inches) = average storm peak (cfs) = baseflow recession rate = summer volume (inches) = winter volume (inches) = summer storms (inches) = winter storms (inches) =
Make a Model Run with WinHspfLt 12/19/ Open C:\Basins41\modelout\EXT directory Browse to UCI file Drag file to WinHspfLt icon on desktop Alternatively, doubleclick WinHspfLt icon, browse to UCI file, and click Open
Make a Model Run with HSPEXP+ 12/19/ Start – Programs Browse to HSPEXP+ and start the program Use Browse button to browse to the UCI file; click Open Toggle (√) the following 5 items: Run WinHspfLt before calculating the statistics Calculate Expert Statistics Produce Land Use and watershed area reports Produce standard monthly flow…graphs Produce Water Balance reports Enter the following numbers in the red-bordered box: 312,316,415,418,434,507,508 (these are outlet reach ID numbers where observed flow data are available) Click Start
HSPEXP+ 12/19/
Create Time Series Graphs with BASINS 4.1 Open BASINS from Start Menu.
Go to File Open Data Select WDM Time Series Click OK
Browse to Desired WDM Time Series Click OK
View the Data File(s) opened by BASINS 4.1 The file is opened by BASINS 4.1 and you may view the file by going to File Manage Data
Go to Analysis Graph From the Data Sources Window or BASINS 4.1 window.
Click No at the selection Window
Data Selection Window
Select Desired Locations and Constituents to see the matching data You can view more attributes of Data by Clicking on Attributes Add and then selecting the desired attributes.
Select Desired Matching Data by Clicking on it. Selected Data shows up on Selection Pane near bottom
Verify the Time Periods, Time Step; Change if Needed, and Click OK.
Select the Types of Graphs Desired and Click Generate
Selected Graphs are Produced
Edit Graph This window can be used to edit colors, scale, legend, font, add lines etc.
On Scatter Plot go to Edit Graph Lines
Click on Add with Equation Y (second Add Button) to get a Y=X line Edit the label and color of the new line. Click on Apply.
Make a model run to see the effects of parameter changes 12/19/ EXT model Adjust parameters for the PERLNDs that contribute to RCHRES 415 (Little Wekiva River) RCHRES 415 represents SJRWMD gage at Springs Landing Increase INFILT (PWAT-PARM2 table) to try to reduce the peak flows; make use of comments in UCI file to locate the PERLNDs that contribute to RCHRES 415 Save modified UCI file and then run the model and generate outputs using HSPEXP+ Compare hydrographs, duration plot, cumulative flow graph, and expert statistics for RCHRES 415 for the original run and your new run See next slide for instructions
Run HSPEXP+ 12/19/ Make changes to UCI file with a text editor Save file Start HSPEXP+ Browse to EXT.uci Check the boxes shown in the figure Enter 415 in the red box Click Start Review/compare results for the two runs