Pemodelan Curah Hujan Dasarian Menggunakan Hidden Markov Model Periode September, Oktober, dan November (SON) di Kabupaten Kulon Progo

Wardani, Adelia Ayunda (2026) Pemodelan Curah Hujan Dasarian Menggunakan Hidden Markov Model Periode September, Oktober, dan November (SON) di Kabupaten Kulon Progo. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5003221096-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (3MB) | Request a copy

Abstract

Perubahan iklim menyebabkan peningkatan variabilitas curah hujan, yang dapat memengaruhi kestabilan sumber daya air dan sektor pertanian, sehingga pemahaman terhadap pola hujan menjadi sangat penting. Penelitian ini menggunakan data curah hujan dasarian (10 hari) dari empat wilayah pengamatan di Kabupaten Kulon Progo selama periode September, Oktober, dan November (SON) tahun 2004 hingga 2023. Pemodelan dilakukan menggunakan Hidden Markov Model (HMM) yang mengaitkan kondisi hujan laten (hidden staes) dengan distribusi yang sesuai, diestimasi melalui metode Expectation-Maximization (EM). Hasil analisis menunjukkan bahwa model distribusi Gamma memberikan representasi terbaik untuk keempat stasiun, dengan jumlah hidden state optimal berbeda sesuai wilayah. Kalibawang, Panjatan, dan Samigaluh menghasilkan model 3 hidden states dengan AIC masing-masing -460,1961; -329,8105; dan 19,2682, sedangkan Gembongan menghasilkan 4 hidden states dengan AIC -126,2409. Untuk tiga stasiun pertama, state 1 merepresentasikan kondisi kering, state 2 kondisi normal, dan state 3 kondisi basah. Sementara itu, di Gembongan, meskipun menghasilkan 4 hidden state, dua state pertama sama-sama merepresentasikan kondisi kering karena data tidak cukup membedakan kedua kondisi tersebut. Lalu state 3 menggambarkan kondisi normal dan state 4 kondisi basah. Pemodelan ini berhasil menangkap dinamika temporal curah hujan pada periode SON, sehingga dapat menjadi dasar bagi perencanaan pengelolaan sumber daya air, mitigasi risiko bencana hidrometeorologi, dan mendukung sektor pertanian dalam menghadapi tantangan perubahan iklim.
=====================================================================================================================================
Climate change causes increased rainfall variability, which can affect the stability of water resources and the agricultural sector, so understanding rainfall patterns is very important. This study uses decadal (10-day) rainfall data from four observation areas in Kulon Progo Regency during the September, October and November (SON) from 2004 to 2023. Modelling was carried out using a Hidden Markov Model (HMM) that links latent rainfall conditions (hidden states) to appropriate distributions, estimated using the Expectation-Maximization (EM) method. The analysis results indicate that the Gamma distribution model provides the best representation for all four stations, with the optimal number of hidden states varying by region. Kalibawang, Panjatan, and Samigaluh produced a model of 3 hidden states with AIC values of -460.1961; -329.8105; and 19.2682 respectively, while Gembongan produced 4 hidden states with an AIC of -126.2409. For the first three stations, state 1 represents dry conditions, state 2 normal conditions, and state 3 wet conditions. Meanwhile, at Gembongan, although it produced 4 hidden states, the first two states both represent dry conditions because the data was insufficient to distinguish between the two conditions. Then, state 3 describes normal conditions and state 4 wet conditions. This modelling successfully captures the temporal dynamics of rainfall during the SON period, thus providing a basis for water resource management planning, hydrometeorological disaster risk mitigation, and supporting the agricultural sector in facing the challenges of climate change.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Curah Hujan Dasarian, Distribusi Gamma, Expectation-Maximization, Hidden Markov Model, Ten-day Rainfall, Gamma Distribution, Expectation-Maximisation, Hidden Markov Model
Subjects:	Q Science > QA Mathematics > QA274.7 Markov processes--Mathematical models. Q Science > QC Physics > QC866.5 Climatology--Forecasting. Q Science > QC Physics > QC925 Rain and rainfall
Divisions:	Faculty of Science and Data Analytics (SCIENTICS) > Statistics > 49201-(S1) Undergraduate Thesis
Depositing User:	Adelia Ayunda Wardani
Date Deposited:	04 Aug 2026 07:58
Last Modified:	04 Aug 2026 07:58
URI:	http://repository.its.ac.id/id/eprint/143308

Actions (login required)

View Item