Understand change signal better

jwohland / kliwist_modelchain

Evaluating the GCM-RCM modeling chain for wind energy applications.

MIT License

0 stars 0 forks source link

Understand change signal better #30

Closed jwohland closed 2 years ago

jwohland commented 2 years ago

To test the two main hypothesis (land use change important onshore, thermal stability important offshore), additional plots are needed.

[x] Compute changes in monthly mean wind speeds (e.g., January 1986-2005 minus January 2080-2100) for all months
[x] plot them
[x] Compute changes in SST and surface temperature, as well as SST and surface temperature individually
[x] Plot them
[x] Compute changes in sea ice fraction
[x] Plot them

jwohland commented 2 years ago

[x] Need to check that the tas, ts, and sic ensemble are same size as sfcWind

Some combinations are similar (cordex sfcWind vs. tas; cmip sfcWind vs. tas vs tas-ts vs. ts) and other are very different (no sic in CMIP5, ensemble half as large for cordex sic; only 1-2 models report ts cordex)

Keep in mind during interpretation

grafik

jwohland commented 2 years ago

sic only found in cordex, not in cmip

jwohland commented 2 years ago

Something is wrong in CORDEX sic in december in rcp26 and rcp85

and ts in december also odd in rcp26

Maybe something happening in the very last month? Last december could be imcomplete

jwohland commented 2 years ago

ts only exists for a small set of models

[x] look at ensemble size

jwohland commented 2 years ago

Actually, a very large proportion is the strength of the large-scale circulation