Assessing the spatiotemporal variation of phytoplankton biomass in Nandoni reservoir in the Vhembe District (South Africa) using LANDSAT satellite imagery

Abstract

Chlorophyll–a (chl–a) is an optical active compound used as proxy for phytoplankton biomass to determine the trophic state of the aquatic ecosystem. Traditional approaches for monitoring aquatic system are time consuming, expensive, and non–continuous, therefore, Remote Sensing technologies are qualitative for monitoring the status for water quality in large scale and low cost. The aim of this study was to assess the spatial and temporal variation of phytoplankton biomass in Nandoni reservoir, Limpopo to examine the relationship that exist between the physico–chemical variables and chl–a concentration using readily available Landsat multispectral images. Multispectral resolution of (30 m) Landsat 7 ETM+ and Landsat 8 OLI images for June to December 2008 to 2020 were used to derive the distribution of chl–a concentration. The spatial distribution of chl–a concentration in wet and dry season of these years was obtained. By using regression techniques, in situ measured chl–a was related to construct and validate Landsat predicted chl–a to determine the distribution of chl–a in the reservoir. The results indicate that Landsat derived chl–a was similar with the observed measured chlorophyll–a (R2 = 0.91). There was a negative significant correlation among Land Use and Land Cover with water quality (P > 0.05). Using permutational multivariate analysis of variance (PERMANOVA) analysis, there was significant differences for chl–a concentration in sites, seasons, and zones. There was positive significant correlation observed on water temperature with strong negative significant with salinity and TDS. A strong perfect linear association among predicted vs measured chl–a were found. Chlorophyll–a concentration in Nandoni reservoir was derived using Landsat remote sensing images, suggesting that Landsat sensor is suitable for monitoring small reservoir in a short timescale. Remote sensing techniques can be used to control the development of an early warning system of this study and other reservoirs.