Streamflow decreasing with time

[Sherly Shelton]

Hello everyone,

I ran a WRF-Hydro simulation for one month at 4 km resolution using the CONUS404 dataset. The model ran smoothly; however, during my analysis of streamflow, I observed a high peak at the beginning of the simulation followed by an exponential decrease over time.

I am trying to understand whether this behavior is due to insufficient model spin-up or if it is related to an issue with the precipitation forcing data. In the CONUS404 dataset, precipitation is provided as accumulated grid-scale precipitation over the previous 60 minutes, which I converted to mm/s by dividing by 3600.

I have attached the streamflow plots for the simulated period for reference. I would greatly appreciate any guidance on whether this pattern indicates a spin-up issue or a potential problem with the forcing or model setup.

Thank you for your help.

Sherly 

[aubrey]

Hi Sherly:

It is hard to tell the cause without a few more details. How long did you spinup the model, and are you warm starting all components (land model, routed soil water, and baseflow model)? In other words, for a warm start, you should be providing a RESTART file for the land model states, a HYDRO_RST file for the routing states, and you should be specifying rst_typ = 1 and GW_RESTART = 1 (depending on the physics you have active) in the hydro.namelist.

The hydrograph you sent has a very pointed increase and peak, which are odd for model behavior. What are your timesteps? Are you plotting infrequent timesteps and interpolating? Are you simulating a specific rainfall event, or a longer time period with a range of different precip events?

If you can provide some additional info, we can try to help diagnose the issue. We do see cold start runs with this behavior, as well as when we run the model with idealized forcings.

Thanks!

Aubrey

[Sherly Shelton]

Dear Aubrey,

Thank you for your response.

I performed a one-year spin-up simulation from January 2018 to December 2018 and then ran the simulation for the year 2019. For the restart run, I used HYDRO_RST.2018-12-01_00:00_DOMAIN1 in hydro.namelist and RESTART.2018122700_DOMAIN1 in namelist.hrldas.

In addition, I enabled the following restart options:

• GW_RESTART = 1
• rst_typ = 1

I have attached the namelist files for your review. Please let me know if you notice any issues with the restart configuration.

Thank you for your time and assistance.

Best regards,

Sherly