From Space to Land: exploiting satellite-derived water quality variables for climate studies

Despite making up less than 1% of the world water area, lakes are an important resource that provide drinking water, biodiversity, and recreational opportunities—all of which are tied to sustainable development goals. Increasing urbanisation and population growth have led to eutrophication, hydrological changes and loss of ecosystem services. Invasive species, land use changes and climate change are recognized as the main drivers of species loss in freshwater environments, which may be five times faster than in terrestrial environments. In the coming decades, climate change and global warming, and in particular the increase in extreme weather events, are expected to have more widespread and significant impacts on biodiversity, species composition, hydrology, land cover and nutrient cycling. Using in-situ data to monitor such water bodies and comprehend their complex behavioral changes on a global scale is not feasible. Open access satellite-derived data represent a way forward in understanding the ecological processes and in assessing the impact of the main drivers of changes on freshwaters. The Lakes_cci (Climate Change Initiative) project provides global and consistent satellite observations of lake specific essential climate variables (ECVs): Lake Water Level and Extent, Surface Water Temperature, Ice Cover and Water-Leaving Reflectance (LWLR), which capture both the physical state of lakes and their biogeochemical response to physico-chemical and climatic forcing. With the release of version 2.1, the products cover the period 1992-2022 and provide daily data at 1 km resolution for over 2000 relatively large lakes. The project has explored multiple use cases that examine long-term time series of biophysical water quality parameters to understand possible causes of their trends, including the unique response of shallow lakes globally and the effects of heatwaves on lakes. In the first use case, we selected a globally distributed subset of shallow lakes (mean depth < 3m; n=347) to investigate long-term time series/trends (2002-2020) of chlorophyll-a (Chl-a) and turbidity derived from LWLR. Shallow lakes and wetlands form a major component of inland waters and provide many ecological services, particularly important for carbon storage and biodiversity. Due to their large surface-to-volume ratio, they are vulnerable to environmental changes influenced by nutrient and pollutant loads and are sensitive to climate change. According to the trend analysis, turbidity increased significantly in 60% of the shallow lakes and decreased in 17% of the lakes, while Chl-a increased significantly in 45% of the lakes and decreased in 22% of the lakes. Further investigation revealed that in most lakes the parameters turbidity (50%) and Chl-a (48%) increased simultaneously with LSWT, suggesting an impact of climate warming on lake water quality. Chl-a and turbidity in most lakes increased positively with population and gross regional product, according to a structural equation model-based analysis used to model the interactions between climatic, socio-economic and total water conditions. This finding suggests that human population growth in a lake’s catchment represents an important pressure on lake water quality. In the second use case, exploiting the high frequency and coverage of observations, we were able to gain insight into the response of lakes to sequential extreme weather events, such as heat waves and monsoon rainfall events, that occurred in India in 2019. Indian lakes are an excellent test subset across Lakes_cci variables as monsoon dynamics require that lake turbidity, chlorophyll-a, LWL and climatic variables are considered together at seasonal and annual scale. We examined the water quality response using time series and TAM (Time Alignment Measurement) analysis, which measures the degree of synchrony with the heatwave event, followed by cluster analysis of Chl-a and turbidity patterns. The TAM analysis showed that the rainfall time series was closer in phase with air temperature and turbidity, but less so with Chl-a, indicating a driving influence of rainfall on turbidity, probably due to the strong influence of the monsoon. The available LWL data showed high variability over a short period of time. Cluster analysis revealed two main groups of turbidity patterns: northern lakes showed peaks most likely driven by spring snowmelt, while southern lakes were dominated by peaks driven by the summer monsoon. In contrast, Chl-a patterns were less related to hydro-morphology, and likely more influenced by local nutrient dynamics and changes in LWL, helping to focus further studies centred on individual lakes.