Great differences between SWE values in three snow routines

5 views
Skip to first unread message

amir kk

unread,
Jul 26, 2018, 7:18:15 AM7/26/18
to shyft

 

Hello everyone,

 

           I am doing my master thesis on snowpack on Neanidelva catchment with SHyFT. The goal of this thesis is comparing the SWE and SCA with three snow routines (Snow Distribution_Gamma , Skaugen Snow and HBV Snow). First, I did the calibration for all three routines and then I used the calibrated parameters for simulation and the model generated the SWE of the mentioned routines. Surprisingly, the value of SWE for HBV and Skaugen snow are too small in comparing with Gamma snow. I checked out the values with observed value and found that value of Gamma snow SWE is much more accurate. The range of SWE for Gamma snow ( 0 -600 mm) and for HBV and Skaugen (0 – 25 mm) but they share the same pattern on catchment. It seems that it needs a scaling factor for HBV and Skaugen, something like 20-25 for my case to get close these three SWE values.

I used snow state for HBV and Skaugen and response for Gamma snow, is there any relevant with response, state and the differences between their SWE ?

 

swe = simulator.region_model.gamma_snow_response.swe([],idx)

swe = simulator.region_model.hbv_snow_state.swe([],idx)

swe = simulator.region_model.skaugen_snow_state.swe([],idx)

 

I am looking forward to hearing your help.

 

Best Regards,

 

Amir

 

sigbjorn.helset

unread,
Aug 2, 2018, 5:57:22 PM8/2/18
to shyft
Hi,

I got a issue report from Felix, regarding issue with hbv_snow and skaugen snow for simulation timestep different than 1h.
This is now resolved, and new package for win64 provided on conda channel sigbjorn.


then fix and pull-request that solves the issue


Hope this help progress your work,

Br. Sigbjørn
Reply all
Reply to author
Forward
0 new messages