POSTER SESSION 2

10:30–12:30, Tuesday, September 15

Poster Session | 1 | 2 | 3 | 4 | Schedule at a Glance

ABSTRACT 1223 | POSTER 234

ESTIMATING TOTAL IRON IN DUTCH INLAND WATERS FROM SENTINEL-2

Iron is not represented in operational satellite water-quality products, yet in lowland catchments it can indicate groundwater influence and is linked to phosphate retention and nitrogen dynamics across redox transitions, making it a potential optical indicator during low-productivity periods. We evaluate whether total iron can be estimated from Sentinel-2 reflectance in optically complex Dutch inland waters, and whether spectral information improves estimation beyond environmental context.

We combined in situ iron measurements with quality-controlled Sentinel-2 MSI matchups from 2015 to 2024. Pixel selection was limited to inland Water Framework Directive water bodies, excluding brackish and tidal waters, applying 5 to 200 m width filters, and retaining peat, sand or clay matchups. This yielded 1096 observations. Spectral separability was quantified with mean absolute difference and spectral angle mapper, using threshold testing and comparisons of total iron against co-measured suspended matter, organic carbon, transparency and chlorophyll-a.

Iron contributed substantially to Sentinel-2 spectral variation, but separability was context-dependent. At 0.5 mg/L, iron exceeded suspended solids and organic carbon in spectral-shape separation in most contexts, but was generally weaker than transparency and chlorophyll-a. Separability increased in clearer, low-chlorophyll-a waters, and was strongest in narrow, colder waters. For classification above 0.5 mg/L, adding reflectance-derived predictors to environmental predictors improved balanced accuracy from 0.74 to 0.82 and PR-AUC from 0.86 to 0.93. Continuous estimation improved OOF R squared from 0.14 to 0.45. These results indicate that iron is a promising optical variable for inland-water remote sensing and mapping, motivating compound-specific inherent optical property characterisation of iron phases.

Marit van Oostende, Vrije Universiteit Amsterdam, [email protected], https://orcid.org/0000-0003-3364-3785

Ype van der Velde, Vrije Universiteit, [email protected], https://orcid.org/0000-0002-2183-2573

Poster Session | 1 | 2 | 3 | 4Schedule at a Glance

Keep up to date

Sign up to receive email updates to be sure to catch all the meeting news.

Questions?

Contact Jenny Ramarui,
Conference Coordinator,
at [email protected]
or (1) 301-251-7708

Translate »