ajdawson / windspharm

A Python library for spherical harmonic computations on vector winds.
http://ajdawson.github.io/windspharm
MIT License
82 stars 36 forks source link

Problem with computing vorticity of NWP model output #114

Closed tctrkrpiluvr closed 3 years ago

tctrkrpiluvr commented 3 years ago

Hello, I find this module very convenient, but have encountered a problem when computing the vorticity of a numerical weather prediction (NWP) model output. While there is no error during the computation, the results turn out to be quite non-physical.

In my opinion, it appears to be due to the longitude and latitude increments being different. The model I used here is the regridded version of GDAPS-UM, the Korean model borrowed from the Met Office, with a horizontal resolution of 0.234375 (769 latitude points) x 0.3515625 (1024 longitude points). Other models were interpolated to 0.125 x 0.125, 0.25 x 0.25, 0.5 x 0.5, or even 0.140625 x 0.140625, but you can see that the resolution is the same in the x and y directions.

I have used the same function in the standard interface:

        w = ws.standard.VectorWind(u, v, rsphere = 6371229.)
        rvo, div = w.vrtdiv()

This is the GDAPS-UM result for vorticity (shaded). 2019080612_GDAPS

These are results of ECMWF-IFS, GFS, and the Korean Integrated Model, or KIM (vorticity shaded in each figure). 2019080612_ECMWF 2019080612_GFS 2019080612_KIM

The u and v seem to be similar, considering these are different forecasts and each have different resolutions. However, only the GDAPS-UM model has a peculiar vorticity pattern.

I can't quite find out what is wrong with the code. Could anybody help me on this matter? Thank you in advance.

P.S. Please mind that the model outputs cannot be publicly released.