Tracking blue ice in time: deriving Antarctic blue ice fraction from MODIS images using spectral unmixing
- 1Institute for Marine and Atmospheric research Utrecht (IMAU), Utrecht University, Utrecht, The Netherlands
- 2Department of Geoscience Remote Sensing, Delft University of Technology, Delft, The Netherlands
- 3German Remote Sensing Data Center (DFD), German Aerospace Center (DLR), Wessling, Germany
- 4Applied Geoinformation Science Lab, Geography Department, Humboldt Universitaet zu Berlin, Berlin, Germany
Antarctic Blue Ice Areas (BIAs) are a sensitive indicator for climate change. They can be formed either by wind and sublimation or by surface melt, and vary over time. In this regard, distinguishing different blue ice types and observing their change over time can enhance our understanding of climate change in Antarctica. Presently, the areal extent of BIA is retrieved using Earth observation satellites. However, such products rarely provide time series of BIA extent over the entire continent. To fill this gap, we derived blue ice fraction over Antarctica from the moderate resolution imaging spectroradiometer (MODIS) using spectral mixture analysis. Blue ice fraction is defined as the fraction of each MODIS pixel that is covered by blue ice. The results provide a continuous time series of blue ice fraction during the austral summers 2000 to 2021. This time series shows Antarctic blue ice abundance and exposure over time, and indicates that melt-induced BIAs are more variable in time than wind-induced. According to the accuracy assessment based on high-resolution Sentinel-2 images over six selected test sites in coastal East Antarctica, the blue ice fraction results have an overall uncertainty of around 15% and 25% in wind- and melt-induced BIAs, respectively. The uncertainties mainly arise due to the very similar spectral profiles among melt streams, lakes, and ponds. Overall, our results show great potential in (1) generating annual BIA maps, (2) separating wind-and melt-induced BIAs, (3) evaluating (regional) climate model outputs, and (4) deriving temporal variations in blue ice abundance and exposure.
How to cite: Hu, Z., Kuipers Munneke, P., Lhermitte, S., Dirscherl, M., Ji, C., and van den Broeke, M.: Tracking blue ice in time: deriving Antarctic blue ice fraction from MODIS images using spectral unmixing, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-5343, https://doi.org/10.5194/egusphere-egu22-5343, 2022.