Impact of the atmospheric thickness on the atmospheric downwelling longwave radiation and snowmelt under clear-sky conditions in the Arctic and Subarctic

Studies show that the energy available to melt snow at high latitudes is al most exclusively provided by radiation. Solar energy determines the period of possible snowmelt, while downwelling atmospheric longwave radiation modi fies the timing and triggers the onset of snowmelt. Atmospheric thickness, defined as the vertical distance between the 500- and 1000- mb pressure sur faces, is directly related to the mean temperature and water vapor path of an atmospheric layer and thus has a direct influence on the downwelling lon gwave radiation and snowmelt. A comprehensive radiative transfer model was applied to calculate the downwelling longwave radiation to the snow surface over the period of snowmelt from 1980 through 1991 using radiosonde data o btained at Barrow and McGrath, Alaska, under clear- sky conditions. The res ults indicate that the atmospheric thickness has a positive impact on downw elling longwave radiation, which ranges from about 130 W m(-2) for an atmos pheric thickness of 4850 m to about 280 W m(-2) for an atmospheric thicknes s of 5450 m. This study demonstrates that atmospheric water vapor path has a greater impact on atmospheric downwelling longwave radiation to the snow surface than the mean atmospheric temperature. This study also indicates th at when the near- surface air temperature is used to infer downwelling long wave radiation, significant errors can occur. Thus, compared with the resul ts obtained from the atmospheric radiative transfer model, the empirical fo rmula due to Parkinson and Washington underestimates the downwelling longwa ve radiation when the near- surface air temperature is relatively high and overestimates it when the near- surface air temperature is relatively low. Investigations of the relationship between the atmospheric thickness and th e snowmelt onset were conducted. Results indicate that for the period from 1980 through 1991, an atmospheric thickness of 5250 m at Barrow and 5200 m at McGrath in Alaska was sufficient to trigger the onset of snowmelt. The d ifference in the threshold values of the atmospheric thickness may be due t o differences in the atmospheric structure and different contributions of o ther energy sources such as sensible and latent heat to melt snow. This stu dy also demonstrates that snow cover disappears earlier during warm and wet (higher atmospheric temperature and precipitable water path, and greater a tmospheric thickness) springs and later during cold and dry (lower atmosphe ric temperature and precipitable water path, smaller atmospheric thickness) springs. Atmospheric precipitable water path has a greater impact on snowm elt than the mean atmospheric temperature. Generally, higher atmospheric te mperature is correlated with higher atmospheric water vapor path and since atmospheric temperature is closely coupled to the atmospheric water vapor p ath in the Arctic and Subarctic and since it can be obtained through routin e numerical weather prediction models, the atmospheric thickness may be use d as a reliable indicator of regional- scale snowmelt in the Arctic and sub arctic.