Streams are modeled using the stream package, and only base flows are represented. Reservoirs are modeled as stream segments with high conductivity and stage set to the observed reservoir stage. An absurdly large flow rate is specified at the inflow to reservoir stream segments to provide a virtually infinite supply of water for reservoir leakage calculations.
Stream flows at selected locations were extracted using the HYDMOD package. Since only baseflows are modeled, the instantaneous model predicted flow rate can be directly compared to the estimated baseflows. The results can be viewed as a hydrograph or as a scatter plot of predicted vs. observed flows.
The annual flow rate is calculated by integrating the model results over all the time steps in each calender year. This can then be directly compared to the annual estimated baseflow.
In order to evaluate winter flows, the model predicted instantaneous result for the first January time step which ends in the middle of January is compared to the January estimated baseflow.
Similarly, in order to evaluate summer flows, the model predicted instantaneous result for the first July time step which ends in the middle of July is compared to the July estimated baseflow.
Model predicted impacts are evaluated exclusively as the predicted impact on stream flows. At each compliance point, the difference in model predicted stream flow between two simulations is calculated. One of these simulations represents historic conditions. The other simulation represents a "what-if" scenario where all pumping and the associated return flows within a state are suppressed or where all surface water imports in the mound area and the associated return flows and canal leakage are suppressed. The graphs show the magnitude of the change relative to the estimated base flows.