SWAT Modeling of Priority Watershed- Phase II
description
Transcript of SWAT Modeling of Priority Watershed- Phase II
SWAT Modeling of Priority Watershed- Phase II
2013 NPS Conference – September 19, 2013
Dharmendra SaraswatAssoc. Prof,/Ext. Engineer-
Geospatial501-681-5987 (mobile)[email protected]
Naresh PaiPost-Doctoral Associate
Mike DanielsExtension Water Quality and
Nutrient Management Specialist
Tom RileyInterim Assistant Director- CED and
Director, Public Policy Center
Project# 11-900
OBJECTIVE
Project Objective
Prioritize 12-digit HUCs using SWAT model output
2013 NPS Conference – September 19, 2013
BACKGROUND
Phase II(2011 – 2013)
SWAT Modeling Project12-digit HUC Prioritization
1
2
3
2013 NPS Conference – September 19, 2013
OVERALL PROJECT APPROACH
Land-use/management
practices
SWATSWAT
Topography
Soils
Temperature Precipitation
Calibration/Validation
Calibration/Validation
Prioritization(Flow-weighted concentration)
Prioritization(Flow-weighted concentration)
PBIASNSER2
RSR
Point Sources
2013 NPS Conference – September 19, 2013
BEST MODELING PRACTICES (BMPS)
Sensitivity analysis
Check potential model problems using SWAT Check*
Annual calibration
Monthly calibration/validation
Qualitative comparison with AWRC data (monthly data from Oct 2011-Sept 2012 and then storm samples through March 2013)
Uncertainty analysis
Subwatershed prioritization*White et al., 2012
2013 NPS Conference – September 19, 2013
PRIORITIZATION RESULTSPoteau
2013 NPS Conference – September 19, 2013
Critical source area analysis (CSA) suggested that merely 5% of the watershed area was responsible for 26%, and 34% of the sediment and TP overland loads, respectively. In contrast, the nitrate loading was relatively uniform in this watershed.
UNCERTAINTYPoteau
2013 NPS Conference – September 19, 2013
Results suggest that 55% of observed data at Cauthron and 49% of observed
data at Hackett, respectively was found within 95% confidence interval of the
best simulations as shown by the 95PPU plot.
SWAT model uncertainty band (i.e. 95PPU, shown in green) at Cauthron and its comparison with the observed (shown in blue) and best simulated (shown in red) data.
SWAT model uncertainty band (i.e. 95PPU, shown in green) at Hackett and its comparison with the observed (shown in blue) and best simulated (shown in red) data.
PRIORITIZATION RESULTSStrawberry
2013 NPS Conference – September 19, 2013
Critical source area analysis (CSA) suggested that merely 5% of the watershed area was responsible for 85%, 22%, and 16% of sediment, TP, and NO3-N loads, respectively.
PRIORITIZATION RESULTSUpper Saline
2013 NPS Conference – September 19, 2013
Critical source area (CSA)
analysis indicated that merely 5% of
the watershed area was responsible
for 28%, and 13% of the sediment
and TP overland loads, respectively.
In contrast, the nitrate loading was
relatively uniform in this watershed.
COMPARISON WITH MONITORING DATA
2013 NPS Conference – September 19, 2013
Source: Massey et al., 2013
Overall Summary
1)the load comparisons were favorable across all three watersheds;
2)The mean concentration comparison during base flow conditions at the selected HUC 12 level showed relations in the ranks of the sites within the Poteau and Upper Saline Watersheds for NO3-N and TP, but not TSS. The monitoring data and SWAT output were not related at the Strawberry Watershed.
3)These results increase our confidence in the subwatershed prioritization by SWAT model for the Poteau and Upper Saline Watersheds, but not necessarily for the Strawberry Watershed.
ACKNOWLEDGEMENTS
EPA, Region VI
2013 NPS Conference – September 19, 2013