Open AccessTemporal analysis of drought coverage in a watershed area using remote sensing spectral indexes
Author(s) -
Débora Joana Dutra,
Marcos Antônio Timbó Elmiro,
Carlos Wagner Gonçalves Andrade Coelho,
Marcelo Antônio Nero,
Plínio da Costa Temba
Publication year - 2021
Publication title -
sociedade and natureza
Language(s) - English
Resource type - Journals
eISSN - 1982-4513
pISSN - 0103-1570
DOI - 10.14393/sn-v33-2021-59505
Subject(s) - normalized difference vegetation index , watershed , vegetation (pathology) , environmental science , enhanced vegetation index , geostatistics , remote sensing , seasonality , hydrology (agriculture) , physical geography , spatial variability , geography , vegetation index , climate change , statistics , geology , computer science , mathematics , medicine , oceanography , geotechnical engineering , pathology , machine learning
The development of several time series analysis programs using satellite images has provided many applications based on resources from geostatistics field. Currently, the use of statistical tests applied to vegetation indexes has enabled the analysis of different natural phenomena, such as drought events in watershed areas. The objective of this article is to provide a comparative analysis between NDVI and EVI vegetation index data made available by MOD13Q1 project of MODIS sensor for drought mapping using vegetation condition index (VCI) in the Serra Azul stream sub-basin, MG. The methodology adopted the Cox-Stuart statistical test for seasonality analysis and Pearson's linear correlation to verify the influence of different indexes on delimitation of drought in a watershed. The results indicated the NDVI vegetation index as more efficient than EVI in spatial characterization of studied watershed region, mainly in identification of seasonality. The VCI proved to be highly feasible for monitoring drought in study period between 2013 and 2018, allowing the effective delimitation of drought conditions in the Serra Azul stream sub-basin. In addition, the effectiveness of MODIS sensor data in characterizing drought events that affected the study area was proven.