Multi-Sensor NDVI Analysis of Eruption Impacts and Vegetation Recovery at Mount Kelud, East Java (2013–2025)

Abstract

This study examines the effects of the February 2014 volcanic eruption of Mount Kelud, East Java, Indonesia, on vegetation dynamics and the subsequent recovery process using multi-temporal satellite-derived Normalized Difference Vegetation Index (NDVI) data. Although NDVI has been extensively used in the past to analyze post-disaster conditions, there is very little long-term, multi-sensor analysis in tropical volcanic regions. In this research, there is a combination of Landsat 8 OLI and Sentinel-2 MSI images to assess spatial and temporal vegetation changes from 2013 to 2025. Findings reveal that the NDVI decreased drastically (by more than 0.50 units) shortly after the eruption and further confirm that a significant amount of vegetation was lost due to ashfall, pyroclastic flows, and lahars. However, the vegetation recovery was quick, as the NDVI values returned to the pre-eruption values after one year and surpassed the baseline values in subsequent years (0.80-0.87). Damage classification indicated 89% of sampled sites were low-severity disturbed, and there was spatial heterogeneity in eruption effects. The statistical analysis indicated that there was no significant seasonal effect on the variability of NDVI (r = −0.095, p > 0.05), which suggested that vegetation dynamics are more influenced by disturbance as opposed to climatic seasonality. The results indicate that multi-sensor NDVI analysis is effective in tracking recovery after an eruption and gives an understanding of the resilience of the tropical volcanic ecosystems.