Skip to content

Instantly share code, notes, and snippets.

@denis-bz
Last active December 20, 2022 17:35
Show Gist options
  • Save denis-bz/a8703cdd1207c5a291ddce43c9c42cc4 to your computer and use it in GitHub Desktop.
Save denis-bz/a8703cdd1207c5a291ddce43c9c42cc4 to your computer and use it in GitHub Desktop.
Wind speeds at 100m in Bayern, ERA5 vs. globalwindatlas 20 Dec 2022

Wind speeds at 100m in Bayern, ERA5 vs. globalwindatlas

Keywords: wind power, Germany, python, ERA5, globalwindatlas

Purpose: look at windspeed estimates at 100m in Bayern from two different sources:
ERA5: https://confluence.ecmwf.int/display/CKB/ERA5
GWA: https://globalwindatlas.info/area/Germany/Bayern

20dec-era5-vs-gwa-100m

even though these are not directly comparable:
ERA5 has average windspeeds on 0.25° x 0.25° regions (~ 30 x 20 km), whereas
GWA has data on a fine grid, 0.0025° x 0.0025° (here plotted coarser).

How does ERA5 take averages, how does GWA interpolate / extrapolate ? As an old engineer, I like to have a rough idea of where data comes from -- what's measured, what's extrapolated -- and I like spot checks. For GWA, I see neither. Can you help ?

Why are globalwindatlas windspeeds so much higher than ERA5 ? Their capacity factors are correspondingly high:
31 % DEU_capacity-factor_IEC3.tif (at 100m ?) in Bayern, vs.
22 % from gist Sun-wind-biomass-electricity-Germany.

See also

GWA caps vs real caps at the sites of several windfarms in the Netherlands:

gwacap realcap  MW  GWh              Name  ix  iy    lat    lon knmi50 knmi100 knmi150 gwa50 gwa100 gwa150
    59      37 429 1400   Noordoostpolder 115 108 52.702 5.6148   6.89    7.71    8.26  7.89   8.95   9.91
    63      45 383 1500           Fryslân 105 121  53.01   5.28   8.37    8.98     9.4  8.65   9.48  10.04
    ...

in gist windspeeds-Netherlands-wikipedia-KNMI-GWA

gist Windspeeds and windfarms in Bayern

Links

https://confluence.ecmwf.int/display/CKB/ERA5:+data+documentation
https://confluence.ecmwf.int/display/CKB/Model+grid+box+and+time+step
some sort of average value over the grid scale. However, models don't represent the grid scale very well either

Comments welcome

cheers
— denis-bz-py t-online.de 20 Dec 2022

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment