Integrating a very fast simulated annealing optimization algorithm for crop leaf area index variational assimilation

Leaf area index (LAI) is a major indicator for crop growth monitoring and yield estimation. Data assimilation as an effective tool for crop LAI estimation fully considers the properties of actual observations and physical model simulations. In this work, we present a new data assimilation scheme, introducing a very fast simulated annealing (VFSA) optimization algorithm into the process of crop LAI assimilation with a four dimensional variational data assimilation (4DVAR) algorithm. Firstly, calibrating the input parameters of a crop growth simulation model based on history observations. Secondly, quantitatively describing the relationship between fields observed data and model simulated data by the cost function of 4DVAR algorithm. Finally, the optimization process of the cost function is accomplished by the VFSA optimization algorithm, and further the optimal solution is taken as the best combination of input parameters of physical model for LAI estimation. Winter wheat in Beijing is taken as an experimental object. The numerical results show not only the improved time efficiency of this proposed assimilation scheme, but also enhanced assimilation accuracy of all LAI assimilations, especially for LAI  ≥≥  3.00. Theoretical analysis and practical experiments confirm the application prospect of VFSA optimization algorithm in LAI variational assimilation.