UAV and AI for large-scale monitoring of the effects of climate change on mountain lake areas: the ACLIMO cross-border cooperation project

Abstract. Studying the effects of climate change is essential to understanding the impact of human activities on the environment and to developing effective mitigation and adaptation strategies. This is the objective of the ACLIMO project, which applies a multi-sensor and multiscale approach to analyse the effects of climate change on a border alpine area and specific ecosystems. In this contribution, we focus on the large-scale objectives, analysing vegetation coverage around two alpine lakes (Lake Brocan and Lake Vej del Bouc) and their bathymetry. In particular, through automatic classification of drone imagery with an Object-Based Image Analysis (OBIA) workflow, five machine learning algorithms (Bayesian, Random Forest, Support Vector Machine, K-Nearest Neighbours, Decision Tree) were tested for automatic classification of vegetation. Field surveys were conducted to collect in situ vegetation data, providing ground-truth points for classification validation. In addition, bathymetric mapping was carried out using a USV (Uncrewed Surface Vessel) equipped with a single-beam echo sounder, serving as ground-truth for bathymetric models derived via Structure-from-Motion analysis of UAV images. This integrated and in-depth methodology enabled the generation of detailed land cover maps highlighting dominant vegetation species and accurate 3D bathymetric models, allowing for a comprehensive ecological assessment of this alpine environment under ongoing climate change conditions and establishing a starting point (t0 data) for future monitoring and change detection analyses.