<![CDATA[East Nusa Tenggara (NTT) Province faces complex multi-hazard threats from hydrometeorological and geological disasters. This study developed a multi-hazard risk typology for 22 regencies/cities in NTT and examined temporal trends in disaster frequency over 2008–2025 (complete years; partial 2026 data excluded). The dataset comprised 1,158 records, adjusted to 1,128.5 real events using fractional weights across eight disaster types. Clustering was evaluated across three feature schemes and two algorithms (K-Means and Agglomerative Ward), with the number of clusters validated through a consensus of Silhouette, Davies-Bouldin, and Calinski-Harabasz indices while avoiding degenerate solutions. Temporal trends and change points were evaluated using the Mann-Kendall and Pettitt tests, preceded by lag-1 autocorrelation (ACF) screening. The selected hazard-only Agglomerative Ward model (k = 6; Silhouette = 0.2105, DBI = 1.205, CHI = 5.28) categorized regions into six distinct typologies, from low-risk to high-frequency flood/extreme-weather clusters, with Sabu Raijua identified as an impact-severity outlier. Auto correlation checks confirmed the validity of test assumptions. The Mann-Kendall and Pettitt tests consistently revealed no statistically significant monotonic trend or change point (p > 0.05) across the province, dominant disaster types, and all clusters. This meaningful null result demonstrates that disaster occurrences in NTT are driven by sharp inter-annual fluctuations rather than long-term trends, emphasizing the necessity of continuous preparedness and differentiated, typology-based mitigation strategies.]]>
Copyrights © 2026
