Early Detection of Forest Calamities in Homogeneous Stands -- Deep Learning Applied to Bark-Beetle Outbreaks

TL;DR

The paper tackles the challenge of early bark-beetle disturbance detection in homogeneous spruce stands using a memory-efficient, unsupervised LSTM Autoencoder applied to Sentinel-2 time series. By training on healthy stand data and evaluating across window lengths, the approach achieves high accuracy and demonstrates substantial early detection of anomalies before visible defoliation, outperforming the BFAST method on multiple vegetation indices. A novel Pre-Defoliation Score is introduced to reward timelier warnings in the absence of precise ground-truth labels, contributing a practical metric for operational monitoring. The work advances wall-to-wall forest health monitoring by combining multivariate spectral-temporal analysis with real-time anomaly detection, paving the way for scalable, data-efficient forest management tools.

Abstract

Climate change has increased the vulnerability of forests to insect-related damage, resulting in widespread forest loss in Central Europe and highlighting the need for effective, continuous monitoring systems. Remote sensing based forest health monitoring, oftentimes, relies on supervised machine learning algorithms that require labeled training data. Monitoring temporal patterns through time series analysis offers a potential alternative for earlier detection of disturbance but requires substantial storage resources. This study investigates the potential of a Deep Learning algorithm based on a Long Short Term Memory (LSTM) Autoencoder for the detection of anomalies in forest health (e.g. bark beetle outbreaks), utilizing Sentinel-2 time series data. This approach is an alternative to supervised machine learning methods, avoiding the necessity for labeled training data. Furthermore, it is more memory-efficient than other time series analysis approaches, as a robust model can be created using only a 26-week-long time series as input. In this study, we monitored pure stands of spruce in Thuringia, Germany, over a 7-year period from 2018 to the end of 2024. Our best model achieved a detection accuracy of 87% on test data and was able to detect 61% of all anomalies at a very early stage (more than a month before visible signs of forest degradation). Compared to another widely used time series break detection algorithm - BFAST (Breaks For Additive Season and Trend), our approach consistently detected higher percentage of anomalies at an earlier stage. These findings suggest that LSTM-based Autoencoders could provide a promising, resource-efficient approach to forest health monitoring, enabling more timely responses to emerging threats.

Figures (16)

• Figure 1: Overview of the study area and training data.
• Figure 2: Autoencoder: Reconstruction of a time series.
• Figure 3: Components of a LSTM-Unit.
• Figure 4: LSTM Autoencoder Architecture.
• Figure 5: Methodology Outline