ma-ai/CropNet

CropNet数据集是一个开放、大规模、面向深度学习的数据集，专门针对美国本土县级气候变化感知的作物产量预测。它由三种模态数据组成：Sentinel-2卫星图像（提供高分辨率农业图像和归一化植被指数）、WRF-HRRR计算数据集（包含每日和每月气象参数）和USDA作物数据集（提供作物产量和生产信息）。这些数据在时空上对齐，覆盖2017-2022年超过2200个美国县。数据集旨在促进研究人员开发深度学习模型，通过考虑短期生长季节天气变化和长期气候变化的影响，及时准确地预测县级作物产量。The CropNet dataset is an open, large-scale, and deep learning-ready dataset, specifically targeting climate change-aware crop yield predictions for the contiguous United States at the county level. It is composed of three modalities of data: Sentinel-2 Imagery (providing high-resolution agriculture imagery and normalized difference vegetation index), WRF-HRRR Computed Dataset (containing daily and monthly meteorological parameters), and USDA Crop Dataset (offering crop yield and production information). These data are aligned in both spatial and temporal domains, spanning over 2200 U.S. counties from 2017 to 2022. The dataset is designed to facilitate researchers in developing deep learning models for timely and precise crop yield predictions by accounting for the effects of short-term growing season weather variations and long-term climate change.