Saving water in irrigated agriculture (where 75% of water is used worldwide) can be achieved through the use of Earth Observation (EO)-derived information in operational irrigation scheduling at farm and field scale. End-users of the information are the farmers, who experience benefits in the form of "more crops per drop" (enhanced water productivity) and "more jobs per drop" (boost of rural development). Space-assisted Irrigation Advisory Services at community level provide the EO-derived irrigation scheduling information to them, interacting with water management decision makers at river basin level, and serving as a potential policy instrument at national and European scale.

The operational procedure is based on a virtual constellation of all available high-resolution EO satellites, to achieve required space-time resolution (10-30 m, weekly). The virtual constellation is tied together by means of a novel inter-satellite cross-calibration procedure. Due to its excellent operational availability and low cost, Landsat is the backbone of this virtual constellation. The other high-resolution satellites (IRS, Spot, Terra-Aster, ALI) are used mainly to fill gaps in the time series.

The operational space-assisted irrigation scheduling uses a time series of weekly images during the entire growing season of crops in a given area (4-8 months). Maps of crop evapotranspiration and crop water requirements are derived from each image (using mature EO methodology) and introduced directly in existing routine irrigation advisory operations, which are based on standards recommended by the Food and Agriculture Organization (FAO). The EO-derived information replaces and complements data previously obtained by resource-intensive field work or averaged tables. Introduced into a GIS, it opens the access to the spatial dimension and thus, a more accurate prediction of irrigation water needs.

Combined with leading-edge IT tools, it offers a personalized information transmission service to the farmers, where each farmer can receive easy-to-use information on the status and irrigation scheduling of each of their fields and additional information on weather, environmental issues, and best farming practises.

The concept is currently implemented as the DEMETER (DEMonstration of Earth observation TEchnology in Routine irrigation advisory services) prototype in three pilot zones with national up-scaling perspective in Spain, Italy, and Portugal. Pilot campaigns were conducted in each zone during the entire 2003 growing season, providing a large sample of ground truth validation data. The campaigns also provide the first demonstration cases of fully operational EO-derived irrigation scheduling products, delivered to the users two days after the Landsat satellite overpass. This paper describes the methodology, the operational processing chain, and selected campaign results, including a demonstration sequence of weekly images and derived irrigation recommendation.