Uncertainty-aware tree height change regression

2026-07-01

AI Summary

This study addresses the limitations of conventional binary paradigms in remote sensing change detection, which overlook the continuous nature of canopy height dynamics and associated label uncertainty. To bridge this gap, the authors construct a high-resolution (3 m) continuous Canopy Height Change (CHC) dataset spanning 10,598 km² across Spain, introducing spatially explicit uncertainty annotations for the first time. They propose a novel uncertainty-aware change regression task and develop tailored fine-tuning strategies leveraging PlanetScope time-series imagery and geospatial foundation models (GFMs), alongside dedicated evaluation metrics. This work advances remote sensing change monitoring from classification toward regression-based paradigms. Experimental results demonstrate both the potential and limitations of GFMs in estimating continuous change, offering a new methodological framework and foundational dataset for high-precision carbon sink monitoring.

Abstract

Monitoring canopy height change is essential for understanding carbon sinks and forest dynamics. Remote sensing enables consistent, large-scale observations of such changes, increasingly integrated with deep learning architectures such as Geospatial Foundation Models (GFMs). However, existing methods and datasets frame the problem as binary change detection, which overlooks both the continuous nature of change, especially for vegetation, and the inherent uncertainty in labels. We present the Canopy Height Change (CHC) dataset, providing 3 $\mathrm{m}$ resolution continuous canopy height differences and associated spatially resolved uncertainties across 10598 $\mathrm{km}^2$ of northern and western Spain. The dataset is paired with a co-located time series of PlanetScope satellite imagery. Based on the dataset, we introduce the task of uncertainty-aware change regression, associated metrics and strategies for fine-tuning GFMs. Furthermore, we evaluate state-of-the-art GFMs and highlight promising directions and remaining challenges for advancing continuous canopy height change estimation.

Problem

• canopy height change
• uncertainty-aware regression
• remote sensing
• continuous change
• Geospatial Foundation Models

Innovation

• uncertainty-aware regression
• canopy height change
• Geospatial Foundation Models
• continuous change detection
• remote sensing