= GSWP Forcing =

== Tair ==

[[Image(Graphics_GSWP3e-1.png, width=40%, center)]]

== Qair ==

[[Image(Graphics_GSWP3e-3.png, width=40%, center)]]

== LWdown ==

[[Image(Graphics_GSWP3e-0.png, width=40%, center)]]

== SWdown ==

[[Image(Graphics_GSWP3e-2.png, width=40%, center)]]

== Potential Evaporation ==

[[Image(Graphics_GSWP3e-4.png, width=40%, center)]]

== Difference in Potential Evaporation for June==

[[Image(CheckMaps-2.png, width=40%, center)]]

== Wind ==

The wind is given in the forcing at 10m and the model interpolates it to the same level as Tair and Qair. Thus the values in the output of the model are lower than in the forcing. 

[[Image(Graphics_GSWP3e-Wind.png, width=40%, center)]]

== PSurf ==

[[Image(Graphics_GSWP3e-PSurf.png, width=40%, center)]]

== RadT ==

[[Image(Graphics_GSWP3e-RadT.png, width=40%, center)]]

== Testing other hypothesis for the meaning of the average of the fluxes ==

In the case of GSWP3 we tested the following 3 options : mean(start), mean(center) and mean(end).
It was confirmed by Hyungjun Kim that the correct meaning of the fluxes is the one of ECMWF and thus the time stamp positioned at the end of the averaging interval.

Impact on potential evaporation :

[[Image(CompareRuns_driver-3.png, width=40%, center)]]