2020–2022年三江平原主要农作物种植分布遥感监测数据集

Abstract

Sanjiang Plain, renowned for its expansive and fertile black soil, serves as a crucial hub for commodity grain production in China. The per capita cultivated land area and per capita grain output in the region are five times the national average. Accurate crop acreage information is of great significance for understanding regional food security and agricultural development planning in Sanjiang Plain. In this paper, we used the Sentinel-2 satellite remote sensing data of time series and the survey data of typical ground features in Sanjiang Plain to screen the feature bands of major crops and typical ground features of surrounding areas. Using the random forest classification algorithm, we extracted the data to produce a dataset of remote sensing monitoring of planting distribution for major crops (rice, corn and soybean) in Sanjiang Plain from 2020 to 2022. Based on field survey data, it has been verified that the overall accuracy of the three crops extraction in 2020, 2021 and 2022 stands at 95.18%, 95.0% and 94.5%, and the Kappa coefficients of these three years are 0.924, 0.925 and 0.919. This dataset can not only offer basic data for the analysis of temporal and spatial changes of crop planting distribution in Sanjiang Plain, but also contribute to informed decision-making in agricultural production management for the region. Furthermore, it can support the development of agricultural informatization as well as the preservation and utilization of black soil resources.