Model Probabilistik Pengaruh Indikator Agroklimatologi Terhadap Produksi Padi di Pulau Jawa

Maharani, Anggita (2025) Model Probabilistik Pengaruh Indikator Agroklimatologi Terhadap Produksi Padi di Pulau Jawa. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 2043221018-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (2MB) | Request a copy

Abstract

Ketahanan pangan yang stabil merupakan fondasi penting bagi kestabilan sosial dan ekonomi. Padi sebagai sumber utama konsumsi masyarakat Indonesia memiliki peran strategis, terutama di Pulau Jawa yang menyumbang lebih dari 50% produksi nasional. Namun, sejak 2018 produksi padi di wilayah ini menunjukkan tren penurunan, dipicu oleh ketidakpastian iklim yang memengaruhi kondisi tanah dan lingkungan. Permasalahan tersebut dapat dianalisis melalui pendekatan agroklimatologi yang mempelajari hubungan antara faktor cuaca dan sistem pertanian. Faktor yang dikaji mencakup curah hujan, suhu udara, radiasi matahari, kelembapan udara, tutupan awan, kecepatan angin, kelembapan permukaan tanah, dan kelembapan zona akar. Penelitian ini bertujuan memodelkan hubungan kausal antara faktor-faktor agroklimatologi dan produksi padi di Pulau Jawa menggunakan pendekatan Copula Bayesian Network (CBN). Metode ini dipilih karena mampu menangani ketergantungan nonlinear dan tidak terikat pada asumsi distribusi tertentu. Data yang digunakan berupa data bulanan tahun 2018–2024. Hasil penelitian menunjukkan bahwa faktor agroklimatologi yang berpengaruh signifikan berbeda antarprovinsi, dengan ketergantungan yang didominasi Copula Archimedean, khususnya Clayton dan Gumbel. Curah hujan dan kelembapan zona akar muncul sebagai variabel yang paling konsisten berasosiasi signifikan dengan produksi padi. Model CBN memberi performa prediksi titik yang terbatas karena sifatnya yang probabilistik, namun memberikan cakupan ketidakpastian (PICP) yang cukup baik sebesar lebih dari 60% Jawa Timur, Jawa Barat, Banten, dan DKI Jakarta. Kelembapan zona akar memiliki sensitivitas positif paling konsisten di seluruh provinsi, sedangkan curah hujan dan kecepatan angin cenderung menunjukkan sensitivitas negatif. Hal ini menunjukkan perlunya strategi adaptasi sesuai kondisi agroklimatologi masing-masing provinsi di Pulau Jawa.
======================================================================================================================================
Stable food security is an important foundation for social and economic stability. Rice, as the main source of consumption for the Indonesian people, plays a strategic role, especially on Java, which contributes over 50% of national production. However, since 2018, rice production in this region has shown a downward trend due to climate uncertainty affecting soil and environmental conditions. This issue can be analyzed through an agroclimatology approach, which studies the relationship between weather factors and agricultural systems, including rainfall, air temperature, solar radiation, air humidity, cloud cover, wind speed, soil surface moisture, and root zone moisture. This study aims to model the causal relationship between agroclimatological factors and rice production in Java using the Copula Bayesian Network (CBN) approach. This method was chosen because it handles nonlinear dependencies and is not bound to specific distribution assumptions. Monthly data from 2018 to 2024 were used. Results showed that significant agroclimatological factors differed between provinces, with dependencies dominated by Archimedean Copulas, particularly Clayton and Gumbel. Rainfall and root zone moisture were the most consistently associated variables with rice production. The CBN model provided limited point prediction performance due to its probabilistic nature but achieved a PICP over 60% in East Java, West Java, Banten, and DKI Jakarta. Root zone moisture showed the most consistent positive sensitivity, while rainfall and wind speed tended to have negative sensitivity. These findings highlight that rice production dynamics in Java strongly depend on adaptation strategies aligned with local agroclimatological conditions of Java.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Agroklimatologi, Copula Bayesian Network, Produksi Padi, Agroclimatology, Rice Production
Subjects:	H Social Sciences > HA Statistics > HA31.3 Regression. Correlation. Logistic regression analysis.
Divisions:	Faculty of Vocational > 49501-Business Statistics
Depositing User:	Anggita Maharani
Date Deposited:	15 Jun 2026 06:10
Last Modified:	15 Jun 2026 06:10
URI:	http://repository.its.ac.id/id/eprint/129134

Actions (login required)

View Item