Changes between Version 31 and Version 32 of DevelopmentActivities/CMIP6/DevelopmentsCMIP6/soil_physic

Timestamp:

2015-02-27T15:19:04+01:00 (9 years ago)

Author:

peylin

Comment:

--

DevelopmentActivities/CMIP6/DevelopmentsCMIP6/soil_physic

@@ -299 +299 @@
 
 Non, je n'ai pas désactivé le gel pour les petites couches. En revanche, j'ai créé une variable qui stocke en mémoire l'énergie libérée au cours du gel, sur l'ensemble du profil. Au moment du dégel, une fois que le sol a complètement dégelé, on compare ce stock d'énergie à l'énergie consommée pour dégeler le sol. Si cette énergie est supérieure à la première, on considère qu'on a trop "fondu" et qu'on aurait dû réchauffer à la place, donc on réajuste la température, sur l'ensemble du profil.
+
 Ta prochaine question sera sans doute pourquoi j'ai travaillé sur l'ensemble du profil et pas par couches. La réponse est que j'avais la vieille version de la neige, implicite dans le sol. Donc, les "couches de sol" changeaient dans le temps en fonction de la hauteur de neige, donc il était difficile de garder une trace de l'énergie du gel/dégel du sol par couche.
 
@@ -320 +321 @@
  - new common scheme with 2m for W and 8m for T: V3
 
-Some results for T:[[BR]]
-
-5 meters is not enough for diunal/seasonal cycle of T (criterial is that the amplitude of the cycle at lowest layer should lower by a factor "E-3" of that of the first layer ???)
-
-we need a depth of at least 9.5 m for sand soil when using new USDA texture classes.
-
-
-For soil moisture:[[BR]]
-
-Overall the differences between V2 and V3 are not too large for the fluxes ?
-
-V2 need a longer spin-up (at least 40 yr compared to 20 yr for V3). Soil moisture at t0 = 0.2 m3/m3
- 
-Impact of the drainage is large for JJA, but additional graphics are needed to better quantify the changes
-
-The impact on the river dischages is only significant for some rivers: additional test ?
+Some results for T:
+ - 5 meters is not enough for diunal/seasonal cycle of T (criterial is that the amplitude of the cycle at lowest layer should lower by a factor "E-3" of that of the first layer ???)
+ - we need a depth of at least 9.5 m for sand soil when using new USDA texture classes.
+
+For soil moisture:
+ - Overall the differences between V2 and V3 are not too large for the fluxes ?
+ - V2 need a longer spin-up (at least 40 yr compared to 20 yr for V3). Soil moisture at t0 = 0.2 m3/m3
+ - Impact of the drainage is large for JJA, but additional graphics are needed to better quantify the changes
+ - The impact on the river dischages is only significant for some rivers: additional test ?
 
 
@@ -349 +343 @@
 
 Soil Freezing:
-
-
+ - change significantly soil heat capacity;
+ - we cannot anticipate the energy needed to freeze water as we dont know at the begining of the time step if the temperature will cross the zero before calculating it (thus we can not anticipate the release of energy). 
+ - the scheme of Isabelle Goutevin has been implemented in the TRUNC.
+ - As for energy conservation the scheme buffers the energy released during freezing for all layers, then calculates the consumed energy during the next thawing period. If the the two energy do not match she distribute the excess/deficit of energy by re-adjusting the Temperatures of the whole soil profile. There is thus only energy conservation throughout the season (see email above by Isabelle). She could not do it per layers as with the old snow scheme the depth of the layers varies in time.
+ - With the new snow scheme we can envisage a new scheme that could conserve the energy each day or time step and possibly for each layer. To be discussed.