Detection and localisation of wadi flow events utilizing seismic sensors

In Oman, the frequency of flash floods has significantly increased in recent years. This phenomenon is correlated with climate change, resulting in an intensification of the atmospheric water cycle. Consequently, a further escalation of flash floods can be anticipated in the future. In Oman, the issue of flash floods is exacerbated by the frequent occurrence of tropical cyclones. Furthermore, the rapid expansion of urban areas, in some cases extending directly into wadis, coupled with the advancing sealing of the ground and insufficient drainage systems, leads to an increased risk of flooding. This is accompanied by substantial property damage and recurring loss of life.

Despite the growing danger posed by flash floods, there is currently no early warning system for precise prediction of these events in Oman. To establish such a system, densely distributed networks for rainfall and water level measurements would be required. However, due to the challenging topography and vastness of the country, implementing such networks is currently not feasible.

Recent studies have shown that seismic sensors could be used for measuring flow conditions. Further, seismic networks could be utilized to detect and track extreme flow events. The increasing availability of low-cost seismic sensors opens up the possibility of instrumenting previously ungauged wadi systems. However, the question remains if seismic networks can pick up smaller flow events and flow events happening in multiple smaller catchments at the same time.

In this study we used flow data from wadi gauge stations in the Al-Batinah Region (NW Oman) and data from broadband seismometers of the Earthquake Monitoring Center to research how flow events of various sizes can be detected by seismic networks. Initial results suggest that flow regimes in wadi systems offer favourable conditions for detection, as they mainly change between flow and no flow conditions. As the amplitude of seismic signals decreases with distance from the source, detection range is limited by background noise. To overcome this, low-cost seismic sensors have recently been installed in a wadi system together with camera based river gauges. Further work utilizing this data is currently ongoing.