Summary
Snow depth from offline LDAS and coupled LDAS based on CY48R1 has been compared. Open-loop experiment by offline LDAS has positive biases around the edge of snow-covered areas against coupled LDAS. However, the differences become smaller by assimilating IMS in offline LDAS. Offline snow depth analysis will make it possible to be more consistent between initial conditions for reforecast and operational forecast.
Experiment settings
Experiment name | ID | Model version | Horizontal resolution | Atmospheric forcing | Snow depth analysis | Soil moisture analysis | Start date | End date |
---|---|---|---|---|---|---|---|---|
Open-loop by offline LDAS | hsts | 48r1 | TCo319 | ERA5 | No | No | Sep 1998 | Dec 2021 |
Offline LDAS | htyv | 48r1 | TCo319 | ERA5 | Yes | Yes | Sep 1998 | Dec 2021 |
Coupled LDAS (48r1 control) | hut8 | 48r1 | TCo399 | Coupled | Yes | Yes | Dec 2020 | Feb 2021 |
Results (monthly averages of snow depth in Jan 2021)
hut8 (coupled LDAS) | htyv (Offline LDAS) | hsts (open-loop by offline land surface model) |
htyv - hut8 | hsts - hut8 | |
Other related results
You can refer to my poster presentation at the EARSeL workshop in Feb 2023: EARSeL_Ochi.pdf
2 Comments
Patricia de Rosnay
Hi Kenta,
Excellent, we clearly see that the differences between the coupled and offline snow are much reduced when IMS is assimilated. Is it SWE?
Kenta Ochi
Thank you very much for your comment. These plots show snow depth (unit: m). I think it will be similar if we have a look at SWE.