Assessment of water status in almond trees using optical indices at high and medium spatial resolution

The almond tree is one of Spain’s most widespread woody crops, with an annual growth rate of 4% of cultivated area. In recent years, management practices focused on the intensification of plantations to increase almond production have led to higher irrigation requirements in regions with recurrent water scarcity. In this context, an accurate assessment of canopy water stress is key to successfully apply deficit irrigation strategies. They are critical to optimize water resources without causing severe yield reductions. The usefulness of the Crop Water Stress Index (CWSI) for monitoring transpiration and water status in almond trees has been successfully demonstrated, which uses thermal information acquired remotely at very high spatial resolution to target individual tree crowns. However, canopy temperature in open vegetation orchards is currently limited to sensors installed in manned or unmanned aerial vehicles, which could significantly increase production costs in commercial fields.

This work aims to evaluate the ability of a set of optical indices applied to airborne hyperspectral imagery to assess the water status of an almond tree orchard located in Southern Spain during the 2018 campaign. The field was subjected to different deficit irrigation treatments: fully irrigated, moderately stressed and severely stressed. The analysis has been carried out at different spatial scales to explore the effects of pixel size in detecting water stress situations in an attempt to extrapolate the methodology to Sentinel-2 satellite imagery at medium resolution.

The indices were compared with stem water potential measurements collected in randomly selected trees within areas with deficit irrigation treatments. The results support the potential of the shortwave infrared-based indices, Normalized Difference Water Index (NDWI) and Moisture Stress Index (MSI) to monitor the water stress of this complex crop with open canopy structure when thermal data are not available at sufficient spatial resolution.