Central America has an elevated seismic risk, resulting from the vulnerability of the building stock and steady population growth. Earthquake Early Warning (EEW) aims to provide warning in advance of imminent shaking, allowing recipients to take action and reduce casualties during damaging motions. The Swiss Seismological Service (SED) has been collaborating with local seismic agencies to develop national EEW systems across Central America, which can potentially benefit nearly 47 million inhabitants. We conducted a public survey to comprehend the desire for EEW, the preferences for EEW attributes, and the current behaviour of people during earthquakes and the driving factors behind it. We recruited participants from Nicaragua (N = 513), Costa Rica (N = 1350), Guatemala (N = 559), and El Salvador (N = 491). In all four countries, participants consider it necessary to have an EEW system, are tolerant of false alerts, and are likely to react promptly to alerts. The desirable alert threshold is for low felt intensities, ranging between MMI III to IV. We found that a significant number of respondents already take protective action when earthquakes strike, and appropriate reactions are expected to increase when EEW is available. Our survey is unique in providing insights into the social dimension of EEW systems in low-income regions with high earthquake risk and where no operational EEW system yet exists.

A noticeable decrease in seismic noise was registered worldwide during the lockdown measures implemented in 2020 to prevent the spread of COVID-19. In Central America, strong lockdown measures started during March of 2020. In this study, we used seismic stations from Costa Rica, Guatemala, El Salvador, and Nicaragua to study the effects of these measures on seismic records by characterizing temporal variations in the high-frequency band (4–14 Hz) via spectral and amplitude analyses. In addition, we studied the link between the reduction in seismic noise and the number of earthquake detections and felt reports in Costa Rica and Guatemala. We found that seismic stations near the capitals of Costa Rica, Guatemala, and El Salvador presented a decrease in their typical seismic noise levels, from 200 to 140, from 100 to 80, and from 120 to 80 nm, respectively. Our results showed that the largest reduction of ∼ 50 % in seismic noise was observed at seismic stations near main airports, busy roads, and densely populated cities. In Nicaragua, the seismic noise levels remained constant (∼ 40 nm), as no lockdown measures were applied. We suggest that the decrease in seismic noise levels may have increased earthquake detections and the number of felt reports of low-magnitude earthquakes. However, the variations observed in several seismic parameters before and after the lockdown are not significant enough to easily link our observations or separate them from other contributing factors. Our results imply that the study of seismic noise levels can be useful to verify compliance with lockdown measures and to explore their effects on earthquake detection and felt reports.

The national seismic networks in Central America have been developing network-based early warning since 2016 for Nicaragua, 2018 for El Salvador and 2019 for Costa Rica. This effort is part of a project with the Swiss Seismological Service (ETH Zurich) including funds for accelerograph deployment. At each network, delay for first earthquake parameter estimations have been significantly reduced by optimizing data acquisition, metadata quality, and configuration of the EEW algorithms implemented in SeisComP3, i.e. Virtual Seismologist and the Finite fault rupture Detector. Issues remain with significant numbers of deployed instrumentation that for a variety of reasons, do not optimally contribute to the EEW systems. Building on our experience so far, we design national network upgrades that will optimize the earthquake early warning performance in the Central America region, mitigating the current issues with velocimeter clipping during large events, datalogger delays, and incomplete network coverage. The new instruments have been selected after testing all available EEW-capable accelerographs natively compatible with SeisComP3 including class A force balance accelerometers as well as MEMs. To justify our instrument selection, we summarize the performance of these different instruments. We model and discuss reference maps for performance expectations, and present planned instrument vaults. Our primary focus is on minimizing first alert times but we also wish to accentuate the broad value of the network upgrade for seismological monitoring showing changes in the magnitude of completeness in the region. We demonstrate the value of the network upgrade for earthquake early warning with real-time processing simulation using synthetic data for the maximum magnitude earthquake expected for the Central America subduction zone.