US-LandCover-Dataset

Monitoring land cover changes is essential for understanding the environmental impact of natural processes and human activities, such as deforestation, urbanization, and agriculture. We present a comprehensive dataset covering the entire United States, spanning from 2016 to 2024. This dataset integrates multi-band imagery from the Sentinel-2 satellite with pixel-level land cover annotations from the Dynamic World dataset, providing a valuable resource for land cover change detection research. The dataset is publicly available and includes six spectral bands (Red, Green, Blue, NIR, SWIR1, SWIR2) as well as labels indicating land cover types. To address the steep learning curve associated with remote sensing, we also provide accompanying code to facilitate data loading and perform basic analyses, enabling researchers to focus on their core work without being hindered by data preprocessing challenges. This dataset facilitates research in fields such as environmental monitoring, urban planning, and climate change, offering an accessible tool for understanding landscape dynamics over time.