Evaluation of MERIS and Sentinel-3 OLCI Compatible Algorithms for Chlorophyll-a Retrieval in UK Lakes
Hunter, Peter1; Tyler, Andrew1; Spyrakos, Evangelos1; Riddick, Caitlin1; Pratt, Amanda1; Maberly, Stephen2; Carvalho, Laurence2; Groom, Steve3
1University of Stirling, UNITED KINGDOM; 2Centre for Ecology & Hydrology, UNITED KINGDOM; 3Plymouth Marine Laboratory, UNITED KINGDOM

There are some 304 million natural lakes on Earth. These cover >3% of the terrestrial land surface and as such represent an ecosystem of global significance. Lakes also play a key role as conduits and reactors in the global carbon cycle, act as regulators and sentinels of climate change and provide many essential ecosystem services to society. The ability to observe changes in the physical, chemical and biological characteristics of lakes is crucial to understanding their responses to environmental change at local, regional and global scales. However, the number, size and extensive geographical distribution of lakes are major constraints on the development of effective observational strategies. Currently, our inability to monitor lakes at appropriate spatial and temporal scales is a major source of uncertainty in our understanding of their global status and a constraint on our capacity to predict future responses to environmental change. The use of satellite observations for monitoring change in the status of lakes has progressed significantly in recent years but the complexity in the optical properties of these water bodies and the continentality of the overlying atmosphere still presents notable challenges for the accurate retrieval of in-water biogeochemical constituents. In addition, strong land adjacency effects also diminish the accuracy of satellite observations over lakes. Consequently, algorithms for Chla retrieval developed for ocean and coastal waters have limited applicability to turbid lakes. Recently, several new algorithms have been developed for the estimation of Chla in lakes and other turbid waters that are compatible with MERIS and Sentinel-3 OLCI data but currently these have not been widely validated for different lake types and geographical regions. As part of the GloboLakes project, we have been undertaking a retrospective validation of MERIS-compatible algorithms for Chla estimation in lakes and other Case 2 waters using data from several UK lakes (including Lough Neagh, Loch Lomond, Loch Leven and Windermere) encompassing a large trophic gradient. This includes existing neutral network inversion algorithms for Case 2 waters (e.g., Case-2-Regional, Boreal Lakes, Fub WeW) and alternative near infrared-red band ratio algorithms. The satellite products were validated against in situ Chla measurements from routine lake monitoring programmes. This talk will present some initial findings from this validation exercise and outline future work planned on the development and validation of algorithms for MERIS and Sentienl-3 OLCI to be undertaken during the GloboLakes project.