Potential Evapotranspiration (PET) Prediction using Machine Learning Algorithms

Abstract

The principle of Petro-Evaporative Rainfall.  With the use of AI, a system may be built to predict the PIC and establish the range.  Climate, humidity, wind, and sunlight are the causes of transpiration and evaporation.  Using PET is crucial in hydrology, agronomy, and managing water resources.  It is useful for calculating how much water a crop needs throughout its growth season and for assessing the water balance of an area.  Using Machine Learning techniques for customized PET predictions is the goal of this work. In order to create their PET predictions, the meteorologists used the Koppala region's weather records.  Predictions were made using the Random Forest Regressor (RFR), Linear regression (LR), K Nearest Neighbors (KNN), and Sequential vector regression techniques.  The SVR model is unique among all of them; it approaches the PET prediction requirements with an R2 of 0.98 and an RMSE of 0.1.  Among the projected approaches, the model performs the best and operates as expected.