A new automated approach for remote sensing recognition and yield estimation of cultivated alfalfa crop based on Sentinel-2 NDVI time-series data: A case study of Hexi Corridor, China

Abstract

Alfalfa (Medicago sativa) is an important forage source for grassland agricultural development, so it would be worthwhile to explore accurate and fast methods of alfalfa remote sensing identification and yield estimation. However, the traditional methods of identifying large areas of crops and yield estimation have some problems, such as the limited spatial resolution of remote sensing data and heavy reliance training data. In this study, based on Sentinel-2 high-resolution images and the Google Earth Engine (GEE) platform to establish a cloud-free normalized difference vegetation index (NDVI) time-series dataset, we proposed an effective method for alfalfa feature extraction and yield estimation method. The results show that (1) The producer’s accuracy, user’s accuracy, overall accuracy, and Kappa coefficient of alfalfa identification using a trough recognition algorithm were 98.51%, 91.67%, 94.26%, and 0.88, respectively. The total area of cultivated alfalfa identified in the study area in 2020 was estimated at 46,793.21 hm2, which was mainly distributed in areas in north of the Qilian Mountains; (2) NDVI had a highly significant correlation with alfalfa hay yield, and the power function regression model was the greatest, with an R2 greater than 0.65; (3) The annual unit hay yield of four alfalfa cuttings was estimated at 17,497.55-32,962.10 kg/hm2, with a total hay yield of 48.38×107 kg and an average hay yield of 4,464.95 kg/hm2. The method proposed has important application potential for automatic and rapid remote sensing identification and yield evaluation of large-scale cultivated alfalfa.