Sea Ice Concentration Algorithm And Study On The Physical Process About Sea Ice And Melt-pond Change In Central Arctic

Sea ice concentration is one of most important parameter of Arctic, and it can beretrieved from passive microwave radiation data. A new algorithm for retrieving seaice concentration is derived by the basic equation of the brightness temperature andby dual-polarized brightness temperatures at36.5GHz. This algorithm could becalled DPR algorithm. In the DPR algorithm, the sea ice concentration depends onlyon the ratio of dual-polarized emissivity of sea ice. When sea ice concentration closesto1, the vertically polarized brightness temperature and horizontally polarizedbrightness temperature at36.5GHz assemble a line in the scatter plot, which crossesthe origin of coordinates. And the ratio of dual-polarized brightness temperature notonly is the ratio of dual-polarized emissivity of sea ice, but also is slope of the line.The ratio of dual-polarized emissivity of sea ice is little influenced by sea ice types(first year ice or multiyear ice) and it is little changes, about from0.92to0.96. So if areasonable ratio be made sure, the sea ice concentration can be retrieved using DPRalgorithm.The parameter of contrast-ratio can show the changing features of the ratio ofbrightness temperature from open water to sea ice covered area. A threshold value atthe sea ice margins can be determined using the contrast-ratio of the raiodual-polarized brightness temperatures at36.5GHz. The value and the number of theratio are all change at the sea ice margins because of sea water, so contrast-ratiorapidly change. The study indicates that the contrast-ratio rapidly changed and itsgradient appears extreme value when the ratio of brightness temperature changesaround0.92. In the scatter plot, the dual-polarized brightness temperature at36.5GHzapparently assemble a zero-crossing line when their ratio larger than0.92. So the ratioof dual-polarized emissivity of sea ice in DPR algorithm is taken as0.92.Contrast-ratio is also used to determine the open water margins by the the ratio ofvertically polarized brightness temperature at18.7and36.5GHz. When the ratiochanges around0.90, the gradient of contrast-ratio appears extreme value. So the threshold value of open water is0.90.The ice concentration retrieved by the DPR algorithm is compared with the NT2and ABA algorithms. In higher sea ice concentration covering area, their difference isvery little (usually less than5%), however the bigger difference takes place in margi-nal ice zones. The comparison of different algorithms found that all algorithms takefull advantage of the features of brightness temperature at higer sea ice concentrationarea, but there are different approaches at lower sea ice concentration area. Forexample, NT2needs to identify many parameters to ensure the reasonable results andABA not only needs to identify many parmeters but also needs identify the position ofthe line which is assembled by brightness temperature. In DPR algorithm, only theratio of dual-polarized emissivity of sea ice needs to been identified and otherparameters are not allowed to be human adjustment. The mean error,root-mean-square error and mean absolute error of the DPR algorithm are relativelybetter than those from the other two algorithms according to the verification results.Because the microwave emissivity of melt-pond closes to open water, the results ofthe DPR algorithm would be closer to the actual ice concentration when the ice undera pond fully melts in summertime. So melt-pond can not be igored when usingmicrowave radiation data retrieve sea ice concentration.The rapid change of Arctic sea ice also profound influences the sea iceconcentration change in the central Arctic. In2010summer, an extremely low iceconcentration (ELIC) appeared in the central Arctic no matter what kind of sea iceconcentration algorithms. The navigation speed of the icebreaker identified thisphenomenon in central Arctic. It is verified based on the in situ ice thicknessobservation that the open water is not caused by ice melting. The driftingtrajectories of ice buoys showed a divergence in the central Arctic and Atlantic sector,which belongs to the double-gyre pattern of ice drifting, being more effective insummer to dilacerate pack ice and to create the ELIC. The open water absorbed moresolar energy under the ELIC condition. The increased energy of ocean resulted in anaccelerated retreat of summer sea ice cover and produced intense evaporation throughair-sea interface. A drizzle occurred at about87°N could be attributed to the feedback of low ice concentration to the atmosphere. As size of ice block is becoming smallerin recent years, the low ice concentration phenomenon will be easily to be producedand amplified as the small pieces of sea ice are more mobile to respond the windforcing. The ELIC will result in a low sea surface pressure in central Arctic as itsclimatologic effect on Arctic change.During2010summertime, the double-gyre pattern of ice drifting had animportant impact to the depth of melt-pond. The observation indicate that the latitudeis higher the deeper for melt-pond when the latitude highter than77°N. Earth’ssurface solar radiation calculated results show that the higher latitude the more solarradiation reaching earth when the latitude highter than70°N in summer. Sea ice occurmainly eastward drift under double-gyre pattern of ice drifting, but it is small inlongitude direction of change. So high-latitude melt-pond maintains a relatively moresolar radiation and more sea ice would be melt than low latitude sea ice. Therefore,solar radiation distribution and the ice drifting pattern are the mainly reasons thatmelt-pond depth changes significantly with latitude at Pacific sector.A one-dimensional and thermodynamic model of melt-pond is used to sudy thesea ice melting process. The model results indicate that the sea ice melting undermelt-pond is not consistent with the changes of melt-pond depth. Sensible heat, latentheat and net longwave radiation are forms for melt-pond losing heat, but all theseforms are relativevely small compared to the incident solar radiation. A large numberof solar radiation is absorbed by melt-pond water (more than50%). In the melt-pond,convective mixing in the vertical direction is drived because of solar radiation, and thesolar energy absorbed by melt-pond water continuously reach to the bottom to heatingsea ice and melting sea ice. Therefore, melt-pond temperature show distinct diurnalvariation at the influence of solar radiation, but the range is relatively small, generallymaintained at0-0.3°C. The study also indicate that sea ice melting rate relate tomelt-pond depth. When melt pond depth is less than0.4m, the change of sea icemelting rate is very fast. However when melt pond depth is larger than0.4m, the seaice melting is relatively stable and maintained around1cm/day.