Spectral-Temporal Analysis Of Wetland Fires In The Pantanal-Brazil Supported By UAV Multispectral Imagery

The Wetlands of the Pantanal in Brazil are highly sensitive to environmental changes, and events such as wildfires pose a significant threat to biodiversity. In 2020, approximately 80% of its area was affected by high-intensity fires. Therefore, this study aimed to analyze, both spectrally and temporally over a three-year period (2019, 2020, and 2021), the behavior of four macrohabitats located within two study areas of the Private Reserve of Natural Heritage SESC Pantanal (RPPN SESC Pantanal), situated in the state of Mato Grosso, Brazil. For the analysis conducted over the three-year period in the study areas, the Micasense Altum multispectral camera was employed, along with processing methods involving spectral and temporal analysis. The results revealed a drastic decrease in reflectance within the red-edge and near-infrared (NIR) spectral bands in 2020, following the fire event, in both mapped areas. A subsequent recovery was observed in 2021, although reflectance levels remained below those recorded in 2019 (pre-fire conditions). The Acurizal and Tabocal macrohabitats exhibited the highest reflectance amplitudes and the greatest variability over the years, particularly in longer wavelengths (NIR). The Campina macrohabitat showed the lowest reflectance values, due to its vegetation being composed predominantly of shrub and herbaceous species. The Dry Forest (Mata Seca) displayed the highest spectral stability and demonstrated a continuous downward trend in average reflectance, indicating a loss of species diversity following the fire event. The findings contribute to the enhancement of conservation measures in wetland ecosystems, the management of protected areas, and the effectiveness of public policies, highlighting the potential of high-resolution multispectral data for spectral monitoring as a tool for detecting environmental changes.