Avaliação dos índices NDVI, SAVI e EVI na vegetação do Parque do Cocó (Fortaleza, CE) via imagens Landsat

Abstract

Vegetation cover plays a fundamental role in maintaining environmental balance in urban areas, making its monitoring essential for environmental planning and management. In this context, this study aimed to evaluate the performance of vegetation indices in the Cocó State Park through the quantitative analysis of spectral indices and the visual interpretation of satellite images. Landsat 5 and Landsat 8 images acquired in 1985, 1995, 2005, 2015, and 2025 were processed using the Google Earth Engine platform to calculate the NDVI, SAVI, and EVI vegetation indices. The generated products were analyzed through thematic maps and descriptive statistics, allowing the assessment of temporal changes in vegetation cover and the comparison of the behavior of the three indices. The results indicated an overall increase in vegetation cover throughout the analyzed period, although a reduction was observed in 2005, possibly associated with the rainfall deficit recorded that year. The three vegetation indices showed similar temporal trends but differed in their representation of vegetation cover. NDVI presented higher values and a tendency to overestimate dense vegetation, whereas EVI showed a greater capacity to differentiate vegetation cover based on the visual interpretation of the generated cartographic products. However, this finding is limited to the visual assessment conducted in this study, as no field validation was performed. The results suggest that the establishment of Cocó State Park as a protected area may have contributed to the conservation of its vegetation cover over the last decades, while also demonstrating that each vegetation index has specific strengths and limitations that should be considered when interpreting the results.