(2014) used this sensor to observe sea ice development. (2002) demonstrated the potential of Landsat 7 ETM+ (L-7 30 m) for the classification of summertime sea ice surface conditions and retrieved the spatial distribution of ponded/unponded ice and open water ( Markus et al., 2003), while Landy et al. Studies based on optical sensors with a higher spatial resolution (≤ 30 m), however, are rare. Despite a high temporal and spatial coverage of the abovementioned sensors (swath widths: AVHRR = 2,900 km, MODIS = 2,330 km, S-3/OLCI = 1,270 km), their coarse spatial resolution (AVHRR: 1.1–4 km, MODIS: 250–1,000 m, S-3/OLCI: 300 m) impedes detailed observations of sea ice features such as melt ponds and ridges, which exist at spatial scales of meters to tens of meters ( Figure 1). Recently, Istomina and Heygster (2017) published a retrieval algorithm for sea ice albedo and melt pond fraction from S-3/OLCI observations. The Sentinel-3 satellites (S-3) from the European Space Agency's (ESA) Copernicus program, carrying the Ocean and Land Color Instrument (OLCI), continue this tradition ( Donlon et al., 2012 Malenovský et al., 2012). (2008)) have long been used in Arctic research ( Pope et al., 2014 Nasonova et al., 2017) providing observations of different parameters such as sea ice extent, sea ice thickness and albedo. (2007)) and the MODerate Resolution Imaging Spectroradiometer (MODIS e.g., Rösel et al. Optical sensors such as the Advanced Visible High Resolution Radiometer (AVHRR e.g., Huck et al. The retrieval of parameters, such as the spectral albedo of sea ice or snow and properties of melt ponds and leads, relies on spectral information the retrieval of optical properties of target surfaces and waterbodies therefore is necessary. The limited accessibility of Arctic sea ice makes satellite remote sensing an important tool for synoptic observations in this region. Medians of average reflectance and NDMI range from 0.80–0.97 to 0.12–0.18 while medians for TOA are 0.75 and 0.06, respectively. To illustrate consequences of processor selection on secondary products we computed average surface reflectance of six bands and normalized difference melt index (NDMI) on an image subset. Parameterization based on external data, therefore, was necessary to obtain reliable results. ACOLITE estimated AOT within the uncertainty range of AERONET measurements. ATCOR, iCOR and Sen2Cor failed in the image-based retrieval of atmospheric parameters (aerosol optical thickness, water vapor). ACOLITE, ATCOR, and iCOR performed well over sea ice and Polymer generated the best results over open water. For open water, we additionally evaluated intensities. We used spectral shapes to assess performance for ice and snow surfaces. We evaluate the results of the different processors using in situ spectral measurements of ice and snow and open water gathered north of Svalbard during RV Polarstern cruise PS106.1 in summer 2017. We therefore provide an evaluation of five currently available atmospheric correction processors for S-2 (ACOLITE, ATCOR, iCOR, Polymer, and Sen2Cor). For an accurate retrieval of parameters such as surface albedo, the elimination of atmospheric influences in the data is essential. The MultiSpectral Instrument on board the Sentinel-2 (S-2) satellites of the European Space Agency offers new possibilities for Arctic research S-2A and S-2B provide 13 spectral bands between 443 and 2,202 nm and spatial resolutions between 10 and 60 m, which may enable the monitoring of large areas of Arctic sea ice.
![google earth msi google earth msi](https://www.msi.com/Landing/creator-awards/uploads/images/09b550a31c18a66bb6c6e9ebc9def8c8.jpg)
![google earth msi google earth msi](https://www.get-itsolutions.com/wp-content/uploads/2021/08/Best-Computer-for-Microsoft-Office-Excel-Laptop-and-Desktop-150x150.png)
Multispectral remote sensing may be a powerful tool for areal retrieval of biogeophysical parameters in the Arctic sea ice. 2Institute of Coastal Research, Helmholtz-Zentrum Geesthacht, Geesthacht, Germany.1Earth Observation and Modelling, Department of Geography, Kiel University, Kiel, Germany.Marcel König 1 *, Martin Hieronymi 2 and Natascha Oppelt 1