Water quality parameters as early warning indicators in earthquake risk management: A case of study Mount Kinabalu in the district of Ranau, Sabah in Malaysia

Abstract

Abstract On June 5, 2015, a 6.0 magnitude earthquake shook Mount Kinabalu in the region of Ranau, Sabah. By integrating and increasing synergies among the components of effective early warning systems, impact-based forecasting and warning services aim to bridge the gap between warning information producers and users. This study evaluates the water quality of the Liwagu River, which is located in the earthquake zone of the 2015 Ranau Earthquake, to see if it may be used as an early warning indicator for earthquake risk management in the area. From 2013 to 2019, data on 11 parameters was gathered and recorded monthly to assess their impact on water quality before, during, and after the 2015 Ranau earthquake. This is done by computing the mean value for each water quality parameter for the whole year under normal conditions, as well as for a set period before and after the June 2015 Ranau earthquake. The data is tabulated and projected onto a graph to look for any patterns, and it can be seen that some parameters, such as Aluminum, Color, Dissolved Oxygen, Iron, Manganese, Nitrate, and Turbidity, showed clear patterns. Data from the aforementioned parameters were then fitted to any seismic activities on relevant dates and tested using mathematical and computational methods to predict an event, in this case, an earthquake. By making time-based modifications to the model inputs, which are the water quality parameters that show promising patterns after screening, a mathematical model is utilized to forecast earthquakes. To accurately calculate a dynamic system reaction to water quality data, a NonLinear AutoRegresive with the eXogenous model is first identified (NLARX). All parameters must fulfill at least 89% of the best-fit data for modeling and validation. As a result, the formulated model based on all factors can be utilized as an early warning system for earthquake prediction in the future with high confidence, limited to the parameters and the area.