Improving the Thermal Infrared Monitoring of Volcanoes: A Deep Learning Approach for Intermittent Image Series
This work addresses the problem of improving volcanic unrest forecasting for observatories and risk management by handling intermittent thermal imagery, but it is incremental as it builds on existing deep learning architectures with specific adaptations.
The paper tackled forecasting volcanic thermal imagery from intermittent satellite data using deep learning, achieving a lowest RMSE of 4.164°C with a proposed architecture (ConvLSTM + Time-LSTM + U-Net) compared to other methods (4.217-5.291°C). It also found that models best at forecasting imagery did not perform best on derived time series, and training on individual volcanoes generally worsened performance relative to a multi-volcano dataset.
Active volcanoes are globally distributed and pose societal risks at multiple geographic scales, ranging from local hazards to regional/international disruptions. Many volcanoes do not have continuous ground monitoring networks; meaning that satellite observations provide the only record of volcanic behavior and unrest. Among these remote sensing observations, thermal imagery is inspected daily by volcanic observatories for examining the early signs, onset, and evolution of eruptive activity. However, thermal scenes are often obstructed by clouds, meaning that forecasts must be made off image sequences whose scenes are only usable intermittently through time. Here, we explore forecasting this thermal data stream from a deep learning perspective using existing architectures that model sequences with varying spatiotemporal considerations. Additionally, we propose and evaluate new architectures that explicitly model intermittent image sequences. Using ASTER Kinetic Surface Temperature data for $9$ volcanoes between $1999$ and $2020$, we found that a proposed architecture (ConvLSTM + Time-LSTM + U-Net) forecasts volcanic temperature imagery with the lowest RMSE ($4.164^{\circ}$C, other methods: $4.217-5.291^{\circ}$C). Additionally, we examined performance on multiple time series derived from the thermal imagery and the effect of training with data from singular volcanoes. Ultimately, we found that models with the lowest RMSE on forecasting imagery did not possess the lowest RMSE on recreating time series derived from that imagery and that training with individual volcanoes generally worsened performance relative to a multi-volcano data set. This work highlights the potential of data-driven deep learning models for volcanic unrest forecasting while revealing the need for carefully constructed optimization targets.