The &ldquo\;polar data gap&rdquo\; is a regi on around the North Pole where satellite orbits do not provide sufficient coverage for estimating se a ice concentrations. This gap is conventionally m ade circular and assumed to be ice-covered for the purpose of sea ice extent calculations\, but rece nt conditions around the perimeter of the gap indi cate that this assumption may already be invalid. We present partial differential equation-based mod els for estimating sea ice concentrations within t he area of the data gap. In particular\, the sea i ce concentration field is assumed to satisfy Lapla ce&rsquo\;s equation with boundary conditions dete rmined by observed sea ice concentrations on the p erimeter of the gap region. This type of idealizat ion in the concentration field has already proved to be quite useful in establishing an objective me thod for measuring the &ldquo\;width&rdquo\; of th e marginal ice zone&mdash\;a highly irregular\, an nular-shaped region of the ice pack that interacts with the ocean\, and typically surrounds the inne r core of most densely packed sea ice. Realistic s patial heterogeneity in the idealized concentratio n field is achieved by adding a spatially autocorr elated stochastic field with temporally varying st andard deviation derived from the variability of o bservations around the gap. Testing in circular re gions around the gap yields observation-model corr elation exceeding 0.6 to 0.7\, and sea ice concent ration mean absolute deviations smaller than 0.01. This approach based on solving an elliptic partia l differential equation with given boundary condit ions has sufficient generality to also provide mor e sophisticated models which could be more accurat e than the Laplace equation version\, and such pot ential generalizations are explored.

LOCATION:Seminar Room 1\, Newton Institute CONTACT:INI IT END:VEVENT END:VCALENDAR