Introduction
The goal of the audit undertaken was the comparisson of the differences in the implementation of the FAO (Food and Agriculture Organisation of the United Nations) water balance method in two software packages : AMS (AgrometShell) developed by HSS for FAO and GWSI (Gobal Water Statisfaction Index) developed by Alterra for JRC (Joint Research Centre of the European Union). Although the primary focus of this report is the comparison of the water balance implementation, the general architecture of the models will be described as well as the complementarities and the possibilities for data exchange.
Conclusions
Following the analysis the conclusion is drawn that the packages differ only marginally and complement each other well. More specifically:
- AMS could be used to analyze GWSI output. In general AMS gives the user (e.g. analysts at JRC) the possibility to analyze GWSI output statistically as well as in space and time. Subsequently, JRC analysts would have a playground for thematic experts.
- As the input weather data for both models differ (measured values versus ECMWF model estimates) AMS model output can be used to verify the GWSI model outputs in a number of "verification areas" that could be established. These verification areas could consist of continents, sub-continents or individual countries and would have to involve National or Regional Early Warning Units.
- AMS could provide "planting date estimation" functionality to GWSI.
- AMS contains a number of interpolation techniques (SEDI) that might improve GWSI products through the application of a "background" factor. This could be applied to inputs (e.g. rainfall estimates are improved through the application of altitude as a background factor) or outputs (e.g. using water balance indicators with a NDVI as a background factor).
