Rapid Earthquake Source Description Using Variometric-Derived GPS Displacements toward Application to the 2019 Mw 7.1 Ridgecrest Earthquake

Abstract

Using near-field high-rate Global Positioning System (GPS) displacements to invert for earthquake fault slips in real time has the potential to improve the accuracy of earthquake early warning or tsunami early warning. For such applications, real-time retrieval of high-accuracy GPS displacements is essential. Here, we report on rapid modeling of the 2019 Mw 7.1 Ridgecrest earthquake with real-time GPS displacements derived from a variometric approach with readily available broadcast ephemeris. This method calculates station variations in real time by differencing continuous phase observations and does not rely on precise orbit and clock information. The phase ambiguity is also removed, and thus the method does not suffer from a relatively long convergence time. To improve the accuracy of variometric displacements, we use a local spatial filter to decrease the influence of residual errors that cannot be removed completely by the time difference. We invert for the centroid moment tensor, static fault slips, and fault rupture process from the derived displacements. Our results show that all inverted models are available within about 65 s after the origin time of the earthquake and are comparable with models inverted by real-time precise point positioning displacements. This study highlights the great value of variometric displacements for the rapid earthquake source description with only broadcast ephemeris.

Real-Time Source Modeling of the 2022 Mw 6.6 Menyuan, China Earthquake with High-Rate GNSS Observations

“…where u k i (t) refers to the ith component of coseismic offset measured at the kth station at the time t; m j (t) is the jth component of the moment tensor at the time t; G k i,j is the green function that relates the jth component of the moment tensor to the kth station. In this study, the real-time static coseismic offsets were retrieved by smoothing RT-PPP-derived displacements with a 30-s sliding window [39]. The Green's functions were determined by interpolating outputs of EDGRN program [56] with 1-D layered velocity structure from Crust 1.0 (https://igppweb.ucsd.edu/~gabi/crust1.html (accessed on 2 September 2022)).…”

“…Considering the relatively low accuracy of GNSS displacements in the up component, the relative weight between horizontal coseismic offsets and up coseismic offsets was set to 3:1. In addition, a distance-related weighting scheme adopted by Zang et al [39] was also used where stations with epicenter distance less than 50 km were weighted with the factor of 4, stations with epicenter distance between 50 and 100 km were weighted with the factor of 3, stations with epicenter distance between 100 and 150 km were weighted with the factor of 2, and stations with epicenter distances exceeding 150 km were weighted with the factor of 1.…”

“…All these studies have demonstrated the great potential of high-rate GNSS for real-time earthquake source modeling. However, many studies focus on large earthquakes [34][35][36][37][38][39][40]; there are limited studies about real-time earthquake source modeling with high-rate GNSS for earthquakes with a magnitude less than Mw 7.…”

Line‐Source Model Based Rapid Inversion for Deriving Large Earthquake Rupture Characteristics Using High‐Rate GNSS Observations

“…Zang et al. (2021) investigated the earthquake source of the 2019 M w 7.1 Ridgecrest earthquake by using variometric‐derived GNSS displacements, but variometric‐derived displacements may have non‐linear drift which is hard to completely remove by the linear filter or the spatial filter (Fratarcangeli, Savastano, et al., 2018; Shu, Fang, Li, et al., 2018).…”

“…The advantage of this technology is that it uses merely broadcast ephemeris instead of precise orbit and clock products like PPP, allowing the GNSS monitoring device to operate independently like an accelerometer (Benedetti et al, 2014(Benedetti et al, , 2015Fratarcangeli, Ravanelli, et al, 2018). Zang et al (2021) investigated the earthquake source of the 2019 M w 7.1 Ridgecrest earthquake by using variometric-derived GNSS displacements, but variometric-derived displacements may have non-linear drift which is hard to completely remove by the linear filter or the spatial filter (Fratarcangeli, Savastano, et al, 2018;Shu, Fang, Li, et al, 2018).…”