Spatiotemporal variations of earthquake source parameters in the aswan high dam reservoir region: constraints on reservoir-modulated seismicity

Abstract

This study investigates earthquake source characteristics and their spatiotemporal variability in the Aswan High Dam Reservoir (AHDR) region using 1,193 earthquakes (\(0.76 \le M_L \le 4.2\)) recorded during 2010–2022. Source-parameter analysis is based on 1,175 events with \(M_L\le 3.0\), which were used for spatiotemporal analyses. Seismic moment (\(M_0\)) and S-wave corner frequency (\(f_c\)) were estimated from path- and site-corrected spectra using Brune’s \(\omega ^2\) source model, and apparent stress drop was calculated assuming a circular seismic source. For the 1,175-event source-parameter dataset, the mean corner frequency is 13.61 Hz and the median apparent stress drop is 0.688 bar. Earthquake-level bias-corrected and accelerated bootstrap resampling yields 95% confidence intervals of 13.46–13.79 Hz for mean \(f_c\) and 1.251–1.578 bar for mean apparent stress drop. Generalized Orthogonal Regression gives \(\log _{10}(f_c/\textrm{Hz}) = 6.698-0.306\log _{10}[M_0/(\mathrm {dyn\,cm})]\), with a 95% bootstrap confidence interval of \([-0.3377,-0.2808]\) for the slope. This interval includes the self-similar value of \(-1/3\), indicating that the source scaling of the \(M_L\le 3.0\) earthquakes is statistically compatible with constant-stress-drop scaling under the adopted source and error models (error-variance ratio \(\lambda =0.005\); this compatibility is not retained for \(\lambda \ge 0.006\)). Inclusion of all 1,193 fitted events yields a shallower slope of \(-0.2706\) (95% CI: \([-0.2968,-0.2481]\)), indicating sensitivity of the inferred scaling to the analyzed magnitude range. The 1,175 events show significant spatial clustering but no statistically significant recurrent monthly seasonality (Friedman test across years, \(\chi ^2=18.5\), 11 d.f., \(p=0.071\)). Gutenberg–Richter analysis of the complete catalog gives b-values of \(0.88\pm 0.06\), \(0.90\pm 0.04\), and \(1.92\pm 0.38\) for 2010–2013, 2014–2017, and 2018–2022, with completeness magnitudes of 1.8, 1.5, and 2.0; the 2018–2022 estimate rests on only 25 events and is poorly constrained. The results indicate substantial source heterogeneity within a fault-controlled seismic system. Although hydrological modulation is physically plausible, the observational record does not independently establish reservoir forcing, as it lacks concurrent measurements of both reservoir level and subsurface pore pressure.