Andini, Syifa Fauziyah (2026) Pemodelan Emisi Karbon Dioksida (CO₂) Di Indonesia Menggunakan Spatially Constrained Hierarchical Clustering (SCHC) Dan Multivariate Adaptive Regression Splines (MARS). Other thesis, Institut Teknologi Sepuluh Nopember.
![[thumbnail of 5003221178-Undergraduate-Thesis.pdf]](http://repository.its.ac.id/style/images/fileicons/text.png) |
Text
5003221178-Undergraduate-Thesis.pdf - Accepted Version Restricted to Repository staff only Download (3MB) | Request a copy |
Abstract
Dalam beberapa dekade terakhir, perubahan iklim telah menjadi tantangan yang dihadapi hampir seluruh negara. Salah satu faktor yang mendasari kondisi tersebut adalah meningkatnya emisi karbon dioksida (CO₂), sebagai komponen utama gas rumah kaca. Di Indonesia, tingkat dan struktur emisi CO₂ menunjukkan variasi antarwilayah akibat keberagaman karakteristik sektor energi, kehutanan, dan proses industri. Keberagaman ini menegaskan bahwa pendekatan kebijakan yang seragam tidak lagi memadai dalam mengakomodasi perbedaan karakteristik antarwilayah. Penelitian ini bertujuan menganalisis karakteristik emisi CO₂ sektoral, melakukan segmentasi wilayah berbasis keterhubungan spasial menggunakan Spatially Constrained Hierarchical Clustering (SCHC), serta memodelkan hubungan antara emisi CO₂ dan variabel prediktor menggunakan Multivariate Adaptive Regression Splines (MARS). Data yang digunakan berupa data panel 34 provinsi di Indonesia periode 2015–2024 yang bersumber dari Badan Pusat Statistik, Kementerian Lingkungan Hidup dan Kehutanan, serta Kementerian Energi dan Sumber Daya Mineral. Hasil penelitian menunjukkan bahwa sektor energi mendominasi emisi CO₂ nasional dengan kontribusi sebesar 68%, diikuti sektor kehutanan sebesar 27% dan IPPU sebesar 5%. Penerapan SCHC menghasilkan sembilan klaster wilayah yang mampu merepresentasika heterogenitas emisi CO₂ Indonesia secara spasial. Klaster 1 terdiri atas 15 provinsi di kawasan timur Indonesia dan Bali, Klaster 2 mencakup 11 provinsi di Sumatera dan Jawa, Klaster 3 terdiri atas Jawa Tengah dan Jawa Timur, sementara Kalimantan Tengah, Sumatera Selatan, Riau, Sumatera Utara, Kalimantan Timur, dan Jawa Barat masing-masing membentuk klaster tunggal karena memiliki karakteristik emisi yang sangat khas. Selanjutnya, pemodelan MARS menunjukkan bahwa faktor penentu emisi CO₂ bervariasi antarklaster. Secara umum, jumlah penduduk dan jumlah kendaraan bermotor merupakan prediktor yang paling konsisten muncul pada model emisi CO₂ di Indonesia. Hasil penelitian menunjukkan bahwa kombinasi metode SCHC dan MARS berhasil mengidentifikasi tipologi wilayah serta prediktor emisi CO₂ pada setiap klaster, sehingga dapat mendukung penyusunan strategi mitigasi yang lebih terarah sesuai karakteristik spesifik wilayah, guna mendukung pencapaian target net zero emissions Indonesia.
=====================================================================================================================================
Over the past few decades, climate change has become one of the most pressing global challenges, driven largely by increasing carbon dioxide (CO₂) emissions, the primary contributor to greenhouse gases. In Indonesia, the magnitude and composition of CO₂ emissions vary substantially across provinces due to differences in the energy, forestry, and industrial processes and product use (IPPU) sectors. This spatial heterogeneity highlights the need for region-specific mitigation strategies rather than uniform policy interventions. This study aims to analyze the characteristics of sectoral CO₂ emissions, identify spatially coherent regional typologies using Spatially Constrained Hierarchical Clustering (SCHC), and model the relationship between CO₂ emissions and their driving factors using Multivariate Adaptive Regression Splines (MARS). The analysis is based on panel data from 34 Indonesian provinces covering the period 2015–2024. The results indicate that the energy sector is the largest contributor to national CO₂ emissions (68%), followed by the forestry sector (27%) and the IPPU sector (5%). The SCHC method identifies nine spatially constrained clusters that capture the heterogeneity of CO₂ emissions across Indonesia. Cluster 1 consists of 15 provinces in Eastern Indonesia and Bali, Cluster 2 includes 11 provinces in Sumatra and Java, and Cluster 3 comprises Central Java and East Java. Meanwhile, Central Kalimantan, South Sumatra, Riau, North Sumatra, East Kalimantan, and West Java each form a singleton cluster due to their distinctive emission profiles. The MARS models further reveal that the determinants of CO₂ emissions differ across clusters, with population size and the number of motor vehicles emerging as the most consistent predictors. Overall, the integration of SCHC and MARS successfully identifies regional emission typologies and cluster-specific drivers, providing valuable insights for developing targeted mitigation strategies to support Indonesia's net zero emissions target.
| Item Type: | Thesis (Other) |
|---|---|
| Uncontrolled Keywords: | Emisi Karbon Dioksida, Heterogenitas Spasial, Spatially Constrained Hierarchical Clustering, Multivariate Adaptive Regression Splines, Net Zero Emission Carbon Dioxide Emissions, Spatial Heterogeneity, Spatially Constrained Hierarchical Clustering, Multivariate Adaptive Regression Splines, Net Zero Emission |
| Subjects: | G Geography. Anthropology. Recreation > GE Environmental Sciences > GE300 Environmental management H Social Sciences > HA Statistics > HA30.6 Spatial analysis Q Science > QA Mathematics > QA278 Cluster Analysis. Multivariate analysis. Correspondence analysis (Statistics) |
| Divisions: | Faculty of Science and Data Analytics (SCIENTICS) > Statistics > 49201-(S1) Undergraduate Thesis |
| Depositing User: | Syifa Fauziyah Andini |
| Date Deposited: | 31 Jul 2026 06:48 |
| Last Modified: | 31 Jul 2026 06:48 |
| URI: | http://repository.its.ac.id/id/eprint/140783 |
Actions (login required)
 |
View Item |