Abstract ID: 3.12857 | Accepted as Talk | Talk | TBA | TBA

Drought events are anticipated to become more frequent and intense in the future, resulting in consequences for forest productivity, water use strategies, and ecosystem services. As forests play a crucial role in climate regulation, assessing the impact of droughts on these ecosystems is essential for improving the accuracy of climate change projections. Remote sensing (RS) techniques have been widely used to evaluate droughts impacts over large areas, but their effectiveness is often limited by the lack of ground-data for calibration and validation. This study aims to address this gap by integrating RS data with a unique, extensive time series of ground-based observations. We focus on the drought events of the years 2015, 2018 and 2022 in two alpine valleys: Lötschental, in Switzerland, and Mazia Valley, in South Tyrol, Italy. The RS dataset consists of time series of spectral indices derived from MODIS (Moderate-resolution Imaging Spectroradiometer) and HLS (Harmonized Landsat and Sentinel-2) imagery, estimates of Gross Primary Production (GPP) and other biophysical variables. The ground-based dataset includes physiological observations on coniferous species collected from 2007 to 2023 over our two study areas, such as xylogenesis imagery, dendrometer series, and wood anatomical data. First, we plan to identify RS indicators that are most sensitive to tree-level carbon dynamics in response to meteorological drought conditions. Multivariate modeling will be employed to quantify the relationships between ground-based and RS data. Next, a model will be developed to estimate tree-level carbon impacts based on RS indicators and auxiliary datasets, such as topography and climate data. With this work, funded by the CALEIDOSCOPE project, we aim to provide new insights on the relationships between ground-based carbon sink quantifications and broad-scale, carbon source-derived RS estimates during and after drought events.