Publications

Yackel, JJ; Nandan, V; Mahmud, M; Scharien, R; Kang, JW; Geldsetzer, T (2018). A spectral mixture analysis approach to quantify Arctic first-year sea ice melt pond fraction using QuickBird and MODIS reflectance data. REMOTE SENSING OF ENVIRONMENT, 204, 704-716.

Abstract
Melt ponds play a significant role in the summer decay of sea ice due to the fact that their albedo is significantly lower than surrounding snow and sea ice surface. Despite its requirement for thermodynamic sea ice modeling, measurement of melt pond areal coverage using satellite remote sensing has proven difficult due to significant spatiotemporal variability in the timing and evolution of melt ponds. Less than optimal results from prior studies employing a spectral mixture analysis (SMA) towards the determination of melt pond areal coverage from satellite remote sensing data provided the incentive for a multiple endmember spectral mixture analysis (MESMA) approach. The MESMA was performed on Moderate Resolution Imaging Spectroradiometer (MODIS) imagery using endmember spectra obtained from atmospherically corrected coincident high resolution imagery, surface observations and modeling. Results were validated against a high resolution Quickbird image acquired coincident to the MODIS image. The validation indicates that the best MESMA results provide consistent estimates of melt pond coverage for regions with high pond coverage (within 5% melt pond coverage) but overestimate pond fraction for regions with low pond coverage (by 10% or more). This may be due to deficiencies in the representation of sea ice surfaces within the endmember library used, oversimplified modeling of the ice surface and shortcomings in the validation process. However, it is assumed that with further refinement, the MESMA technique could allow for reliable estimates of the areal coverage of sea ice melt ponds using low resolution (large spatial coverage) optical satellite imagery under a wide variety of spatiotemporal pond evolution and fraction conditions.

DOI:
10.1016/j.rse.2017.09.030

ISSN:
0034-4257