Hi-Freq NEON-VI: High-Frequency Vegetation Indices Dataset for NEON Sites (2022-2025) based on PlanetScope

Abstract This dataset provides high-frequency (Weekly & Monthly), high-resolution (3m), and analysis-ready (ARD) vegetation indices products for 13 representative NEON (National Ecological Observatory Network) sites across the United States. It is designed to address the "data gap" problem caused by cloud cover in public satellite data (e.g., Sentinel-2) and to capture fine-scale phenological changes. Key Features High Frequency: Leveraging Planet SuperDove's daily monitoring capability, we generated Weekly and Monthly composite products using the Max Value Composite (MVC) algorithm. Multi-Indices: Includes NDVI, EVI (Enhanced Vegetation Index), and kNDVI (Kernel NDVI) to support diverse ecological applications. High Quality: Rigorous quality control (QC) was applied, including UDM2-based cloud/shadow masking and radiometric calibration (Top-of-Atmosphere reflectance). Validated: Cross-validated against Sentinel-2 L2A data, showing high consistency (high R-squared) and superior continuity during the growing season compared to public datasets. Data Specifications Time Range: January 2022 – January 2025 Spatial Resolution: 3 meters Format: GeoTIFF (Float32) & Statistical CSV Sites: 13 NEON sites (Covering Forest, Grassland, Shrubland, etc.) Funding & Support This dataset was produced as part of the academic course project at SYSU.