2.1 Study area
Lake Victoria, spanning an area of 68,800 km2, is shared by three nations (Tanzania 49%, Uganda 45%, and Kenya 6%), with a catchment area of 194,000 square kilometers spread across five countries (Juma et al., 2014). Its climate ranges from tropical rain forest with year-round rainfall (117 km3/year) over the lake to a semi-arid climate with occasional droughts in some parts, and temperatures ranging from 12 - 260C (Miriti, 2022). The LVB experiences rainfall in two distinct seasons with the "long rains" season spanning from March to May (MAM) and the "short rains" occurring in October, November, and early December (OND) (Nicholson, 2015). These are influenced by different large-scale forces such as zonal winds over the central Indian Ocean and inter-tropical convergences (Nicholson, 2017). On the other hand, the driest months tend to be June, July, and August. The soil types in the LVB are diverse and heavily influenced by the Great Rift Valley's volcanic activity whereas montane forests, savannahs, grasslands, wetlands, woodlands, and croplands are among the vegetation types found throughout the basin (Odada et al., 2009). LVB is densely populated, with 300 people per km2, growing by 3.5% annually (Marcus, 2022). Major cities such as Jinja, Kisumu, Mwanza have expanded, alongside new towns on the lake shore (Nyamweya et al., 2020). Fig. 1 shows the LVB, its major tributary rivers and elevation from the Shuttle Radar Topography Mission (SRTM).
2.2 Datasets used in the study
In this study, turbidity and chlorophyll-a were the water quality parameters considered as these can be directly derived from ocean-color satellite remote sensing data. Chlorophyll-a indicates phytoplankton abundance and biomass, reflecting trophic status (Keukelaere & Knaeps, 2021), while turbidity indicates water clarity, affected by factors like river run-off, phytoplankton growth, climate, and watershed changes (Crétaux et al., 2020). Satellites like MODIS, MERIS, Sentinel-2, and Landsat enable accurate analysis of WQ parameters through the connection established between in-situ measurements and emitted/reflected radiation in spectral bands such as the green and infrared bands (Watanabe et al., 2018; Papenfus et al., 2020; Ambrose-Igho et al., 2021). Chl-a and TUR are derived from Lake Water-Leaving Reflectance (LWLR); an important indicator of biogeochemical processes and habitats in the water column (Crétaux et al., 2020), using globally validated algorithms (Dogliotti et al., 2015; Keukelaere & Knaeps, 2021).
Two RS WQ products were used i.e. (1) “ESA” data from the Lakes Project of the European Space Agency Climate Change Initiative (ESA CCI-Lakes) and (2) “VITO” data which is a Lake WQ product from Copernicus Global Land Service (CGLS). The VITO data comprised of monthly turbidity and trophic state index (TSI) at a spatial resolution of 300m derived from the OLCI sensor on board of Sentinel-3. TSI measures phytoplankton productivity and eutrophication. The VITO data were obtained from the Copernicus Global Land Service website (https://land.copernicus.eu/global/products/lwq) for the period of 2016 – 2022. The retrieval algorithms for this dataset are stipulated in Warren et al. (2021). Chl-a was derived from TSI according to the table adapted from (Simis, 2020) shown in Supplementary Material (Table S1). ESA data records of turbidity and chlorophyll-a at a spatial resolution of 100m and daily temporal resolution were acquired from the ESA website (https://climate.esa.int/en/projects/lakes/data/) from 2000 – 2012 (derived from the MERIS sensor on board ESA's ENVISAT satellite) and 2016 – 2019 (derived from the OLCI sensor on board of Sentinel-3). The “Algorithm Theoretical Basis Document” for this data product, readily available on website, provides a full explanation of the algorithms and corrections used to create these estimates. Both datasets were already preprocessed and ready for use. Chlorophyll-a estimates were measured in mgm-3 whereas turbidity in NTU.
Past records of in-situ measurements of Chl-a and TUR data were collected from the National Water and Sewerage Corporation (NWSC), Uganda and these were used to validate RS Chl-a and TUR data. The measurements, available irregularly, were gathered monthly from March 2013 to June 2022 at 26 sampling locations in the Inner Murchison Bay (IMB). Approximately 7% of the data was missing, reflecting occasional gaps in the monthly records.
Monthly precipitation records at 0.05° spatial resolution for the period of 2000 to 2022 were retrieved from the CHIRPS website (https://www.chc.ucsb.edu/data/chirps) and used to analyze changes in rainfall across the LVB over time. Annual land cover maps at a spatial resolution of 300m were obtained from the Land cover dataset from the European Space Agency Climate Change Initiative (ESA-CCI). These were acquired from the website (https://www.esa-landcover-cci.org) for the years 2000, 2005, 2010, 2015 and 2020.
2.3 Validation of WQ RS with in-situ measurements
We carried out an accuracy assessment and validation of the RS water quality data using past in-situ measurements. WQ parameters from the 26 sampling locations in the IMB were averaged at a monthly scale and compared with the ESA and VITO RS data. The sampling locations can be seen in Fig. S1(Supplementary Material). The RS Chl-a and TUR raster files were realigned and resampled to the same pixel size (0.00833o) at a monthly timestep. Statistical metrics, such as mean, median, standard deviation, correlation coefficients, time series and graphical criteria were used to compare the RS and in-situ data. The evaluation aimed at assessing the accuracy and correlation between in-situ and RS data.
