Hello,

I am developing a lumped daily hydrological model using the WEAP Soil Moisture Method.

The catchment currently generates surface runoff, interflow, flow to groundwater, and groundwater discharge to the river. I calibrated the model using the PEST SCE-UA algorithm.

To maintain a conceptual bucket-type structure, I replaced the standard WEAP groundwater wedge formulation with a groundwater-to-surface-water discharge equation inspired by the GR4J routing store. The standard groundwater wedge did not perform well in my current configuration.

During calibration, the optimized Root Zone Conductivity consistently reaches its lower bound, approximately 3 mm/day, while the maximum NSE remains around 0.5. This suggests that the soil reservoir may be compensating for insufficient temporal routing by slowing both interflow and percolation.

I would like to introduce a conservative daily routing component between the catchment and the river, similar to the CK1/CK2 routing time constants in the NAM model. The routing should delay and smooth runoff without changing its total volume..

Is there a recommended way to implement this in WEAP?

For example, could I use:

-a loss-free linear reservoir;
-a cascade of conceptual reservoirs;
-a discrete unit hydrograph using PrevTSValue();
-or an external function called from a WEAP expression?

I would also need to intercept the catchment runoff before it reaches the river, without double counting it.

Any guidance or example would be very helpful.

Best regards,

Guillaume

Guillaume,

One strategy you could use:

You could achieve this by routing the infiltration arc to a stream object (this object is a fictitious stream and does not represent a stream in real life- that's OK), putting a reservoir on the stream, then routing the fictitious stream into the real stream.  The discharge from the reservoir can be controlled by a flow requirement, which can be made a function of the storage in the reservoir.