Electronic Theses and Dissertation

Dua dekade pasca Tsunami Samudra Hindia 2004, tantangan utama Kota Banda Aceh tidak lagi berfokus pada rekonstruksi fisik, melainkan pada keberlanjutan resiliensi masyarakat dalam jangka panjang di tengah peluruhan memori risiko (risk memory decay) dan keterbatasan ekonomi struktural. Pendekatan analisis statistik konvensional yang bersifat statis dinilai kurang mampu menangkap dinamika perubahan perilaku adaptif masyarakat yang berlangsung secara temporal dan spasial. Penelitian ini bertujuan untuk memodelkan dan menganalisis dinamika resiliensi jangka panjang (20 tahun) serta intensi migrasi masyarakat di Kecamatan Baiturrahman menggunakan pendekatan Agent-Based Model (ABM). Model dikembangkan berdasarkan data empiris dari 100 responden yang ditransformasikan menjadi agen heterogen dengan atribut sosial, ekonomi, dan psikologis, serta disimulasikan dalam lingkungan komputasional berbasis Python. Seluruh proses pengembangan model, pelaksanaan simulasi, dan analisis hasil dilakukan menggunakan platform Google Colaboratory yang mendukung komputasi berbasis cloud dan reprodusibilitas analisis. Simulasi dijalankan melalui lima skenario kebijakan, yaitu Business As Usual (BAU), Intervensi Edukasi, Intervensi Ekonomi (subsidi), Intervensi Fisik (penataan ruang dan infrastruktur evakuasi), serta Intervensi Gabungan (policy mix). Hasil simulasi menunjukkan bahwa skenario BAU menghasilkan penurunan resiliensi secara gradual akibat erosi memori risiko, yang berimplikasi pada meningkatnya intensi migrasi dan kecenderungan hollowed-out community. Analisis perbandingan antar skenario mengungkap bahwa intervensi edukasi secara tunggal berpotensi memicu fenomena poverty trap ketika peningkatan pengetahuan tidak diimbangi oleh kapasitas ekonomi kelompok berpendapatan rendah. Sebaliknya, intervensi ekonomi efektif dalam meningkatkan kapasitas adaptif, sementara intervensi fisik memperkuat rasa aman (place attachment) dan menekan laju depopulasi melalui peningkatan akses terhadap infrastruktur evakuasi. Secara keseluruhan, skenario intervensi gabungan menunjukkan kinerja paling optimal dalam menjaga stabilitas resiliensi melalui sinergi antar instrumen kebijakan. Temuan penelitian ini memberikan implikasi penting bagi perencanaan wilayah, khususnya perlunya penerapan pendekatan penataan ruang berbasis risiko yang terintegrasi dengan kebijakan sosial-ekonomi serta pengelolaan memori bencana secara berkelanjutan guna menjamin keberlanjutan permukiman di wilayah rawan bencana, seperti Kecamatan Baiturrahman.
Kata Kunci: Agent-Based Model, Resiliensi Komunitas, Memory Decay, Jebakan Kemiskinan, Intensi Migrasi, Baiturrahman

Two decades after the 2004 Indian Ocean Tsunami, the main challenge facing Banda Aceh City is no longer physical reconstruction, but the long-term sustainability of community resilience amid the erosion of risk memory (risk memory decay) and structural economic constraints. Conventional static statistical approaches are considered insufficient to capture the dynamic changes in adaptive community behavior across temporal and spatial dimensions. This study aims to model and analyze long-term community resilience dynamics (20 years) and migration intentions in Baiturrahman District using an Agent-Based Model (ABM) approach. The model was developed based on empirical data from 100 respondents, which were transformed into heterogeneous agents with social, economic, and psychological attributes and simulated within a Python-based computational environment. The entire process of model development, simulation execution, and result analysis was conducted using Google Colaboratory, which supports cloud-based computation and analytical reproducibility. The simulations were implemented under five policy scenarios: Business As Usual (BAU), Education Intervention, Economic Intervention (subsidies), Physical Intervention (spatial planning and evacuation infrastructure), and Combined Intervention (policy mix). Simulation results indicate that the BAU scenario leads to a gradual decline in community resilience due to the erosion of risk memory, resulting in increased migration intentions and a tendency toward hollowed-out communities. Comparative analysis across scenarios reveals that education-only interventions may trigger a poverty trap when increased knowledge is not accompanied by adequate economic capacity among low-income groups. In contrast, economic interventions effectively enhance adaptive capacity, while physical interventions strengthen the sense of security (place attachment) and reduce depopulation through improved access to evacuation infrastructure. Overall, the combined intervention scenario demonstrates the most optimal performance in maintaining resilience stability through policy synergy. These findings provide important implications for spatial planning, particularly the need to implement a risk-based spatial planning approach integrated with socio-economic policies and the sustainable management of disaster memory to ensure the long-term sustainability of settlements in disaster-prone areas, such as Baiturrahman District. Keywords: Agent-Based Modeling, Community Resilience, Memory Decay, Poverty Trap, Migration Intention, Baiturrahman.