Adjusting parameters to increase baseflow

[Jaeger]

Hello All,
I am calibrating a small watershed within the Verde River basin in
Arizona, USA using a 10-year daily USGS streamflow record. Model
evaluation includes both graphical techniques (visual comparison of
the simulated and measured daily hydrograph) and statistics on the
monthly water yield divided by the drainage area. Statistics include
the Nash-Sutcliffe Efficiency coefficient (NSE), the Percent Bias
(PBIAS), RMSE-observations standard deviation ratio (RSR), R squared,
and Wilcoxon Rank Sum comparison of medians.

My main problem is overestimated peak discharge and underestimated
baseflow. The SWAT user guide suggests as a solution to adjust
infiltration, interflow, and baseflow recession parameters, but does
not specifically identify model parameters. I’ve adjusted the curve
number and ESCO, which were identified in the sensitivity analysis as
the most sensitive parameters. I’ve also adjusted the alpha_bf,
gw_delay, and sol_awc.

Adjusted parameter values (specifically alpha_bf, gw_delay, and
sol_awc) result in hydrographs that lack baseflow, but achieve the
best statistics on the monthly values. Looking at other parameters
that might achieve both, I’m wary of being a slave to the calibration
evaluation criteria I've set and settling upon adjusted model
parameter values that are not congruent with each other and don’t make
physical hydrologic sense.

To that end, I have a few questions regarding groundwater parameters.

GQ_QMN: My understanding is that if the shallow aquifer water depth is
equal to or exceeds GW_QMN, return flow can occur. Does this mean,
that at GW_QMN=0 (the model default value), all water that infiltrates
and does not evaporate is available for return flow? When I set it to
1000, it essentially eliminates all baseflow in my model simulation.
Is that water that had been baseflow now lost to the deep aquifer?

Is there a general directional trend when simultaneously adjusting
GW_QMN, GW_REVAP, GW_delay, alpha_BF, or other groundwater or soil
parameters?

Finally, what are parameters that reflect interflow? I’m thinking
parameters that would affect soil evaporation like sol_awc and
something like soil hydraulic conductivity, but what else?

Thanks very much for any advice that can be offered,

Jaeger

[Aman Bhatta]

Did you find your answer?