TY - JOUR

T1 - Further Notes on the New Empirical Earthquake Source-Scaling Laws

AU - Thingbaijam, Kiran Kumar

AU - Mai, Paul Martin

N1 - KAUST Repository Item: Exported on 2020-04-23
Acknowledgements: The research presented in this article is supported by King Abdullah University of Science and Technology (KAUST) in Thuwal, Saudi Arabia

PY - 2020

Y1 - 2020

N2 - We articulate a few additional points, relevant to the empirical earthquake source-scaling laws
(Thingbaijam et al., 2017). First, we report that adjustments, due to improved theoretical
derivation of error-variance ratios, do not impact the general orthogonal regressions
significantly. Hence, the earlier results are not affected. Second, we perform partial correlation
analysis based on the residuals, to investigate influencing parameter(s) on the scaling relations. It
is observed that rupture length L correlates poorly with rupture width W, if not influenced by
magnitude MW. Therefore, overall relation between L and W is controlled by how they scale with
MW independently. On the other hand, average slip correlates negatively with L and more
strongly so with W, if influence of MW is eliminated. Third, we demonstrate that Monte Carlo
approach would be appropriate while applying the scaling laws for predictions, to account for the
associated data scatter. In case of uncertain independent variable, such approach allows capture
the error propagations effectively. Lastly, we evaluate the effect of finite seismogenic depth, in
estimation of maximum magnitude for near-vertical strike-slip faults.

AB - We articulate a few additional points, relevant to the empirical earthquake source-scaling laws
(Thingbaijam et al., 2017). First, we report that adjustments, due to improved theoretical
derivation of error-variance ratios, do not impact the general orthogonal regressions
significantly. Hence, the earlier results are not affected. Second, we perform partial correlation
analysis based on the residuals, to investigate influencing parameter(s) on the scaling relations. It
is observed that rupture length L correlates poorly with rupture width W, if not influenced by
magnitude MW. Therefore, overall relation between L and W is controlled by how they scale with
MW independently. On the other hand, average slip correlates negatively with L and more
strongly so with W, if influence of MW is eliminated. Third, we demonstrate that Monte Carlo
approach would be appropriate while applying the scaling laws for predictions, to account for the
associated data scatter. In case of uncertain independent variable, such approach allows capture
the error propagations effectively. Lastly, we evaluate the effect of finite seismogenic depth, in
estimation of maximum magnitude for near-vertical strike-slip faults.

UR - http://hdl.handle.net/10754/661553

M3 - Article

JO - Bull. Seis. Soc. Am.

JF - Bull. Seis. Soc. Am.

ER -