Integration of a Crop Simulation Model and Remote Sensing Information

The monitoring of irrigation requirements at district or regional scale can be based on the use of ecological process-based models and remote sensing data. The former simulates the time evolution (usually at daily scale) of the main biophysical variables which determine crop photosynthesis and water consumption rates; the latter allows to provide the spatial distribution of these variables over a region of interest at a time interval ranging from few days to one month. The evaluation of water balance components and, in particular, the estimate of actual evapotranspiration and the partitioning between soil evaporation and plant transpiration, are crucial issues in semi-arid regions where the scarcity of water resources is becoming an important limiting factor crop growth and yield. The research focused an integrated approach to combine field data, simulation crop model and remote sensing information. Historic data about topography, soil, climate and land cover were collected and organised into a Geographic Information System, which is routinely updated with remotely sensed images. A mechanistic crop growth model has been coupled with the dynamic soil water balance and root uptake model based on finite difference solution of Richards’ equation in order to describe accurately the main crop physiological processes, the soil water fluxes at the interfaces groundwater-soil and soil-atmosphere and water re-distribution in the soil profile. The integration of these two models allows for an improvement of results accuracy for the crop types considered in this study in Mediterranean environmental conditions. Remote sensing images from optical and radar satellite sensors at different spatial scales (from 10 to 50 m) have been collected during crop cycles. Information about land use and leaf area index will be assimilated dynamically by the model, to increase the effectiveness of simulation. The final product is a Decision Support Systems with the purpose to integrate remote sensing images, to estimate crop and soil variables related to drought, to assimilate these variables into a simulation model at district scale and, finally, to estimate evapotranspiration, plant water status and drought indicators. The structures of the model and Decision Support System are described and the first results of the field and simulation activity are reported.