Can water stress be accurately modeled using Remote sensing data?

_I created a irrigation recommendation model that can inform farmers when to irrigate and for how long based on water stress

With increasing temperatures and extreme rainfall fluctuations, climate change is causing harsher conditions for growing crops. Farming is becoming increasingly difficult as farmers struggle to produce enough food to feed a growing population. As a result, agriculture stands to gain a great deal from data driven models to help inform crop production.
Irrigation is a critical input for agricultural crop production globally. 70% of the world's fresh water is used for irrigation. However, irrigation increases farmers' production costs. At the same time, increased drought events have led to water stress. Water stress from drought will have major impacts on biomass, yield, and quality. However, drought stress can be alleviated with proper irrigation. Thus, having models to quantify water stress in plants can inform farmers on irrigation needs for gains at harvest and reduce inputs to fields. This can save thousands of gallons of water, reduce the production cost for farmers, and protect ecosystems.

I tried to solve the problem of whether it is possible to accurately model crop water stress with remote sensing technology deployed throughout a growing season at the level of a crop's growth. Climate and plant-derived parameters sampled from remote-sensed data sources can be utilized to develop a model to predict plant water stress. We will aim to create a model using remote sensed, in field measurements to determine whether a corn crop is undergoing water stress at a point in time in a growing season for a corn crop. Our goal was to be able to create a model which can take in field specific remote sensed values and predict with a strong degree of accuracy whether the crop in the field is water stressed using the methodology laid out in the Dejonge paper on CWSI. Ultimately, I built out a model to see how close we could get to being able to create the conditions where someone could view the data with strong confidence to make an irrigation decision. The implications of being able to do this are to provide growers with the tools to irrigate more efficiently, to save natural resources, to reduce equipment and labor costs, and to obtain the best outcomes that they can given their climatic conditions.