Volcano Forecasting Methods Using Statistical Data Assimilation

Posted on 09.13.2018
Inflation of six volcanoes revealed by an averaged 2006-2009 ground velocity map of west Sunda, Indonesia, from ALOS InSAR. Positive velocity (red colors) represents movement towards the satellite (inflation). Insets: zoom into the six inflating volcanoes. Three erupted (Sinabung, Kerinci and Slamet). From Chaussard & Amelung [2012].

Breakthrough technological advances in the form of near-real time geodetic observations from Interferometric synthetic aperture radar (InSAR) can provide critical early warning of volcanic activity. Eruptions are typically preceded by the ascent of new magma to shallow storage levels, resulting in swelling of the ground surface, which is observable with satellites. However, there are only few examples in which InSAR contributed to crises assessment, largely because not enough observations were available, because the interpretation of inflation in terms of eruption potential was ambiguous, and because it was difficult to resolve small changes because of the InSAR noise.

We are currently funded by NASA and the National Science Foundation to improve on previous approaches for eruption prediction along three fronts. First, we are utilizing near-real time SAR imagery provided for volcanic crises by the Global Earth Observation System of Systems (GEOSS) to derive near-real time geodetic data (“volcano event Supersites”). Second, we are utilizing finite element methods and geodynamical magma chamber evolution models in which ground deformation depends on the chamber geometry, the host-rock viscosity, and the magma overpressure. Third, we are integrating these modeling approaches with the real-time InSAR data using a model-data assimilation framework based on Ensemble Kalman Filtering (EnKF). This framework will provide an updated estimate of the magma overpressure whenever a new SAR image is acquired. The deliverable of this project is an operational methodology for eruption prediction for the DESDYnI era.

We are looking for graduate students to participate in this project! Please click on the link for prospective students for more information on how to apply.

COLLABORATORS: Falk Amelung (U of Miami), Estelle Chaussard (SUNY Buffalo), Yosuke Aoki (University of Tokyo), Zhong Lu (Southern Methodist University), Jeff Freymueller (Michigan State University), Hélène Le Mével (Carnegie DTM), Dennis Geist (NSF / Colgate University)

FUNDING: This work is currently funded by grants from NASA and the National Science Foundation.

PUBLICATIONS & PRESENTATIONS:
§ Student Advisee 

F. Albino, F. Amelung, and P. M. Gregg, The Role of Pore Fluid Pressure on the Failure of Magma Reservoirs: Insights From Indonesian and Aleutian Arc Volcanoes. J. Geophys. Res, 123, 1328-1349, 10.1002/2017JB014523, 2018. 

§J. Albright, P. M. Gregg, Z. Lu, J. Freymueller, Tracking reservoir stability through multi-data stream statistical data assimilation: Application to the 2008 eruption of Okmok, AK, AGU Fall Meeting, 2018. 

P. M. Gregg, Y. Zhan, J. A. Albright, Z. Lu, J. Freymueller, F. Amelung, Imaging volcano deformation sources through geodetic data assimilation, UNAVCO Science Workshop, 2018, INVITED.

P. M. Gregg, H. Le Mével, Y. Zhan, J. A. Albright, H. E. Cabaniss, Linking thermomechanical models with geodetic observations to assess magma reservoir evolution and stability, AGU Chapman Conference on Merging Geophysical, Petrochronologic, and Modeling Perspectives of Large Silicic Magma Systems, January 2018, INVITED.

§Y. Zhan, P. M. Gregg, E. Chaussard, Y. Aoki, Sequential Assimilation of Volcanic Monitoring Data to Quantify Eruption Potential: Application to Kerinci Volcano, Sumatra, Front. Earth Sci., doi.org/10.3389/feart.2017.00108, 2017.

§Y. Zhan and P. M. Gregg, Data assimilation strategies for volcano geodesy, J. Volcanol. Geotherm. Res., doi:10.1016/j.jvolgeores.2017.02.015, 2017.

§J. A. Albright, P. M. Gregg, Z. Lu, J. Freymueller, Hind-casting the 2008 eruption of Okmok, AK using multi-data stream statistical data assimilation, IAVCEI Scientific Assembly, Portland, OR, 2017.

§Y. Zhan, P. M. Gregg, E. Chaussard, Y. Aoki, Investigation of volcanic unrest in Indonesia using statistical data assimilation with a two-source magma chamber model, IAVCEI Scientific Assembly, Portland, OR, 2017. 

F. Albino, F. Amelung, P. M. Gregg, How pore-fluid pressure due to heavy rainfall influences volcanic eruptions, example of 1998 and 2008 eruptions of Cerro Azul (Galapagos), Cities on Volcanoes 9, Chile, 2016.

§Y. Zhan, P. M. Gregg, E. Chaussard, Y. Aoki, Investigation of volcanic unrest in Indonesia using statistical data assimilation with a two-source magma chamber model, IAVCEI Scientific Assembly, Portland, OR, 2017.

F. Albino, F. Amelung, P. M. Gregg, How pore-fluid pressure due to heavy rainfall influences volcanic eruptions, example of 1998 and 2008 eruptions of Cerro Azul (Galapagos), Cities on Volcanoes 9, Chile, 2016. 

§*Y. Zhan and P. M. Gregg, Data Assimilation Strategies for Volcano Geodesy, Goldschmidt, 2016.

J. C. Pettijohn and P. M. Gregg, A model-data fusion approach for assessing volcanic unrest, Goldschmidt, 2016.

P. M. Gregg, H. Le Mével, J. Dufek, Linking ground deformation to magma injection and volatile exsolution in a rapidly evolving magma chamber, Goldschmidt, 2016.

P.M. Gregg and J.C. Pettijohn, A multi-data stream assimilation framework for the assessment of volcanic unrest, J. Volcano. Geotherm. Res, 10.1016/j.jvolgeores.2015.11.008, 2016.

P. M. Gregg and J. C. Pettijohn, A multi-data stream assimilation framework for the assessment of volcanic unrest, AGU Fall Meet. Suppl., 2015

J. C. Pettijohn, P. M. Gregg, and Y. Zhan, Sequential data assimilation strategies for utilizing ground deformation data to assess rapidly evolving magma reservoirs, AGU Fall Meet. Suppl., 2015.