Observing trends and extreme events impacts on lakes using ESA CCI Satellite Data Package

Lakes are already responding rapidly to climate change and in coming decades it is projected that global warming will have a more persistent and stronger effects on hydrology, nutrient cycling, and biodiversity (Cardoso et al. 2009; Carpenter et al. 2011). Recent projections estimate that lakes will get warmer for longer periods, with heatwaves potentially spreading across multiple seasons (Woolway et al. 2021). In some regions, heatwaves can add to existing pressure from drought which can lower lake levels and areal extent resulting from reduced inflows, increased evaporation and extraction for anthropogenic purposes (M?ynski et al. 2021; Zhao et al. 2022). The double heatwave event that occurred in the summer of 2019 in Europe was one of the top five warmest summers since 1500 (Sousa et al. 2020). In lakes, intense phytoplankton blooms might be the result of consecutive heatwave (e.g., Søndergaard et al. 2003; Free et al. 2021). When high temperatures are combined with low humidity, low rainfall, dry vegetation there is an increased risk of wildfire in case there is a source of ignition. Wildfires can have a significant hydro-geomorphological impact on watersheds in relation to post-fire rainfall events that can trigger erosion and transport processes leading to potential alteration of water quality (Smith et al., 2011). Despite the increased concern on the impact of wildfires on lake water quality, an uneven coverage of their geographical distribution has been observed (Shakesby and Doerr, 2006). Moreover, the topic has mainly been addressed at small scales while there remains a poor understanding at larger scales. The European Space Agency (ESA) Climate Change Initiative (cci) products could fill these gaps by providing long term, global coverage of both fire and water quality satellite-derived data also for the remote regions. The Lakes_cci project develops products covering Lake Water Level (LWL), Lake Water Extent (LWE), Lake Surface Water Temperature (LSWT), Lake Ice Cover (LIC) and Lake Water-Leaving Reflectance (LWLR) with the overarching objective to produce and validate a consistent long term dataset. The first phase of the project has recently been completed with the release of the last version (v2.0.2) of the dataset, including about 2000 lakes for the period 1992-2020. The dataset (netCDF file format) is hosted at the Centre for Environmental Data Analysis (https://catalogue.ceda.ac.uk/uuid/a07deacaffb8453e93d57ee214676304). A third user survey was conducted to collect user feedback on the data exploitation (https://climate.esa.int/en/projects/lakes/news-and-events/news/a-new-survey-for-users/). Phase 2 of the project started in July 2022. One scope of the project is the integration of different satellite-derived products across ESA CCI projects. For this reason, a study on wildfire and lakes is ongoing aiming to investigate the relationship between fires and lakes water quality over a wide range of geographical regions and fire regimes. The Fire_cci project, already at phase 2, focuses on several issues relating to fire disturbance including analysing and specifying scientific requirements relating to climate, production of burned area datasets, and product validation and product assessment. In this study potential of the CCI dataset is explored in three different case studies in the Eurasian region. I.Long-term trends in the ECV "Lakes" The dataset was explored for two Italian lakes and one Swedish lake of different depth and trophic state. The lakes are part of the Long-Term Ecosystem Research (LTER) network. In situ data from the LTER dataset were used to compare and integrate satellite products. Time-series of satellite data were then explored to examine trends in the context of key meteo-climatic variables. LSWT, chlorophyll-a (Chl-a), turbidity and ice cover data covering 16 years (2003-2018) were extracted. Daily climate data (wind, air temperature, precipitation) were obtained from ERA5. North Atlantic Oscillation (NAO) daily values were obtained from NOAA-CPC. The analyses revealed variations in the water quality variables, including a significant alteration in the concentrations of Chl-a in the lakes under study. Another aspect highlighted in the study is the variation in response to climate change in lakes in different geographical regions and with different trophic and morphological characteristics, when comparing northern Europe to southern Europe. For example, in Lake Trasimeno (shallow-eutrophic) Chl-a was higher with more positive values of the NAO, lower lake levels and warmer temperatures; in Lake Garda (deep oligotrophic) a shift was found in the timing of the traditional Chl-a peak (Fig. 1); the Erken lake time series indicated a significant increase in Chl-a and air temperature. II. Heatwave and storm events impact on lakes Chl-a data were examined for any potential responses during the 2019 double heatwave period for 36 European lakes, evaluating how the response varies depending on latitude, total concentration of phosphorus and the average depth of the lake. The Chl-a concentrations for summer 2019 were extracted from Lakes_cci dataset (v1.1). Data on total phosphorus and lake mean depth were obtained from Waterbase, the European Environment Agency database on water quality (https://www.eea.europa.eu/data-and-m aps/data/waterbase-water-quality-icm-1) or from the Environmental data MVM database for Swedish lakes (https://miljodata.slu.se/mvm/) or from published literature. The results show that the timing and magnitude of the response to the heatwave events depends on lake depth and nutrients (Fig. 2). Deeper lakes respond sooner probably because of higher temperatures leading to stronger stratification thereby improving the light climate but with the response strength dependent on nutrient status (e.g. Maggiore and Geneva). In contrast, shallower lakes and lakes at lower latitudes showed more asynchrony with a greater response after the heatwave event probably as a result of internal and external loading (e.g. Razim and Balaton). III.Effects of wildfires on Lake Baikal Fire_cci and Lakes_cci datasets were explored to highlight any spatial-temporal relationships of burned area, meteo-climatic and water quality parameters for three sub-basins in correspondence to the inflows of the major tributaries into Lake Baikal. Burning vegetation in the basin could cause an increase in erosion and surface transport process, and a subsequent increase in turbidity and/or concentration of Chl-a in the lake waters. The products were extracted for the period 2003-2019. The first analyses focused on trends in the time series observed independently. Results showed a trend towards an increase in burned area, chlorophyll-a and turbidity in the summer months, over the 16-year period, probably related to climate change. The time series of precipitation were analysed through the use of the standardized precipitation index (SPI). It was observed that SPI assumes significantly negative values and drought indicators for the years 2003 and 2015 in which numerous fires were also observed. A non-parametric multiplicative regression model showed that the temporal variable (seasonal and annual) is the main predictor of turbidity and Chl-a. The predictive role of the burned area and wind were limited in the investigated study area. Effectively, on long-time series over a such a deep, pristine and large lake, data did not show clear effects of wildfires on water parameters except for local effects. For example, the spatial-temporal analysis conducted in some years (2003, 2006, 2011 and 2018) of significant interest, because they represent extreme conditions of fires and precipitation, showed an increase in both Chl-a and turbidity following fire events that resulted in a significant burned area and without significant precipitation events observed in the central portion of the lake for the year 2003 (Fig. 3).