SMArtCast: Predicting soil moisture interpolations into the future using Earth observation data in a deep learning framework
This work addresses soil moisture prediction for agricultural monitoring, particularly in areas with limited data, but it is incremental as it applies existing deep learning methods to a specific domain.
The authors tackled the problem of predicting future soil moisture maps for crop health monitoring by using LSTM architectures to analyze satellite-derived soil moisture and vegetation indices, then interpolating these sparse measurements into dense spatial maps, achieving potential advance warnings for inhospitable conditions.
Soil moisture is critical component of crop health and monitoring it can enable further actions for increasing yield or preventing catastrophic die off. As climate change increases the likelihood of extreme weather events and reduces the predictability of weather, and non-optimal soil moistures for crops may become more likely. In this work, we a series of LSTM architectures to analyze measurements of soil moisture and vegetation indiced derived from satellite imagery. The system learns to predict the future values of these measurements. These spatially sparse values and indices are used as input features to an interpolation method that infer spatially dense moisture map for a future time point. This has the potential to provide advance warning for soil moistures that may be inhospitable to crops across an area with limited monitoring capacity.