Spatial Correlation of Surface-Level Variables over Arctic Sea Ice

During SHEBA, the 1997–1998 experiment to study the Surface Heat Budget of the Arctic Ocean, the SHEBA Atmospheric Surface Flux Group made hourly measurements of surface-level meteorological variables and surface fluxes at five sites separated by distances from 200 m to 12 km. The spatial correlation of these observations as a function of separation has implications for modeling air-sea-ice interaction with models of all horizontal resolutions. That is, the spatial correlation of the observations addresses the assumption that a single grid-cell value is adequate for representing an entire grid cell.

Our correlation results are mixed. Mean meteorological variables such as wind speed, 2-meter air temperature, surface temperature, and pressure are well correlated for all spatial separations in our dataset. The radiative fluxes--incoming and outgoing longwave and shortwave radiation--are also generally well correlated except in winter, December, January, and February. The friction velocity and surface sensible heat flux are not as well correlated among the sites, but the friction velocity and the sensible and latent heat fluxes estimated with a bulk flux algorithm are better correlated than the measured fluxes. In the variables that show good correlation among the sites--namely, wind speed, air temperature, surface temperature, pressure, and longwave and shortwave radiation--we see no consistent decay in the correlation with increasing separation up to the largest separation in our dataset, 12 km. This result implies that, for model grid cells of 12 km and less, using a single value for a state variable is a reasonable approximation.