Application of the artificial intelligence approach and remotely sensed imagery for soil moisture evaluation

Abstract The current research attempts to present a modeling framework for determining soil moisture conditions by using remotely sensed imagery products. In this way, identifying various pixels with similar patterns from satellite images could be reliable method have an appropriate view over the condition of particular region. context, artificial intelligence-based self-organizing map (SOM) is employed classify homogenous Phoenix, which located in south Arizona, utilizing parameters extracted images. central clusters are selected as cluster indicator, one each cluster. Then, feed-forward neural networks (FFNNs) consisting three layers input, hidden, and output trained employing time series clusters. Finally, representative simulated models. results reveal suitability SOM-based clustering identify specific points can represent related regions. proposed methodology obtained further used provide cost-effective determine region reducing costs monitoring. HIGHLIGHTS An SOM pixels. pixel ANN. condition.