Observations of Surface Heat Fluxes During the Spring Melt on the North Slope of Alaska

During the spring and summer of 2000, Pacific Northwest National Laboratory (PNNL) operated an acoustic anemometer near Barrow, Alaska and a dual wavelength scintillometer near Atqasuk, 100 km to the south. These systems were intended to measure surface heat fluxes as part of a continuing study by PNNL to explore the feedbacks between surface fluxes and low-level arctic stratus. Both systems were designed to operate unattended for many weeks and to report data via internet or telephone connections.

The acoustic anemometer was installed on a small, low island approximately 15 km northeast of Barrow across Elson Lagoon. As a result, the anemometer had an unobstructed ice or water fetch from most directions. The anemometer was mounted on a 10 m tower that was attached to a grounded barge. It was powered by a solar panel and battery array and transmitted its data directly to a logging computer via radio modem. The scintillometer was installed with a 200 m path over tundra that was characterized by scattered melt pools (after the spring melt) several meters across. The receiver and transmitter were mounted on pilings sunk into the permafrost, and the beam height was approximately 2.5 m.

The spring melt in Atqasuk preceded the melt in Barrow by about two weeks. Prior to the melt, both Barrow and Atqasuk experienced nearly identical diurnal variations in heat flux that reached typical maximum values of about 30 watts per square meter. Following the melt, the diurnal cycle of heat flux nearly disappeared from the Barrow site. The subsequent flux values remained near zero except for negative excursions when warm air from the land was advected over the cold lagoon. At Atqasuk, the melt was marked by an increasingly large amplitude of the diurnal heat flux cycle, with daily maximum fluxes reaching 200 watts per square meter. At the same time, both the mean temperature and its diurnal variation increased at Atqasuk, with Atqasuk becoming nearly 10 deg. C warmer than the Barrow site. We expect that these differences in boundary layer temperature and heat flux will both reflect and generate differences in cloud properties at the two sites. A parallel component of this project is using radiometric measurements to establish cloud properties. We will also compare these observations with the surface characterization of the ECMWF model to evaluate how well the model accounts for surface conditions and the implications of discrepancies on simulated cloud cover over the North Slope.