Error in the observed runoff data in model output file

[Ross Woods]

Hi,

I'm an enthusiastic CAMELS user, and many of my PhD and MSc students are using it. It's a fantastic resource, and it is rare that I come across errors.

 

I've found that observed runoff in the OBS_RUN column of a model output file is incorrectly reported as zero for at least one site, when the measured streamflow are actually missing, not zero. 

Folder for model output:

basin_timeseries_v1p2_modelOutput_daymet.zip\model_output_daymet\model_output\flow_timeseries\daymet\17\09505200_05_model_output.txt

gaugeid: 9505200

Dates: many days from 1982 to 1990, starting from 1982-Oct-01 

Sample line from model output file:

1982 10 01 12   0.0000000   0.0000000   0.0000000   7.1400000   3.1833000   1.9895000   0.0081000   0.0000000

Compare with data from usgs_streamflow\15\09505200_streamflow_qc.txt

Sample line from measured streamflow

09505200 1982 10 01  -999.00 M

So that final blue zero in the model output is actually missing flow data, and should be -999 to be consistent with conventions for missing data. For at least some other gauges, missing data is handled correctly in the model output file, so this is not a generic issue. 

Hopefully you will catch this on the next iteration!

Cheers

Ross