English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Active volcanoes are globally distributed and pose societal risks at multiplegeographic scales, ranging from local hazards to regional/internationaldisruptions. Many volcanoes do not have continuous ground monitoring networks;meaning that satellite observations provide the only record of volcanicbehavior and unrest. Among these remote sensing observations, thermal imageryis inspected daily by volcanic observatories for examining the early signs,onset, and evolution of eruptive activity. However, thermal scenes are oftenobstructed by clouds, meaning that forecasts must be made off image sequenceswhose scenes are only usable intermittently through time. Here, we exploreforecasting this thermal data stream from a deep learning perspective usingexisting architectures that model sequences with varying spatiotemporalconsiderations. Additionally, we propose and evaluate new architectures thatexplicitly model intermittent image sequences. Using ASTER Kinetic SurfaceTemperature data for $9$ volcanoes between $1999$ and $2020$, we found that aproposed architecture (ConvLSTM + Time-LSTM + U-Net) forecasts volcanictemperature imagery with the lowest RMSE ($4.164^{\circ}$C, other methods:$4.217-5.291^{\circ}$C). Additionally, we examined performance on multiple timeseries derived from the thermal imagery and the effect of training with datafrom singular volcanoes. Ultimately, we found that models with the lowest RMSEon forecasting imagery did not possess the lowest RMSE on recreating timeseries derived from that imagery and that training with individual volcanoesgenerally worsened performance relative to a multi-volcano data set. This workhighlights the potential of data-driven deep learning models for volcanicunrest forecasting while revealing the need for carefully constructedoptimization targets.

Improving the Thermal Infrared Monitoring of Volcanoes: A Deep Learning Approach for Intermittent Image Series