Steps towards a status classification of Norwegian lakes using satellite data - method development and case studies

Abstract

Sentinel 2 & 3 data for 2016 to 2020 were processed with C2RCC processor for six lakes in the ØKOSTOR monitoring program: Mjøsa, Selbusjøen, Gjende, Røssvatnet, Snåsavatnet and Femunden. In this report we have taken steps towards an operational remote sensing (RS) service for Norwegian lakes. Maps of multi-year chlorophyll-a concentrations (chl-a) were produced. Chl-a status classification based on RS was assessed and compared with the status based on in-situ sampling. The chl-a satellite data was variable over month, year and for an assessment period of 4-5 years It worked well for some lakes but overestimations of the chl-a retrieval can most likely be improved by the use of other algorithms or by finding a more correct conversions factor between chl-a concentration and chl-a absorption. The amount of data for each lake increases with satellite observations, on average 15 observations per year (Sentinel 2) and 19 (Sentinel 3), compared to 4-6/year or every 4th year with in-situ observations. RS can provide new knowledge of spatial and temporal dynamics and is thus a good complement to ordinary sampling for water quality classifications of Norwegian lakes, but it needs more in-depth work for possible adjustments and corrections.