New mathematical models to estimate wheat Leaf Chlorophyll Content based on Artificial Neural Network and remote sensing data

Supervised Artificial Neural Network such as Feed-forward Multi-layer Perceptron algorithm and advanced Remote sensing technologies play an important role in improving mapping process of various biophysical and biochemical vegetation parameters. Advanced knowledge of the crop status will help farmers, decision makers and policy makers to take precautions to reduce losses and to increase production. Leaf chlorophyll content of the crop can help to obtain information about the physiological status of the plant and it can help in studying of the carbon balance and responses to fertilizer (e.g., nitrogen) application. In this research, Leaf Chlorophyll Content (LCC) for winter wheat in a large area in Lebanon is estimated based on the technologies and algorithms listed earlier. The estimation process is based on the collected field measurements of reflectance data and chlorophyll content of the wheat by a very high resolution spectroradiometer and precise chlorophyll meter. Mathematical models for wheat crop chlorophyll content are created based on the collected data and a supervised Artificial Neural Network algorithm. The created mathematical models are used to investigate the status of the winter wheat crop throughout Lebanon using remote sensing data. The verification of the results is based on the collected LCC measurements using a chlorophyll meter during the same time a remote sensing satellite is imaging the study area. The verification of the results is based on computing the accuracies between the predicted and estimated LCC. The average accuracy of the new mathematical models is about 99.5%.