Remote sensing based monitoring of durum wheat under water stress treatments

Nowadays, the remotely provided information is urgently needed for various decision makers and remote sensing is widely used for different purposes. An evaluation and standardization of the remote sensing techniques is becoming a necessity after the increase of the sensors number and way of data acquisition. The use of remotely sensed information is of crucial importance for agricultural sector as the monitoring of agricultural activities faces special problems related to the spatial and temporal variability, physical landscape, as well as climatic driving variables and agricultural management practices. Requests for objective information will increase in the future, because of the expected modernization of agricultural sector and needs for a more efficient use of resources. This research aims to provide a comparison of two remote sensing techniques (satellite and ground-based) as tools describing the variations of crop stress related indices under different water regimes (case of durum wheat). The experimental layout was established in Policoro (Matera) located in Southern Italy about 3 km far from the Ionian coast. The growing season was from February to June 2015 with three distinguished water management practices (rain-fed, 50% and 100% of irrigation requirements). The Landsat 8 images and ground-based sensing data were acquired regularly in April, May and June together with plant physiological parameters. The overall results indicated no significant differences in terms of both biomass and yield among the irrigation regimes because of the abundant precipitation (355 mm) which limited the irrigation supply. Moderate correlations appear to be evident for satellite and groundbased derived LAI, NDVI and SAVI regressions, with R² values close or higher than 0.5. The best performance was obtained for the WDI with (R²=0.81). Nevertheless, the satellite data could provide reasonable indications about the crop status if applied with higher resolution.