Skip common site navigation and headers
United States Environmental Protection Agency
BASINS
Begin Hierarchical Links EPA Home > Water > Water Science > Water Quality Tools > Models > BASINS > Training > Flow Calibration Tutorial > Calibration Scenarios > Annual Trends and Water Balance End Hierarchical Links

 

Training

Flow Calibration Tutorial

Calibration Scenarios

Annual Trends and Water Balance

     
  1. Observed Flow Less Than Simulated
    1. Make sure there are no unaccounted flow diversions from the watershed. Compare average annual simulated evapotranspiration (TAET) and potential evapotranspiration (PET). If the ET discrepancy is greater than the flow discrepancy, the key factors regulating ET should be the first modified. Begin with increasing LZETP , the parameter that adjusts for the vigor with which vegetation transpires.
    2. Compare average annual simulated ET (TAET) and potential ET (PET). If the ET discrepancy is less than the flow discrepancy, and assuming the estimate for PET is good, simulated ET cannot be increased very much. Therefore, subsurface losses should be increased. DEEPFR  simulates recharge to deep aquifers. This variable should be increased, but the parameter used should be based on a groundwater study of the area.
    3. Compare average annual simulated ET (TAET) and potential ET (PET). If the ET discrepancy is greater than the flow discrepancy, and the PET and LZET parameters are sufficiently high and the subsurface losses are appropriate, then flow can be decreased by increasing LZSN . This increase in lower zone storage provides greater opportunity for evapotranspiration.
    4. If the discrepancy between simulated and observed flow values cannot be reconciled by adjustments in the parameters noted above, infiltration (INFILT ) may need to be first modified. This is particularly likely is there if peak flows are consistently high and low base flows, or vice/verse. Infiltration shifts drainage from rapid response (surface runoff and interflow) to delayed response (base flow), but it often does not have a strong direct influence on water balance. Rather, it may just be too far off for the other parameters to correct for water balance errors.
  2. Observed Flow Greater Than Simulated
    1. Make sure there are no unaccounted flow diversions into the watershed.
    2. Compare the average annual potential evapotranspiration (PET) and lake evaporation for the region. The latter can be obtained using WDMUtil. If PET is already lower than the lake evaporation, PET should not be reduced further. Rather, decrease the LZETP component of evapotranspiration.
    3. If the ET discrepancy is less than the flow discrepancy, and assuming the estimate for PET is good, simulated ET cannot be decreased very much. Therefore, subsurface losses can be decreased. DEEPFR simulates recharge to deep aquifers. This variable should be decreased, but the parameter used should be based on a groundwater study of the area.
    4. If the PET and LZET parameters are sufficiently low, and the subsurface losses are appropriate, decreasing the lower zone storage parameter (LZSN ) will increase flow by decreasing the opportunity for evapotranspiration.
    5. If the discrepancy between simulated and observed flow values cannot be reconciled by adjustments in the parameters noted above, infiltration (INFILT ) may need to be first modified. This is particularly likely is there if peak flows are consistently high and low base flows, or vice/verse. Infiltration shifts drainage from rapid response (surface runoff and interflow) to delayed response (base flow), but it often does not have a strong direct influence on water balance. Rather, it may just be too far off for the other parameters to correct for water balance errors.
  3. Simulated Peaks Characteristically Lag Behind, Or Are Ahead Of, Observed Peaks
    1. Evaluate LSUR. The longer the assumed overland flow plain, the slower water is calculated to move to the pour point. The slower water moves down a surface, the more opportunity it has to infiltrate the land surface, reducing peak flows. Also, LSUR is assumed to vary with slope, with flat slopes having larger values and vice versa.
    2. Recalculate SLSUR. Large slope values model water to move faster than lower slope values, reducing the opportunity for infiltration. Early peaks can be formed if the SLSUR values are inappropriately high.
    3. Modify NSUR. Manning’s n values for overland flow can vary with seasonal land use changes, but they typically have a relatively small impact on both peak flows and volumes as long as they are within normal ranges.
Introduction | Back to Calibration Scenarios

 

Water Quality Standards | TMDL | Contact basins@epa.gov

 
Begin Site Footer

EPA Home | Privacy and Security Notice | Contact Us