20-17. Optimizing the characterization of small earthquakes for seismic forecasting

CLOSED

----------

Small earthquakes form an essential pillar of the methodology in use by the USGS and others for forecasting of larger earthquakes, whether on timescales of days to months (Operational Aftershock Forecasting - OAF) or years (National Seismic Hazard Model - NSHM).  In addition to their spatial distribution, the seismicity rates (Gutenberg-Richter a-values) and associated occurrence rates as a function of magnitude (Gutenberg-Richter b-values) strongly influence forecasted rates at larger magnitudes.  Although advances in seismic monitoring in recent decades have allowed consideration of increasingly small earthquakes in these forecasts, bringing increased spatial and temporal resolution, the magnitude characterizations themselves have not kept pace, introducing potentially large inconsistencies and biases.

An obvious challenge lies in the magnitude determination of small earthquakes.  The moment magnitude (M_w) scale is now established as the standard for moderate and large earthquakes (M_w>4), yet smaller earthquakes remain characterized by an inconsistent patchwork of magnitude scales.  In particular, transitions between reported magnitude types can create discontinuities, disrupting the observed distribution of earthquake magnitudes within the lower magnitude range, which most strongly controls a and b values.

The NSHM has recognized this problem and attempts to homogenize the underlying seismic catalog to “uniform moment magnitude” using a set of regionally dependent conversion relationships proposed by the CEUS-SSC report (EPRI/DOE/NRC, 2012).  However, these conversion relationships are challenging even in theory, as physical properties of the earthquake rupture such as stress drop can significantly affect the relationships among magnitudes.  Conversion relationships in current practice may not perform adequately even on an aggregate basis.  In fact, a recent case study of earthquakes near the Oklahoma/Kansas border showed that these conversion relationships can contribute to systematic biases of up to 0.5 magnitude units, potentially translating to a factor of ~3 error in seismicity rate, assuming typical b-values near 1.0 (Shelly et al., BSSA, 2021).  Furthermore, associated b-values can also have large biases. 

Given inherent real-time requirements of OAF, the current approach extracts the preferred event magnitudes from the ANSS Comprehensive Earthquake Catalog (ComCat), without applying any conversions.  However, this characterization is sensitive to transitions between magnitude types and associated anomalies in the apparent distribution of earthquake magnitudes.  These issues can bias the forecast or require restricting analysis to larger events, reducing the resolution of the forecasts.

One approach to addressing these challenges would be to create an operational system capable of determining moment magnitude for small earthquakes in near-real-time, thus allowing the full range (or nearly the full range) of earthquakes to be characterized by a single magnitude type.  Although this may be possible (for example, based upon coda envelope analysis (Mayeda et al., 2003)), significant resources would be required to bring such a system into routine operation.  Before embarking on such a journey, it is worth taking a step back to examine the use of small earthquakes in seismic forecasting more broadly. 

The research envisioned in this Mendenhall Opportunity is to take a holistic view of the use of small earthquakes in seismic forecasting.  We invite proposals to assess and address current issues, particularly related to earthquake magnitudes, in order to optimize our ability to forecast seismicity using small-magnitude earthquakes.  We expect this work to encompass research into both theoretical and practical aspects of earthquake characterization and forecasting, such as:

1) Research examining the underpinnings of seismic forecasting using small earthquakes.  To what extent do patterns of small earthquakes portend future activity?  And on what temporal and spatial scales? Does existing methodology fully capture the available information? Should we aim to characterize the complete range of magnitudes using M_w?  Is there a gain derived from utilizing multiple types of magnitude determination for a single event?

And,

2) Research focused on addressing related issues from a practical standpoint.  Which aspects have the most impact for seismicity forecasting?  How can we improve magnitudes of small earthquakes operationally in near-real-time?  What further improvements can come from specialized post-processing?  How can we best characterize and incorporate historical seismicity with magnitudes that may have larger uncertainty?  How can we better characterize catalog incompleteness, both on short (e.g., transient incompleteness following a large earthquake) and long timescales?

Interested applicants are strongly encouraged to contact the Research Advisor(s) early in the application process to discuss project ideas.

References:

Electric Power Research Institute/Department of Energy/Nuclear Regulatory Commission (EPRI/DOE/NRC) (2012). Central and eastern United States seismic source characterization for nuclear facilities, Final Report, EPRI, USDOE, and USNRC, Palo Alto, California, available at www.ceus-ssc.com

Mayeda, K., Hofstetter, A., O'Boyle, J. L., & Walter, W. R. (2003). Stable and transportable regional magnitudes based on coda-derived moment-rate spectra. Bulletin of the Seismological Society of America, 93(1), 224-239.

Shelly, D.R.,  K. Mayeda, J. Barno, K. M. Whidden, M. P. Moschetti, A. L. Llenos, J. L. Rubinstein, W. L. Yeck, P. S. Earle, R. Gök, and W. R. Walter (2021), A big problem for small earthquakes: benchmarking routine magnitudes and conversion relationships with coda-envelope-derived M_w in southern Kansas and northern Oklahoma, Bull. Seism. Soc. Am., doi: https://doi.org/10.1785/0120210115

Proposed Duty Station: Golden, Colorado; Moffett Field, California

Areas of PhD: Geophysics, seismology, or related fields (candidates holding a Ph.D. in other disciplines, but with extensive knowledge and skills relevant to the Research Opportunity may be considered).

Qualifications: Applicants must meet the qualifications for:  Research Geophysicist. 

(This type of research is performed by those who have backgrounds for the occupations stated above.  However, other titles may be applicable depending on the applicant's background, education, and research proposal. The final classification of the position will be made by the Human Resources specialist.)

Human Resources Office Contact:  Megan Agy, 303-236-9584, magy@usgs.gov