The Influence of Aerosols on Forest Vegetation Cover in the South-West of Azerbaijan

Abstract

Atmospheric pollution by anthropogenic emissions leads to an increase in the content of aerosols in the air. This causes distortions in the Earth remote sensing data in the visible ranges. Thus, the use of vegetation indices, in particular, the normalized differential vegetation index (NDVI), is difficult. In this regard, interest in the use of satellite images obtained in the short-wave infrared bands SWIR1 and SWIR2 has recently increased. Unlike NDVI, when calculating the AFRI-1600 and AFRI-2100 indices, the values of these bands are taken into account (instead of visible red). Therefore, these indices are called aerosol-free. The article addresses the issues of the NDVI and AFRI indices. The studies have been carried out in three regions located in the south-west of Azerbaijan. The initial data has been taken from 2000 and 2021 images obtained using the Landsat-5 and Landsat-8 satellites, respectively. The scanners of these satellites have ranges of 1600 and 2100 μm. The study has taken place in several stages. At the first stage, the AFRI-1600 indices have been calculated and the areas with high aerosol content for the indicated years have been identified. By studying the dynamics of the amount of aerosols, two types of sites have been identified. Type 1 sites are those where the aerosol content was high in 2000, but dropped in 2021. In type 2 sites, a high value of this indicator was not previously recorded, but was noted in 2021. The NDVI index value has been obtained for both categories, and the classification of the site coverage has been carried out. The results of superimposing index maps on each other have shown that an increase in aerosol content corresponds to a decrease in vegetation density, and vice versa, a decrease in aerosols corresponds to an increase in this indicator. Thus, the increased aerosol content has a negative effect on the condition of forest cover. This is confirmed by the VCI (vegetation condition index) maps.