Impact of iron associated to organic matter on remote sensing estimates of lake carbon content

- Colloidal iron affects lake CDOM remote sensing.
- Iron content in lake water can be mapped with MERIS data.
- Lake water SUVA (DOM quality) can be mapped with remote sensing.

There is a strong need to develop remote sensing methods for mapping lake carbon content on regional to global scales. The use of in situ methods is impractical for monitoring lake water quality over large geographical areas, which is a fundamental requirement to understand the true role of lakes in the global carbon cycle. The coloured component of dissolved organic carbon (DOC), called CDOM, absorbs light strongly in the blue part of the visible spectrum and can be used as a proxy for mapping lake DOC with remote sensing. However, iron associated to organic matter can cause extra browning of waters. Consequently, the remote sensing signal we interpret as DOC may partially be attributed to the presence of iron associated to organic matter, potentially hampering our ability to estimate carbon concentrations.A thorough analysis of biogeochemical parameters was carried out on Lake Mälaren on August 23, 2010, and a MERIS full resolution image was acquired simultaneously. MERIS standard, Case 2 Regional, and Boreal processors were used to calculate remote sensing products, which were compared with different lake water characteristics.The carbon to iron ratio was different from the rest of the lake in one of the basins. MERIS standard and Case 2 Regional processors were sensitive to this difference as the correlation between MERIS CDOM product and DOC was low (R2 = 0.43) for all sampling stations and increased to 0.92 when the one basin was excluded. The Boreal Lakes processor results were less disturbed by the different carbon-iron ratios found in one basin and produced reasonably good results (R2 = 0.65).We found MERIS products (e.g. total absorption) that provided good correlation (R2 = 0.80) with DOC-specific absorbance at 254 nm, called SUVA, which is a metric commonly used to assess drinking water treatability. However, none of the MERIS products were suitable for mapping the total organic carbon in Lake Mälaren. MERIS total suspended matter product was a good (R2 = 0.73) proxy for particulate iron, meaning that the particulate iron content in Mälaren can be mapped from space.