| Abstrakt: |
This study aims to monitor the implications of climate change on savanna ecosystem drought using time series data from the Landsat 8 sensor, spanning from 2013 to 2022. We employed a remote sensing computational approach with the semi-automatic classification plugin (SCP) in the open-source QGIS software. Specifically, we utilized channels from the operational land imager (OLI), including Band 4 Red (0.636-0.673 µm) and Band 5 Near-Infrared (0.851-0.879 µm), as well as Thermal Infrared Sensor (TIRS) channels Band 10 TIRS-1 (10.60-11.19 µm) and Band 11 TIRS-2 (11.50-12.51 µm). These channels were used to calculate the vegetation health index (VHI) using the raster calculator, followed by data reclassification with specific thresholds to compare drought-affected areas. Our findings reveal a significant impact of climate change on savanna ecosystem drought over the decade, with the most extreme conditions observed in 2015 and 2019, where drought coverage reached 42.74% and 26.58%, respectively. Other years exhibited relatively low drought dynamics, affecting less than 3% of the area. This period aligns with the el niño-southern oscillation (ENSO) phenomenon, particularly the transition from El Niño to La Niña, known to cause global weather variations, and significantly influenced by the positive phase of the Indian Ocean dipole (IOD). The novelty of this research lies in two main aspects: firstly, the use of Landsat satellite sensors for this specific region has not been extensively studied before; secondly, the discovered impacts of drought in relation to global climate change phenomena are particularly noteworthy. A limitation of this study is the relatively short investigation period of just one decade, which does not fully capture the long-term impacts of climate change. Future research is recommended to utilize imagery with higher temporal resolution over extended periods to better represent extreme climate events and derive drought patterns over durations exceeding one decade. [ABSTRACT FROM AUTHOR] |