2.4 Assessment of water quality in the lake
A visual assessment of maps and time series was also carried out using processed RS data of chlorophyll-a and turbidity to assess the ecological status of the lake from 2005 to 2022. The analysis involved both ESA and VITO data due to the challenges of missing data and the limited timeframe of the satellite data available. Pollution hotspots, notably the Winam Gulf and the IMB, were identified across the lake, prompting a detailed study to investigate the links between land use changes, precipitation patterns, and water quality variations in these regions. The Winam Gulf and IMB consistently exhibited elevated levels of chlorophyll-a and turbidity throughout the study period.
2.5 Spatial and temporal variability of rainfall
This analysis involved the use of monthly mean CHIRPS precipitation raster files spanning from 2000 to 2022. The mean annual precipitation and coefficient of variation (CV) over LVB were computed. The CV was computed as the ratio of the standard deviation to the mean and was used to classify the degree of variability of rainfall events as less (CV < 20), moderate (20 < CV < 30), and high (CV > 30) (Nkwasa et al., 2022).
The Mann-Kendall (MK) (Mann, 1945; Kendall, 1975) test was then applied to the data to identify trends. This test has previously been used to analyze temporal trends of climatic variables such as precipitation and temperature (H. Wang et al., 2012; F. Wang, 2018; Mallick et al., 2021). Since the test is nonparametric, the data does not have to adhere to a normal distribution. However, it does presuppose that there is no autocorrelation in the time series. Typically, trends are considered significant when they achieve a 95% confidence level (Buo et al., 2021). The magnitude of the trends was also calculated using the nonparametric Theil-Sen estimator(Sen, 1968) which is computed by taking the median of the slopes between each pair of points in the time series data.
2.6 Land use/cover change analysis
The nomenclature of the land cover maps was reclassified from the 36 original classes in the LVB to 8 major land classes i.e., agriculture, forest, grassland, wetland, built-up, sparse vegetation, bare area, and open water as shown in Table S2, Supplementary Material (Mousivand & Arsanjani, 2019). This was done to accommodate classes that are relevant to the study area and represent specific land use changes related to ongoing human activities.
To visually depict LULC changes over time, maps were created to highlight areas that experienced growth and those that remained unchanged. Using the GIS vector geoprocessing tool, land cover shapefiles from two time periods (e.g. 2000-2010, 2010-2020 or 2000-2020) were intersected to identify classes that showed no changes in each area. This intersection indicated the absence of change in the land cover class. The intersected land cover classes were then reclassified as "no change" in the resulting map, which represented the "no change" classes and newly gained areas from the recent land cover map e.g. for 2010 for the period of 2000-2010.
2.6.1 LULC change matrices.
Land cover change matrices are used to analyse how different land cover areas have changed over time. This involves comparing maps of the same location from two distinct points in time and generating a cross tabulation matrix. The matrix shows the area that has changed between different land cover categories. Diagonal entries indicate land persistence, while off-diagonal entries indicate land cover change (Aldwaik & Pontius, 2012). Transition matrices have been widely used in landscape ecology and land use/cover change studies (Han et al., 2009; Takada et al., 2010; Romero-Ruiz et al., 2012).
Three levels of analysis exist i.e. interval, categorical, and transition levels. The interval level examines changes between two time periods, the categorical level assesses the intensity of transformation between categories, and the transition level focuses on the dynamics and intensity of transitions within a category relative to others. Annual change intensities are computed at the interval level, while the magnitude and intensity of gross gains and losses are evaluated at the category level. The transition level investigates changes in categories, their variations, and identifies frequently targeted or avoided categories. These analyses compare observed intensities to uniform measures of transition (Alo & Pontius, 2008; Aldwaik & Pontius, 2012). Areal percentage changes for 3 distinct time periods; 2000-2010, 2010-2020 and 2000-2020, were calculated using Equation 1. Transition matrices were then generated using the Semi-Automatic Classification Plugin in QGIS (Congedo, 2016).
![](data:image/png;base64,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)
2.6.2 Change budget and intensity analysis
The analysis of relative gain, loss, and persistence among different land cover classes offers deeper insights into the dynamics of LULC changes. To assess the change budget, transition matrices were employed to calculate the uniform gain, loss, and persistence of various land cover classes using the following equations.
![](data:image/png;base64,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)
To conduct the intensity analysis, the average values of gains and losses were calculated and then employed to distinguish between active and dormant changes. An active category change is identified when the intensity surpasses the uniform line, indicating a relatively rapid change. On the other hand, a dormant change is observed when the intensity falls below the uniform line, indicating a relatively slow change (Huang et al., 2012).
2.7 Linking impacts of changes in climate and LULC to changes in lake water quality
To assess the effects of climate and land use/cover changes on water quality, Chl-a and TUR time series data were extracted from prominent pollution hotspots in Lake Victoria, namely the IMB in Uganda and the Winam Gulf in Kenya. These areas are both ecologically important, pollution-prone (as seen in the annual average chlorophyll-a map for the lake in Fig. 2), and subject to land use and precipitation changes, making them ideal for studying the relationship between these factors and water quality (Calamari et al., 2006; Kabenge et al., 2016). They also serve as major sources for water abstraction for local communities and major cities like Kampala and Kisumu (Olokotum et al., 2020). The extracted data was plotted and analyzed to understand the trends and variability in water quality parameters and to relate them to climate and land use changes.