MEMS Accelerometers Mini-Array (MAMA) - initial results and lessons learned

Most operational earthquake early warning systems (EEWS) consider earthquakes to be point-sources and have difficulty providing imminent and robust source locations and magnitudes, especially at the edge of the seismic network or where seismic stations are sparse. Mini-arrays have the potential to estimate reliable hypocentral locations by beam forming (FK-analysis) techniques. They can also characterize the rupture dimensions and account for finite-source effects, leading to more reliable estimates of ground motions for large magnitude earthquakes. In the past, the high price of multiple seismometers has made creating arrays cost-prohibitive. Here, we present a setup of two mini-arrays of a new low-cost (<$150) seismic acquisition unit based on a high-performance MEMS accelerometer around conventional seismic stations. The expected benefits of such an approach include decreasing alert-times, improving real-time shaking predictions and mitigating false alarms. We will present our new 24-bit device details, benchmarks, and results from two MEMS Accelerometers Mini-Array (MAMA) deployed at the UC Berkeley and Humboldt State University campuses. The new device shows lower noise levels than the currently available off-the-shelf 16-bit sensors, commonly used by several citizen-science projects (e.g. QCN, CSN, MyShake, etc.). This lower noise level enables us to record and process lower magnitude events. We show examples of back-azimuth calculations of Magnitude 2.7 - 5.1 events and at a range of 5 to 110km from the MAMA center and discuss some of the limitations and considerations of the MAMA deployments.