Influência das métricas de complexidade baseadas em entropia sobre índice de vegetação

Abstract

Entropy-based complexity metrics are utilized for classifying landscapes into areas of high variability, edge-effect regions, and core areas of a specific land use class. Correlating this data with a vegetation index, such as the Normalized Difference Vegetation Index (NDVI), can provide insights into plant health in crops and the quality of agricultural practices adopted. This study aimed to understand the relationship between NDVI and entropy-based metrics He/Hmax, SDL, and LMC. The research was conducted in the cultivable area of an experimental property situated in the municipality of Céu Azul, in the western region of Paraná, known for crop rotation practices from May to August 2022. Satellite images from Sentinel-2 with a spatial resolution of 10 meters, dated in July 2022, were utilized, corresponding to a corn crop according to the agricultural calendar of DERAL/SEAB. By employing spatial data analysis (AEDE) and the Moran's I index for autocorrelation patterns, the study revealed significant interactions at the 1% probability level between the variables. NDVI exhibited a positive correlation with the He/Hmax metric, suggesting that an increase in variability indicates denser and potentially more productive vegetation. In contrast, SDL and LMC had a negative impact on NDVI, indicating regions that may be less productive. The Spatial AutoRegressive (SAR) model generated demonstrated that NDVI is positively influenced by He/Hmax, with a root mean square error of 0.0171 for the red and near-infrared bands.