Remotely Sensed Evapotranspiration Data Assimilation for Crop Growth Modeling

The agricultural sector will require more water in the near future to provide more food, fibre and fuels (Molden et al., 2007). As population increases and development calls for increased demand of food, a change in diet due to increased prosperity, and a recent focus on biofuels. This population growth coupled with industrialization and urbanization will result in an increasing demand for water and will have serious consequences on the conservation of water resources. Therefore, a rational approach to best water management practices is needed to balance water supply and demand. One approach to check if the supply is adequate to meet the demand is to account for the respective components in the water balance. Doing so provides an opportunity to search for possible ways to save water from one application and allocate it to another. Simulation models are strong in this regard; they can simulate the processes in the real system and predict the state variables at every stage in the simulation. The role of simulation models in understanding the processes in the soil-plant-atmosphere system has increased significantly in recent years. This is attributed to increased computing capabilities available today (Ines et al., 2002). Such mechanistic ecophysiological models integrate knowledge from data collection by various methods (e.g. GPS, field sampling, satellite remote sensing, radar etc.) and laboratory research. Simulations from such models are widely used to predict and simulate crop growth, yield, water requirements and greenhouse gas emissions. For monitoring agricultural crop production, growth of crops is modeled, for example, by using simulation models. Estimates of crop growth often are inaccurate for practical field conditions. Therefore, model simulations must be improved by incorporating information on the actual growth and development of field crops, for example, by using remote sensing data. Numerous researchers have also used remotely sensed data in conjunction with crop growth models via data assimilation for the purpose of improving soil moisture estimation (Entekhabi et al., 1994; Van Dams et al., 1997; Reichle et al. 2001; Ines et al., 2002; Kamble et al., 2008). The objective of data assimilation is to obtain the best estimate of the state of the system by combining observations with the forecast model’s first guess. Genetic algorithms (GA) are designed to search, discover and emphasize good solutions by applying selection and crossover techniques, inspired by nature, to supply solutions (Holland, 1975; Goldberg, 1989). GA operates on pieces of information like nature does on genes in the course of evolution. Changes in the genes of individuals from a given population allow selection of