Publications

Ma, YM; Li, WJ; Wang, JW; Liu, SY; Dong, MX; Shi, ZC (2025). Evaluating the consistency between Sentinel-2 and Planet constellations at field scale: illustration over winter wheat. PRECISION AGRICULTURE, 26(2), 31.

Abstract
PurposeSentinel-2 enables precise wheat monitoring, but cloud cover can disrupt observations. PlanetScope's near-daily, 3-m resolution can compensate, but cross-calibration of its different sensors poses challenges. Synergy between Sentinel-2 and PlanetScope can leverage their complementary strengths. Prior to this, consistency analysis is crucial. This study aims to evaluate their consistency and determine optimal time intervals and product levels.MethodsThe wheat Leaf Area Index (LAI) was measured in France and China (2019-2022), alongside collecting Sentinel-2 and PlanetScope (SuperDove, Dove-R) imagery. Surface reflectance, vegetation indices (VIs), and LAI were analyzed for consistency, with satellite LAI retrieved using a PROSAIL-based neural network.ResultsSurface reflectance exhibited strong overall consistency (R2 = 0.89 for Sentinel-2 and SuperDove, and R2 = 0.98 for Sentinel-2 and Dove-R), enabling overall band conversion regression equations. VIs showed high consistency within a 2-day interval: Dove-R and Sentinel-2 achieved R2 = 0.96 and RMSE = 0.06 for Normalized Difference Vegetation Index (NDVI), SuperDove and Sentinel-2 exhibited the strongest agreement in Normalized Difference Red-Edge Index (NDRE) (R2 = 0.79, RMSE = 0.09), highlighting the potential of red-edge-band-based VIs in satellite synergy. LAI consistency demonstrated robust performance across 0-5 day intervals (R2 = 0.89-0.97 for Dove-R and Sentinel-2, and R2 = 0.89-0.97 for SuperDove and Sentinel-2), with an optimal interval of 2 days.ConclusionThe synergy of Sentinel-2 and PlanetScope for continuous crop monitoring is feasible, with LAI being the most robust choice. Future studies should include more high-resolution satellites and other products (e.g., top-of-atmosphere reflectance) to enhance crop monitoring.

DOI:
10.1007/s11119-025-10226-4

ISSN:
1573-1618