Forecasting soil moisture using in-field real-time data

We are introducing an algorithm that forecasts soil moisture behavior. Our approach involves formulating a soil moisture model based on water balance and a two-stage optimization-based estimation strategy. We highlight two key periods: the green window and the blue window. The green window refers to the estimation period in which the algorithm uses the input data to estimate the model's parameters. The red lines in this window represent actual measurements taken from the soil and weather conditions. On the other hand, the blue window depicts the prediction period where the model is solved using the forecasted rain and temperature three days ahead to predict the soil moisture. The blue lines within this window are our model's predictions.

Disclaimer: This video serves as a supplementary resource for our upcoming paper.