Estimation of Potassium Nutrient Content in Tea Plants Using Sentinel-2 Satellite Imagery

Abstract

Abstract Tea plant (Camellia sinensis) is a plantation crop commodity that has a significant role in Indonesian tea industry. Currently, the determination of a single dose of potassium fertilizer on a tea plant begins with soil and leaf laboratory analysis. The cost of laboratory analysis for testing nutritional content of potassium in a tea plant has always been a challenge for many farmers. the necessity of minimizing the estimate of nutritional content of potassium in tea plants affordably and accurately is the goal of this study. This study aims to estimate the nutrient content of potassium in tea plants using sentinel-2 satellite imagery. The study began by sampling tea leaves and taking satellite images. The data obtained were then correlated with multiple linear regression analysis to create a model. The model obtained is K%= 0.619 + 0.001876 b3 - 0.001264 b4 -0.000201 b8, using the grouped data for the maximum time distance of sampling with image acquisition for 5 days and processed using backward regression method. The coefficient of determination (R-sq) obtained is classified as moderate at 50.18%. the model was validated and well characterized in making estimates with a MAPE percentage of 15.18% and a correctness of 84.82%.