OFF TOPIC: voldria comentar sa profunda millora que està preparant GFS, passarà a tenir una ressolució de només 13km (0.25) comparat amb es 0.50 actual. Això vol dir que, en teoria, hauria de ser més competitiu en fenòmens mesoescalars. Pels qui empleeu es programa Zygrib, serà una meravella. Està previst que entri en funcionament es 17 de decembre segons he pogut llegir. De fet, aquest novembre ha aparegut es "Paralel" de GFS per consultar, encara que no sé exactament si funciona com a versió de prova o què...
Vos faig un copy-past de "WeatherZone" sobre ses millores previstes:
1ST UPGRADE:
* Improved horizontal resolution of 13km (0.25 degrees) out to 15 days (compared to 0.5 degrees for current GFS) which is closer to current EC.
* The forecast component of GFS will eventually be renamed GSM (global spectral model) while the overall system it sits in will be NEMS (NOAA environmental modeling system).
* Major physics package upgrade.
* New variables include CAPE and CIN calculated by raising a parcel from the lowest 255hpa layer, Haines Index, sunshine duration, 0 to 6km mean storm motion, wind chill and heat index.
* Faster/less memory version of current post-processor.
* Use high resolution daily SST and sea ice analysis instead of weekly.
* All changes for this year's upgrade have now been finalised.
2ND UPGRADE:
* 4D variational assimilation which in EC's current case, obs data continually accumulates in its database over 2 x 12hr time blocks each day rather than just snapshots at say 00z and 12z.
* Vertical resolution increase from 64 to 90 levels (current EC has 137 levels up to 0.01 hpa due to stratospheric interactions/radiative influences being important).
* Further physics upgrade in which convective gravity wave drag and other non-orographic gravity wave drag modelling is improved.
Si qualcú hi està més enterat i vol afegir coses, s'agrairà