RELATED LINKS

 

Accomplishments

Estimating melt pond concentrations from passive microwave satellite observations.

Two methods were developed to estimate melt pond concentrations from passive microwave satellite observations (SSM/I data) based on approximation of the visual melt pond detection results during SHEBA observations and melt pond concentrations derived from AVHRR data. Melt pond season was determined by dates of melt and freeze onset for perennial (MY) sea ice, and melt onset date and ice disappearance date for annual (FY) ice (Figure 1).

figure (64K)
Figure 1. Regional variability of the Arctic melt season during 1998 (melt and freeze onset dates were estimated with the PMSTA algorithm (Belchansky et al., 2004), sea ice disappearance dates were defined from the analysis of the Bootstrap sea ice concentrations).

According to the SHEBA-based linear model, mean melt pond concentrations on the MY sea ice were 9.5%, 18.6%, and 12.5% for June-August 1998, respectively. Maximum melt pond concentration on the FY peripheral sea ice in summer 1998 was 69%, and the mean was 27%. Figures 2 and 3 show that during Arctic melt season, peak melt pond concentrations typically occur on the FY ice. MY sea ice surface is less affected by the melt than that of the FY sea ice, and peak melt pond concentrations on the MY ice are observed in July.

figure (51K)
Figure 2. Peak melt pond concentrations (%) over the FY sea ice that disappeared during summer 1998 (left), and monthly mean melt pond concentrations (%) over the MY sea ice that survived in summer 1998, derived from the SHEBA-based model.
figure (16K)
Figure 3. Evolution of the melt pond concentrations (area fractions) in SSM/I pixels containing MY (Belchansky et al, 2005) and FY sea ice, survived and disappeared, respectively, in summer 1998. Melt pond concentrations were defined from the SHEBA-based model. FY ice melt pond area fraction decreased together with that of the FY sea ice.

An example of melt pond detection using visible AVHRR channel is shown on Figure 4.

figure (30K)
Figure 4. The AVHRR image and result of its classification, contributing to the integrated melt pond concentration data set.
Figure 5 demonstrates seasonal and latitudinal changes of melt pond concentrations for clear-sky conditions. Comparison of melt pond concentrations from this regression and SHEBA aerial data gives mean difference of -1±5%.
figure (30K)
figure (30K)
figure (30K)
Figure 5. Interannual, seasonal and latitudinal changes of melt pond concentrations for clear-sky conditions derived from the AVHRR (source) and passive microwave observations (fitted).

References

Belchansky, G. I., D. C. Douglas, and N. G. Platonov (2004), Duration of the Arctic sea ice melt season: Regional and interannual variability, 1979-2001, J. Clim., 17, 67-80.

Belchansky, G. I., D. C. Douglas, V.A. Eremeev, and N. G. Platonov (2005), Variations in the Arctic's multiyear sea ice cover: A neural network analysis of SMMR-SSM/I data, 1979-2004, Geophys. Res. Lett., 32, L09605, doi:10.1029/2005GL022395.