Un APPORT DES ARCHIVES SPOT WORLD HERITAGE À L’ESTIMATION DES TENDANCES DE PRODUCTIVITÉ DES PRAIRIES DE MONTAGNE

Abstract

ong-term monitoring of mountain grassland productivity is crucial for understanding these ecosystem’s responses to cli-
mate change. Although Landsat archives are the commonly used archives for such analyses, their temporal frequency (16
days) and frequent cloud cover often complicate the capture of the annual peak vegetation (N DV Imax). This study quanti-
fies the contribution of SPOT World Heritage (SWH) archives to densify time series over the Vosges Mountains Grasslands
(1985-2015). Using a Generalized Additive Model (GAM) on 927,575 observations, we demonstrate that differences in
peak productivity (N DV Imax) are dependent on the timing of SPOT observations (∆DOY), correcting systematic Landsat
underestimations. While the overall consistency between Landsat-only and combined strategies (Landsat + SPOT) is high
(R2 = 0.83), the addition of SPOT modifies the statistical significance of trends for 8% of pixels, particularly in areas with
rapid dynamics. These results highlight the importance of historical archives for stabilizing climate diagnostics in mountain
areas and advocate for the wider availability of Level-2A products to mitigate calibration biases.