Sea ice concentration algorithm

Sea ice concentration is the proportion or fraction of sea ice in a pixel, expressed as a percentage. Sea ice concentration is the fundamental geophysical parameter derived from passive microwave brightness temperature (TB) data [1]. Numerous algorithms to retrieve sea ice concentration from microwave sensor data have been developed. The NORSEX algorithm [1] is used here to calculate first-year (FY) ice, multi-year (MY) ice and open water (OW) fractions of the arctic surface.  According to this algorithm, the TBs  from three-component footprint is assumed to be the sum of individual emitted brightness temperatures, weighted by the concentration of every fraction:

TB=CMYeMYTMY + CFYeFYTFY  + COWeOW272,

where CMY, CFY and COW  are the concentrations of multi-year ice, first-year ice and open water, which are assumed to cover the field of view of the instrument, eMY, eFY and eOW  are emissivities of the respective components and  TMY  and  TFY are the temperatures of two types of ice, which are related to monthly average atmospheric near surface temperature.  The algorithm is highly sensitive to the emissivities of the three types of surfaces and the emissivities of multi-year ice vary considerably both in time and space [2-4], such that it is essential to construct the emissivity model to represent adequately the three-component surface.

The emissivities used for processing Scanning Multi-channel Microwave Radiometer  (SMMR) data are as follows
 
type of surface 18 G Hz 37 GHz
water .63 .69
FY ice .95 .95
MY ice .83 .74

and for SSMI –
 
type of surface 19.35 GHz 37 GHz
water .65 .75
FY ice .97 .97
MY ice .82 .74

The microwave characteristics of MY and FY ice may become confounded during  melting/freezing periods. Therefore, in order to increase the reliability of the MY-FY ice discrimination, we concentrated on the winter period – from November through March in our MY ice study.  Weather filters suggested by [5] and [6] have been used in calculations. In order to improve the algorithm, we: 1) used published field data [1-5] on measured emissivities in the Arctic, 2) took into account temporal changes of multi-year ice characteristics using their relation to atmospheric temperature, and 3) controlled the discrepancies between the shape of each late summer ice pack and multi-year ice maps during the following winter, in order to provide a reasonable correspondence between them [7, 8].

References

[1] E. Svendsen, K. Kloster, B. Farrelly, O.M. Johannessen, J. Johannessen, W. Campbell, P. Gloersen, D. Cavalieri and C. Matzler (1983). Norwegian Remote Sensing Experiment: Evaluation of the Nimbus-7 SMMR for sea ice research. J. Geophys. Res., 88, 2781-2791.
[2] J. Comiso (1986). Characteristics of Arctic winter sea ice from satellite multispectral microwave observations. J. Geophys. Res., 91, 975-994.
[3] T. Grenfell (1992). Surface-based passive microwave studies of multiyear sea ice. J. Geophys. Res., 97, 3485-3501.
[4] F. Carsey, ed., (1992). Microwave remote sensing of sea ice. Geophysical Monograph 68, American Geophysical Union, 462 pp.
[5] D. Cavalieri, J. Crawford, M. Drinkwater, D. Eppler, L.Farmer, R. Jentz and C. Wackerman (1991). Aircraft active and passive- microwave validation of sea ice concentration from the DMSP SSMI. J. Geophys. Res., 96, 21,989-22,008.
[6] P. Gloersen and D. Cavalieri (1986). Reduction of weather effects in the calculation of sea ice concentration from microwave radiances . J. Geophys. Res., 91, 3913-3919.
[7] E. Shalina, O.M. Johannessen and M.W. Miles (1999). Arctic ice transformations: multiyear ice changes in comparison with summer minima. International Geosciences and Remote Sensing Symposium, IGARSS99, 28 June - 2 July 1999, Hamburg, Germany, 1999, 2023-2026.
[8] O.M. Johannessen, E. Shalina and M.W. Miles (1999). Satellite evidence for an Arctic sea ice cover in transformation. Science, 286,1937-1939.