A combined remote sensing and machine learning approach to monitoring crop stress and predicting crop yield

在当前和未来的生长条件下优化作物产量是确保未来粮食安全的关键优先事项。气候变化带来的挑战，包括气温上升、降水量变化增加和干旱，将导致非生物作物压力增加，潜在地限制产量，并危及可靠粮食供应的维持。作物对多种同时胁迫做出生理反应的能力支撑着农业系统的复原力。遥感技术的进步为实时监测作物逆境提供了可能，但从遥感数据中检测特定类型的非生物逆境是具有挑战性的，而且影响了动态干预的能力和对产量的影响。机器学习(ML)在开发新的分析工具以从一系列遥感数据中提取关键生物信息方面显示出巨大的前景。博士项目的总体目标是将遥感测量与深入的植物生理分析相结合，以开发人工智能(AI)/ML算法，以对非生物作物胁迫的不同信号及其对作物产量的影响进行有效预测。

Optimising crop yield under current and future growth conditions is a key priority for ensuring future food security. The challenges of a changing climate, through increasing temperatures, more variable precipitation, and drought will lead to increased abiotic crop stress, potentially limiting yields and compromising the maintenance of a reliable food supply. The ability of crops to physiologically respond to multiple, simultaneous stresses underpins the resilience of agricultural systems. Technological advances in remote sensing (RS) hold the potential for monitoring crop stress in real-time, however detecting the specific type of abiotic stress from RS data is challenging and compromises both the ability for dynamic intervention and also for understanding the impact on yields. Machine learning (ML) shows great promise for developing new analytical tools to extract key biological information from a range of remotely sensed data. The overall aim of the PhD project is to combine RS measurements with in-depth analyses of plant physiology in order to develop artificial intelligence (AI)/ML algorithms to make effective predictions about different signals of abiotic crop stress and the impacts on crop yield.