Skip to content

A next generation, near-real-time volcanic eruption monitoring algorithm for use with both MODIS Aqua/Terra and SNPP VIIRS, providing continuity with the MODIS/EOS era MODVOLC system

Nikola Rogic, Eric Pilger, Andrea Gabrieli, Robert Wright

  • Affiliations: Hawaiʻi Institute of Geophysics and Planetology, University of Hawaiʻi at Mānoa, Honolulu, Hawaii, United States

  • Presentation type: Poster

  • Presentation time: Monday 16:30 - 18:30, Room Poster Hall

  • Poster Board Number: 225

  • Programme No: 2.4.15

  • Theme 2 > Session 4

Abstract

The MODVOLC system has provided 24 years of global data regarding volcanic thermal unrest at near daily temporal frequency (https://modis.higp.hawaii.edu). With the EOS era coming to an end, we have developed a new approach which will be applied retroactively to the Terra/Aqua MODIS archive, as well as for the near-real-time analysis of data from current and future VIIRS sensors. The new approach has taken decadal, global archives of Terra-MODIS, Aqua-MODIS and VIIRS data, to determine how 'hot' each km2 of Earth's land surface has been in the absence of a volcanic heat source. This improved algorithm will work simply by comparing the 4μm radiance for each pixel in a 'new' MODIS/VIIRS image with its geographic counterpart in the historic reference image. We will present results describing how the new generation algorithm was developed, as well as its performance relative to i) the existing MODVOLC algorithm, and ii) the MODIS Fire product for both day and night images. The new algorithm provides improved detection capabilities of as much as 200% (in terms of number of valid detections). Importantly, the computational complexity of the new algorithm is even simpler than the MODVOLC algorithm, involving a single floating-point comparison operation per pixel. This characteristic has an important implication for future detection algorithms for analysis of high spatial resolution thermal data, for which the order of magnitude increase in the number of pixels to be analyzed requires a computationally simple algorithm, especially if processing is to occur on-board the spacecraft for real-time product delivery.