中国科学数据（中英文网络版）2025，Vol.10Issue(4)：38-53,16.DOI:10.11922/11-6035.csd.2025.0105.zh
1990-2024年黄土高原果园空间分布数据集
A dataset of the spatial distribution of orchards on the Loess Plateau(1990-2024)
摘要
Abstract
The Loess Plateau is an important orchard production base in China.Accurately mapping the spatial distribution of orchards in this region is of great significanc for agricultural planning,ecological conservation,and regional sustainable development.Based on Landsat TM/OLI imagery during 1990-2024 from the Google Earth Engine(GEE)platform,and orchard samples from high-resolution World Imagery Wayback imagery,22 feature variable sets were integrated,and the random forest classification method was employed to extract the spatial distribution information of orchards across the Loess Plateau.The results indicate that the area of orchards on the Loess Plateau exhibited a pronounced expansion trend during the study period,from 22,145.66 km2 in 1990 to 42,610.82 km2 in 2024,with an average annual growth rate of 1.92％.The primary orchard regions remain concentrated in the Weibei Loess Plateau in Shaanxi Province,the Yuncheng-Linfen Basin in Shanxi Province,and the Longdong Loess Plateau in Gansu Province,with average annual growth rates of 2.16％,1.91％,and 1.33％,respectively.Accuracy assessments based on confusion matrices confirmed robust performance,with overall accuracy and Kappa coefficients exceeding 80％for all years.This dataset provides critical data to support orchard industry planning,ecosystem management,and regional ecological research across the Loess Plateau,thereby facilitating the sustainable development of the orchard industry.关键词
黄土高原/果园/随机森林/多源特征融合/长时序Key words
Loess Plateau/orchard/random forest/multi-source feature fusion/long time course引用本文复制引用
康亚琴,姚晓军,冯克庭..1990-2024年黄土高原果园空间分布数据集[J].中国科学数据（中英文网络版）,2025,10(4):38-53,16.基金项目
国家重点研发计划(2022YFF0711703) National Key Research and Development Program of China(2022YFF0711703) （2022YFF0711703）
