Improved machine-learning-based open-water–sea-ice–cloud discrimination over wintertime Antarctic sea ice using MODIS thermal-infrared imagery

Paul, Stephan; Huntemann, Marcus

doi:https://doi.org/10.5194/tc-15-1551-2021

Articles | Volume 15, issue 3

https://doi.org/10.5194/tc-15-1551-2021

© Author(s) 2021. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/tc-15-1551-2021

© Author(s) 2021. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 15, issue 3

Research article

|

26 Mar 2021

Research article |

| 26 Mar 2021

Improved machine-learning-based open-water–sea-ice–cloud discrimination over wintertime Antarctic sea ice using MODIS thermal-infrared imagery

Stephan Paul and Marcus Huntemann

Data sets

Manually categorized initial training data for open-water/sea-ice/cloud discrimination Stephan Paul https://doi.org/10.5281/zenodo.4596407

Short summary

Cloud cover in the polar regions is difficult to identify at night when using only thermal-infrared data. This is due to occurrences of warm clouds over cold sea ice and cold clouds over warm sea ice. Especially the standard MODIS cloud mask frequently tends towards classifying open water and/or thin ice as cloud cover. Using a neural network, we present an improved discrimination between sea-ice, open-water and/or thin-ice, and cloud pixels in nighttime MODIS thermal-infrared satellite data.