Detección de estrés hídrico y niveles de NPK en algodón (Gossypium barbadense L.) mediante imágenes térmicas

Introduction. Real-time monitoring of water stress levels in crops is crucial to improve water use efficiency in agriculture. Objective. To determine the feasibility of using thermal images collected by an unmanned aerial vehicle(UAV) to detect water stress in cotton (Gossypium barbadense L.), Molinero extra-long variety. Materials andmethods. The study was conducted at the Irrigation Research Unit of the National Agrarian University La Molina, Lima, Peru, from November 2022 to May 2023. Forty-eight sampling units of 0.80 × 1.5 m were used with dripirrigation activated in response to physical signs of water stress in cotton. The crop water stress index (CWSI) wasmonitored through eight UAV flights equipped with a thermal camera, and the data were corrected using radiometer measurements. Stomatal conductance (gs) was measured using a porometer, and soil moisture (θ) with a sensor. In addition, tests were carried out in pots to cause plant death and induce maximum water stress. Results. A 0.96correlation was obtained for the thermal images. A potential 0.83 correlation between gs and CWSI was observed, along with a 0.88 between θ and CWSI, indicating an adequate representation of water stress variability with irrigation levels. Conclusions. The use of thermal images obtained via UAV allowed estimating a maximum CWSI of 0.39, optimizing irrigation timing and achieving water savings of 50 %. The CWSI proved to be a valuable tool for accurate irrigation scheduling, with potential to improve water use efficiency up to 2.27 kg/m3 in arid regions such as La Molina, Peru.