out_dt setting is ignored?

[Anatolii Anisimov]

Hi All,

We are in the early stage of using WRF-Hydro in our group. We are conducting a study on flash floods for the Caucasus rivers in the Black Sea region. I would appreciate if some could assist and answer a couple of questions.

1. We run WRF-Hydro v.5.1.1 with hourly input from WRF. Following the
default setup, the model works fine when both the forcing timestep, NOAH
timestep, and output timestep is set to 3600 s in namelist.hrldas. The set
of output files from both LSM module and routing modules is produced every
hour, and the streamflow values are reasonable. However, there is another
output frequency setting within the hydro.namelist file (out_dt). We need a
more frequent output from routing modules, so I tried setting out_dt to 10
minutes. However, it turns out that this hydro.namelist setting is ignored
during runtime: chrtout and rtout files are still produced every hour,
although dt_out is set to 10 minutes. Could it be fixed?

2. I tried an alternative way, setting both the NOAH_timestep and
output_timestep to 10 minutes. In this case, the output fields are written
as expected. However, there is a dramatic decrease in the runoff values of
about an order of magnitude. I see from the documentation that when the
noah_timestep is set smaller than forcing_timestep some linear
interpolation is applied to the forcing data to fill the gaps. Could it be
the issue that the precipitation fields are interpolated incorrectly, leading to lower rainfall accumulations?
Otherwise, I can't understand what drives the streamflow decrease when NOAH
timestep is decreased and other settings preserved.

[Anatolii Anisimov]

Ok, I've looked at the source code. It seems that the output hydro_out subroutine is called by HYDRO_drv with the  output_timestep  frequency and that's it. The out_dt functionality remains unclear, maybe it's a leftover from the previous model versions. 

The second question is still not clear.