Improved Rapid Magnitude Estimation for a Community‐Based, Low‐Cost MEMS Accelerometer Network

Abstract

Immediately following the M w 7.2 Darfield, New Zealand, earthquake, over 180 Quake-Catcher Network (QCN) low-cost micro-electro-mechanical systems accelerometers were deployed in the Canterbury region. Using data recorded by this dense network from 2010 to 2013, we significantly improved the QCN rapid magnitude estimation relationship. The previous scaling relationship (Lawrence et al., 2014) did not accurately estimate the magnitudes of nearby (< 35 km) events. The new scaling relationship estimates earthquake magnitudes within 1 magnitude unit of the GNS Science GeoNet earthquake catalog magnitudes for 99% of the events tested, within 0.5 magnitude units for 90% of the events, and within 0.25 magnitude units for 57% of the events. These magnitudes are reliably estimated within 3 s of the initial trigger recorded on at least seven stations. In this report, we present the methods used to calculate a new scaling relationship and demonstrate the accuracy of the revised magnitude estimates using a program that is able to retrospectively estimate event magnitudes using archived data.

Strong-Motion Observations of the M 7.8 Gorkha, Nepal, Earthquake Sequence and Development of the N-SHAKE Strong-Motion Network

“…Rapidly deployed QCN instruments can potentially contribute to real-time assessment of aftershock damage and perhaps ultimately to earthquake early warning. QCN RAMP experiments after the M 7.1 Darfield, New Zealand, earthquake and the M 8.8 Maule, Chile, earthquake have also proven effective for research and education purposes (Chung et al, 2011(Chung et al, , 2015Lawrence et al, 2014) and have provided scientists with a wealth of information to assess ground motion in earthquake prone regions.…”

Investigation of the high-frequency attenuation parameter, κ (kappa), from aftershocks of the 2010 Mw 8.8 Maule, Chile earthquake

“…To date, the sensors have primarily been used for rapid event detection (e.g. Chung et al 2011Chung et al , 2014Lawrence et al 2014), but as shown here may also provide valuable contributions to attenuation and site response studies by providing spatially dense data. Newer, higher resolution QCN sensors (16-bit) have been developed since this study, which yield a lower noise threshold and provide improved data quality for future studies.…”

On the Reliability of Quake-Catcher Network Earthquake Detections

“…Most of the false-positive events occur in M w 4-5 range. The magnitude relation that Lawrence et al (2014) used had an estimation bias centered around M w 4 (Chung et al, 2015). We attribute the fact that most false-positive events are in the M w 4-5 range to the bias in the magnitude formulation.…”