Open HelinWei-NOAA opened 2 years ago
The test with this proposed change is running. I have rerun p8c 5-day forecasts (P8F) with the old way to calculate 2m T/Q. The difference can be see here
It looks that t2m is interpolated from tv and sfctmp? Is it done over vegetated part and bare part separately?
p8c uses the latest update code of NoahMP. So the 2-meter temperature and humidity over land are derived from NoahMP two-tile values instead of using sfc_diag. The diurnal variability becomes too large (too warm daytime and too cold night time) from this change. Now we would like to test using sfc_diag approach to calculate 2-meter temperature and humidity inside NoahMP.