Hydroclimatic Regionalization in Indonesia: A Comparative Study of Fuzzy C-Means and Gaussian Mixture Models

Drought is a climate-related natural hazard that poses significant challenges to environmental sustainability and disaster risk reduction in Indonesia, particularly in the context of achieving Sustainable Development Goal 13 (Climate Action). However, drought-related characteristics vary considerably across regions due to differences in local climatic conditions. This study aims to regionalize drought-related hydroclimatic characteristics across 34 provinces in Indonesia using Fuzzy C-Means (FCM) and Gaussian Mixture Models (GMM). The analysis used climatological data from CHIRPS Daily and ERA5-Land covering the 20152025 period. Four indicators were considered, including rainfall, average temperature, number of rainy days, and consecutive dry days (CDD). The optimal number of clusters was determined using the XieBeni Index for FCM and the Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC) for GMM, supported by cluster stability assessment based on the Adjusted Rand Index and additional validation using the Gap Statistic. Both methods identified two clusters as the optimal solution, consistently separating a relatively drier group of provinces from a relatively wetter group. Cluster quality was evaluated using the Silhouette Coefficient and DaviesBouldin Index (DBI). The results showed that FCM provided marginally better clustering performance, with a Silhouette Coefficient of 0.345 and a DBI of 1.076, compared with 0.332 and 1.115, respectively, for GMM, although both methods produced broadly consistent regionalization. The resulting clusters reveal long-term drought-related regional patterns across Indonesia, providing a quantitative basis for identifying provinces with similar hydroclimatic characteristics to support further, more targeted drought-related research.

Authors

Institutions

Publication Details

Journal
CAUCHY Jurnal Matematika Murni dan Aplikasi
Published
2026-11-16
DOI
https://doi.org/10.18860/cauchy.v11i2.45896
Primary Topic
Hydrology and Drought Analysis
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Hydroclimatic Regionalization in Indonesia: A Comparative Study of Fuzzy C-Means and Gaussian Mixture Models

A’yunin Sofro, Bunga Anovia Putri Ramadhianti
2 links
CAUCHY Jurnal Matematika Murni dan Aplikasi
Hydrology and Drought Analysis
article

Hydroclimatic Regionalization in Indonesia: A Comparative Study of Fuzzy C-Means and Gaussian Mixture Models

A’yunin Sofro, Bunga Anovia Putri Ramadhianti
article en

Abstract

Drought is a climate-related natural hazard that poses significant challenges to environmental sustainability and disaster risk reduction in Indonesia, particularly in the context of achieving Sustainable Development Goal 13 (Climate Action). However, drought-related characteristics vary considerably across regions due to differences in local climatic conditions. This study aims to regionalize drought-related hydroclimatic characteristics across 34 provinces in Indonesia using Fuzzy C-Means (FCM) and Gaussian Mixture Models (GMM). The analysis used climatological data from CHIRPS Daily and ERA5-Land covering the 20152025 period. Four indicators were considered, including rainfall, average temperature, number of rainy days, and consecutive dry days (CDD). The optimal number of clusters was determined using the XieBeni Index for FCM and the Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC) for GMM, supported by cluster stability assessment based on the Adjusted Rand Index and additional validation using the Gap Statistic. Both methods identified two clusters as the optimal solution, consistently separating a relatively drier group of provinces from a relatively wetter group. Cluster quality was evaluated using the Silhouette Coefficient and DaviesBouldin Index (DBI). The results showed that FCM provided marginally better clustering performance, with a Silhouette Coefficient of 0.345 and a DBI of 1.076, compared with 0.332 and 1.115, respectively, for GMM, although both methods produced broadly consistent regionalization. The resulting clusters reveal long-term drought-related regional patterns across Indonesia, providing a quantitative basis for identifying provinces with similar hydroclimatic characteristics to support further, more targeted drought-related research.

CAUCHY Jurnal Matematika Murni dan AplikasiVol. 11(2)
Universitas Negeri Surabaya (ID)
Climate action, Clean water and sanitation
Hydrology and Drought Analysis
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon