AMMA icon
  Home
  Executive Summary
  Experiment
  Calendar
  Participants
  Input Data
  Output Data
  Results
  Meetings
  Presentations
  Publications
  Contact
  AMMA Related:
  ALMIP2
  AMMA CNRM-GAME
  AMMA International
  AMMA US
ALMIP title banner

Results

(last updated: Jan. 9, 2009)


Page Contents:

  1. Comparison with field measurements
  2. Model Intercomparison
  3. Multi-Model Average Fields


Comparison with field measurements

Several ALMIP model evaluation efforts are underway. As an example, a comparison of the daily average ALMIP simulated volumetric water content (from Exp.3: i.e. using the atmospheric state variables from ECMWF, the downwelling radiative fluxes from LAND-SAF and the precipitation from the TRMM 3B42-V6 dataset) integrated over the uppermost one meter of the soil is shown in Fig.1 (for a single pixel using an ensemble of 10 model simulations) compared with the corresponding observation at the Wankama, Niger site for 2006. Both the model simulations and the observations are expressed as anomalies (the means have been removed) as a simple way to normalize the water content values in order to better relate them to surface (evaporation) and hydrological (runoff) mass fluxes. From this figure it can bee seen that the models tend to continue to dry slightly before monsoon onset in contrast to the observations, the wetting stage during the monsoon is captured very well by the models, and finally, the dry down is much slower than is seen in the observations. The model errors compared to the observations are mainly related to differences in the local and large scale forcing data (such as precipitation), the soil and vegetation parameters, and the model evaporation and hydrological parameterizations. The soil moisture data from the Wankama, Niger site have been provided by T. Pellarin (CNRS) of LTHE, Grenoble, France.


Fig.1 A comparison of the ALMIP-simulated (green) verses observed (red) soil moisture anomalies in the upper 1 meter of the soil during 2006 at Wankama, Niger.




Sample JJAS (June-September) Average Multi-Model Plots

Some sample output fields and some basic analysis are presented herein. For a complete listing of model output fields (and a brief description), please see the output data page. If you wish to aquire model output fields, please see the contact page. Please note too that soon such fields will be available via the AMMA-DB.

Images for Experiments 1, 2 and 3 can be can be accessed here (see current listing below). The file name convention is
VAR_JJAS-SVAT_ExpX_YYYY_EXTSN.gif
where X represents the experiment number, and YYYY corresponds to the simulated year. VAR represents the variable name using the ALMA convention (see the output data page for a listing). Note that 2 variables serparated by an underscore here indicates a ratio of the first variable to the second, for example TVeg_Evap represents the ratio of the transpiration to the total evapotranspiration. EXTSN is an added extension, whhich is generally used to indicate the time or the temporal averaging period (e.g. EXTSN=JJAS-AVG for fields averaged from June through September).


Access to images for Exp3, years 2002-2007



Multi-Model Average Fields


Monthly plots (Exp3) for selected variables, 2002-2007

Some sample output fields and some basic analysis are presented herein. For a complete listing of model output fields (and a brief description), please see the output data page. If you wish to aquire model output fields, please see the contact page. The plotted data are available in NetCDF format.

Monthly Qh for 2002
Monthly Qh for 2003
Monthly Qh for 2004
Monthly Qh for 2005
Monthly Qh for 2006
Monthly Qh for 2007

Monthly Qle for 2002
Monthly Qle for 2003
Monthly Qle for 2004
Monthly Qle for 2005
Monthly Qle for 2006
Monthly Qle for 2007

Monthly Rnet for 2002
Monthly Rnet for 2003
Monthly Rnet for 2004
Monthly Rnet for 2005
Monthly Rnet for 2006
Monthly Rnet for 2007

NOTE, the ALMA name conventions (see the output data page for a listing) are used above: Qh=Sensible heat flux (W/m2), Qle=latent heat flux (W/m2), and Rnet=Net radiation (W/m2).