- National Institute of Meteorological Sciences, Forecast Research Department, Korea, Republic of (ddengkkang@gmail.com)
Due to the increase in extreme weather events, climate patterns are emerging that differ from those of the past, and there is a rising need for effective disaster prevention measures to proactively respond to such hazardous weather conditions. In particular, while some areas are currently managed as a single weather alert zone including coastal and insular regions, the need for further subdivision has arisen because meteorological characteristics (such as precipitation and snowfall) vary even within a single weather alert zone. Accordingly, this study conducted a research project on the subdivision of weather alert zones, focusing on six places in the Western Coastal region where the need for subdivision has been raised, while taking into account the meteorological characteristics of each location. This study utilized observation data from the Korea Meteorological Administration’s AWS and ASOS systems from 2018 to 2025 to quantify the occurrence patterns of weather alerts at each station. Based on this, the study investigated methods to classify regions based on similar weather characteristics from those with dissimilar characteristics. To this end, binary matrix, Jaccard-based clustering, hierarchical cluster analysis, and Dynamic Time Warping (DTW) analysis were employed. We evaluated the simultaneity (correlation) among observation stations within the same special weather alert area to verify spatial heterogeneity, and then estimated the optimal number of clusters. In addition, we verified whether the observed clustering differences were attributable to time-lag effect and confirmed whether the regions exhibiting distinct clustering characteristics were indeed validly separated. As a result, an analysis of six weather advisory zones and 30 observation points along the western coastal regions revealed that the characteristics of hazardous weather occurrence differed among some points, suggesting the potential for spatial subdivision within weather alert zones.
How to cite: Shin, M. and Kim, J.: A Study on the Method of Subdivision Weather Alert Areas including Coastal and Insular regions, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-33, https://doi.org/10.5194/ems2026-33, 2026.