Changes between Version 31 and Version 32 of DevelopmentActivities/AccelerationSpinup

Timestamp:

2012-04-18T16:25:06+02:00 (12 years ago)

Author:

dsolyga

Comment:

--

DevelopmentActivities/AccelerationSpinup

@@ -482 +482 @@
 In the first test, we use the complete algorithm of the spinup with strict criterions over biomass(0,1%) and passive pool(0,01%).[[BR]]
 For the second test, we set a huge error on the biomass so the method can start immediately. I obtained in this case 3 snapshots at 100Y, 200Y and 300Y (see figures below).[[BR]]
-I plot for each time the four following variables : total_carbon_soil, nbp, nbp/gpp et nbp/total_carbon_stock_prev (which corresponds to the variation of carbon between two resolutions).  
-
-
+I plot for each time the four following variables : total_carbon_soil, nbp, nbp/gpp et nbp/total_carbon_stock_prev (which corresponds to the variation of carbon between two resolutions).  [[BR]]
+'''Results after 100Y :'''[[BR]]
+[[Image(Spinup_result_100Y.png)]] [[BR]]
+
+'''Results after 200Y :'''[[BR]]
+[[Image(Spinup_result_200Y.png)]] [[BR]]
+
+'''Final results : '''[[BR]]
+[[Image(Spinup_result_nobiomass_test.png)]] [[BR]]
+
+=== MEETING 13/04/2012 ===
+
+Presents : Ncolas Vuichard, Nicolas Viovy, Philippe Peylin, Fabienne Maignan, Patricia Cadule
+see [attachment:Spinup_13042012.pdf].[[BR]]
+
+'''Actions : '''
+ * Simplification of the spinup algorithm : deletion of the biomass tests; start the resolution from scratch; reset the method at every period; still solve the system even if the point is at equiibrium
+ * Diagnostic : keep relative error on passive pool and nbp value diagnostic; the relative error is calculated over the sum of the passive pools over the pixel and not by pft anymore.
+ * TO DO :Test in parallel version + coupling with teststomate + improve scripts.