Analisis Korelasi Land Surface Temperature (LST) Multitemporal dengan Aktivitas Vulkanik Gunung Agung Menggunakan Citra Landsat-8

Fatma Rizqiamurti Arina, Ir. Ruli Andaru, S.T., M.Eng., Ph.D., IPU.

2026 | Skripsi | TEKNIK GEODESI

Suhu permukaan merupakan salah satu parameter penting dalam pemantauan aktivitas vulkanik. Aktivitas vulkanik Gunung Agung dapat dipantau dari dinamika Land Surface Temperature (LST) menggunakan teknologi penginderaan jauh. Namun, respons spasial dan temporal LST di zona kawah maupun lereng belum terpetakan secara pasti dalam merespon tiap tahapan erupsi. Selain itu, keterbatasan sensor satelit seperti tutupan awan dan perbedaan resolusi perlu dievaluasi keselarasan polanya dengan status resmi PVMBG serta LST MODIS. Penelitian ini bertujuan memetakan sebaran LST, menganalisis anomali pada zona kawah dan lereng, serta membandingkan pola LST Landsat-8 dengan status PVMBG dan LST MODIS di Gunung Agung periode 2016-2022.

Data yang digunakan berupa citra Landsat-8 OLI/TIRS Level-2 Collection 2 Tier 1 yang dianalisis dalam lima fase aktivitas vulkanik, yaitu Fase 1 Sebelum Erupsi (1 Juni 2016 – 1 Juni 2017 ; Level I/Normal) sebagai kondisi baseline, Fase 2 Sebelum Erupsi (22 September 2017 – 30 Oktober 2017 ; Level IV/Awas), Fase 3 Saat Erupsi (29 Mei 2018 – 16 Juni 2019 ; Level III/Siaga), Fase 4 Setelah Erupsi (16 Juli 2020 – 12 September 2021 ; Level II/Waspada), dan Fase 5 Setelah Erupsi (12 September 2021 – 12 September 2022 ; Level I/Normal). Ekstraksi LST dilakukan menggunakan algoritma Statistical Mono-Window (SMW) pada platform Google Earth Engine dengan emisivitas permukaan berbasis NDVI. Anomali LST ditentukan menggunakan threshold Mean + 2? berdasarkan baseline kondisi Normal pada masing-masing zona.

Hasil penelitian menunjukkan lonjakan rata-rata LST terbesar di zona lereng terjadi dari Fase 1 ke Fase 2 sebesar 10,31ºC, sedangkan zona kawah mengalami lonjakan terbesar dari Fase 2 ke Fase 3 sebesar 14,79ºC. Selisih LST antara kondisi Normal sebelum dan sesudah periode aktivitas sebesar 0,72ºC, disertai perluasan area bersuhu “sangat tinggi” sebesar 27,52 ha. Anomali LST tersebar pada kelas “sedang” hingga “sangat tinggi” di zona lereng dan “tinggi” hingga “sangat tinggi” di zona kawah. Peningkatan LST dan perluasan anomali mencapai puncaknya pada Fase 2 dan 3, kemudian menurun secara konsisten pada Fase 4 dan 5 menuju kondisi baseline seiring penurunan status PVMBG. Perbandingan dengan LST MODIS menunjukkan pola yang relatif searah meski terdapat perbedaan nilai absolut akibat faktor resolusi spasial dan karakteristik sensor. Dengan demikian, LST Landsat-8 terbukti efektif dan berpotensi sebagai indikator pendukung pemantauan aktivitas vulkanik.

Land surface temperature (LST) is an important parameter for monitoring volcanic activity. Volcanic activity at Mount Agung can be monitored through the dynamics of LST using remote sensing technology. However, the spatial and temporal responses of LST in the crater and slope zones to each stage of the eruption have not been clearly mapped. In addition, satellite sensor limitations, such as cloud cover and differences in spatial resolution, mean that the consistency of Landsat-8 LST patterns with the official alert levels issued by the Center for Volcanology and Geological Hazard Mitigation (PVMBG) and with MODIS LST needs to be evaluated. This study aims to map the distribution of LST, analyze LST anomalies in the crater and slope zones, and compare Landsat-8 LST patterns with PVMBG alert levels and MODIS LST at Mount Agung during 2016–2022.
The data used were Landsat-8 OLI/TIRS Level-2 Collection 2 Tier 1 imagery, analyzed across five phases of volcanic activity: Phase 1, pre-eruption (June 1, 2016 – June 1, 2017 ; Level I/Normal), used as the baseline; Phase 2, pre-eruption (September 22, 2017 – October 30, 2017 ; Level IV/Awas); Phase 3, during the eruption (May 29, 2018 – June 16, 2019 ; Level III/Siaga); Phase 4, post-eruption (July 16, 2020 – September 12, 2021 ; Level II/Waspada); and Phase 5, post-eruption (September 12, 2021 – September 12, 2022 ; Level I/Normal). LST was retrieved using the Statistical Mono-Window (SMW) algorithm on the Google Earth Engine platform, with NDVI-based surface emissivity. LST anomalies were identified using a threshold of Mean + 2?, derived from the Normal baseline condition of each zone.
The result show that the largest increase in mean LST in the slope zone occurred from Phase 1 to Phase 2 (10,31ºC), while the crater zone experienced its largest increase from Phase 2 to Phase 3 (14,79ºC). The LST difference between the Normal conditions before and after the period of activity was 0,72ºC, accompanied by a 27,52 ha expansion of the “very high” temperature class. LST anomalies were distributed across the “moderate” to “very high” classes in the slope zone and the “high” to “very high” classes in the crater zone. The LST increase and anomaly expansion peaked in Phases 2 and 3, then declined consistently in Phases 4 and 5 toward baseline levels as the PVMBG alert level was lowered. Comparison with MODIS LST showed a relatively consistent trend, despite differences in absolute values caused by spatial resolution and sensor characteristics. Therefore, Landsat-8 LST was shown to be effective and has potential as a supporting indicator for volcanic activity monitoring.

Kata Kunci : Gunung Agung, Land Surface Temperature, Landsat-8, Aktivitas Vulkanik, Mount Agung, Land Surface Temperature, Landsat-8, Volcanic Activity