In this paper we study the feasibility of using services offered by a Spatial Data Infrastructure as the basis for agricultural
productions. By developing a prototype we demonstrate that a Spatial Data Infrastructure facilitates rapid development
of applications that solve flood forecasting and controlling problems. The prototype provides clients with a distributed
application that enables the assessment of flooding damage areas based on precipitation data in China mainland. We
present the architecture consists of four models: Time-Space, State-Space, prediction and estimation for flood forecasting
and control. We conclude that the model implementation allows us to predict in time and in space, and to account for
missing data. The model was applied to 56 years of monthly precipitation data from the China land, and seems to capture
the dynamic evolution of the spatial processes associated with the precipitation in this region.