Some aspects of chlorophyll concentration assessment in inland water bodies based on satellite data

As is known, the accuracy of results for estimating chlorophyll concentration in turbid closed water bodies based on multispectral imagery depends on a number of factors. These include the spatial non uniformity and non linearity of bio optical interactions in turbid waters. As with the estimation of suspended particles in water, remote estimation of chlorophyll is based on the relationships between brightness and reflectance in narrow bands. ChlA, which primarily reflects green light, absorbs more energy from violet blue and orange red light. For this reason, research primarily covers the interval from the near infrared to the red spectrum.
In the article under review, algorithms for characterizing the region were determined based on a comparison of data obtained from various sensors in different spectral bands, with the aim of reducing their measurement errors. In the study, in addition to satellite images, in situ measurement results of ChlA from the Mingachevir reservoir were also used.
Numerous methods have been proposed for the measurement of ChlA in turbid waters, and they are divided into two main groups analytical and empirical methods. In the analytical method, based on the radiative transfer equation, the physical relationship between the water parameters and its absorption and scattering coefficients at various wavelengths is determined.
In the empirical method, the relationship between the parameters and the data obtained in various spectral bands is determined.
The research was conducted in June, July, and August 2023 at Mingachevir Reservoir, the largest reservoir in Azerbaijan. Its area is 605 square kilometers, and its average depth is 6 m. Satellite imagery from Landsat 8, Landsat 9, and Sentinel 2A was used.
In the studies, two and three band models proposed by Gitenson were used to reduce the effect of suspended solids and CDOM indicators in the water.
Based on the reviewed models, the chlorophyll concentration for turbid waters was determined using the NDCİ index proposed by Mishra.
Research results show that the chlorophyll concentration in the reservoir varies within a wide range from 5 micrograms per liter to 22 micrograms per liter. These values differ sharply across different seasons.
The Root Mean Square Error coefficient was determined for the Root Mean Square Error between the in situ measurement results and the satellite measurement results.
Along with the calculations, a chlorophyll distribution map across the water surface and a turbidity map of the water were prepared for each recorded study month in the reservoir.