Multi-Decadal Space-Based Observations of Basaltic Effusive Eruptions from MODIS Infrared Data

Abstract

Lava discharge rate, the volume of lava emitted from a vent at any given time, is a critical parameter that volcanologists use to study effusive eruptions. It not only gives us information about the eruption dynamics but also is a key control on how far lava can flow. Thermal infrared satellite remote sensing has shown advantages for deriving lava discharge rate and monitoring temporal changes during an eruption, with a higher temporal resolution than in-situ measurements. In this dissertation, I use NASA’s MODIS (Moderate Resolution Imaging Spectroradiometer) sensor to study 104 basaltic effusive eruptions over a 15-year time period, to retrieve time-averaged discharge rates (TADR). I first show that the theoretical asymmetrical shape of TADR time-series can be used to predict the end of lava flow-forming eruptions from space. Then, I investigate the 2014-2015 Holuhraun eruption, Iceland, the country’s largest basaltic effusive eruption in the past 200 years. I compared the satellite- and ground-based TADR for this exceptional six month-long eruption. Final flow volumes estimated from both techniques were in good agreement. However, systematic differences between satellite and field TADR estimates the first 30 days of the eruption indicated that the satellite-derived method, and some of the assumptions on which it is based, need to be revised to reconcile these differences for lava flows that are areally extensive and highly radiant. Finally, I tested three different thermal satellite-based methods to retrieve TADR (Harris et al., 1997a; Wright et al., 2001a; and Coppola et al., 2012) to determine whether they yield comparable results, and assess the absolute accuracy of these approaches. I used the global TADR time-series database of basaltic effusive eruptions to estimate final flow volumes (by integrating the TADR time-series), which can be very well constrained at the end of an eruption unlike instantaneous discharge rates. Using final flow volumes from the literature and inferred volumes from satellite images, we demonstrated that the Harris et al. (1997a) method yields TADR closer to reality. However, we showed that the accuracy of TADR could be improved by updating the coefficient method (Wright et al., 2001a) with our eruption database.