Smart irrigation forecast using satellite LANDSAT data and meteo-hydrological modeling

The paper discusses advances in coupling satellite driven soil water balance model and meteorological forecast as support for precision smart irrigation use in a case study of an operative farm in the South of Italy where semi-arid climatic conditions holds. Crop water needs forecast are computed with the intuitive idea of forcing the soil water balance model with the meteorological model outlooks. Discussion on the methodology approach is presented, comparing, for a reanalysis period between June and September 2014, the forecast system outputs with observed soil moisture and crop water needs. Two main issues are here in emphasized: the characteristic of soil moisture water balance model, that due to its state variables may be directly calibrated and validated using satellite or near sensing land surface temperatures; the accuracy of those forecast meteorological variables that are the most important in driving the soil water and energy balance. The soil water balance model performances are then discussed highlighting the importance of using a model which state variable (the pixel surface equilibrium temperature) is the same as the data detected by satellite (Land Surface Temperature), so that it can be used for calibrating and validating soil hydrological parameters. Model outputs are also validated with a comparison of ground latent and sensible heat fluxes from an eddy covariance station and soil moisture data. Problems insight into the meteorological modeling, such as temporal and spatial scale, and their influence on soil moisture forecast are discussed showing on the base of several observation periods the need to increase the meteorological forcings accuracy for this type of applications. The obtained results show how the proposed methodology of the forecasting system is able to have a high reliability in soil moisture forecast correctly providing irrigation suggestion.