Estimation of carbon content and dissolved organic matter in inland waters using EnMAP hyperspectral imaging

Abstract. Inland water bodies, despite occupying a small fraction of the Earth’s Surface, play a significant role as carbon sinks or emitters of carbon dioxide and other greenhouse gases. Remote sensing has the potential to provide continuous carbon estimates; however, conventional operational missions are currently limited by the number of bands, which hinders the detailed detection of certain peaks in reflectance spectra. The hyperspectral EnMAP sensor provides images with 260 bands and a spatial resolution of 30 m. The objective of this study is to estimate the carbon content in continental water bodies of the eastern Iberian Peninsula (Spain), with varying trophic states, through four variables: total organic carbon (TOC), dissolved organic carbon (DOC), particulate organic carbon (POC), and colored dissolved organic matter (CDOM). Different indices and combinations of bands have been studied, based on in situ measurements of these variables coinciding with the acquisition of EnMAP images. The results show that TOC, can be estimated with a triband index using the 525, 535 and 551 nm bands (R² = 0.96) and an exponential fitting curve; DOC with a normalized difference using the 599 and 525 nm bands (R² = 0.93) with an exponential fitting curve; CDOM with simple ratio between bands 435 and 721 (R² = 0.95) with a linear fitting curve. POC is calculated as the difference between TOC and DOC (R² = 0.86).