Arctic climate has undergone changes in cloud amount (fraction), cloud particle size, cloud particle phase, surface temperature, surface broadband albedo, and radiation fluxes over the 18-year period 1982-1999 as revealed by the Advanced Very High Resolution (AVHRR) Polar Pathfinder (APP) data set. Results show that the Arctic surface has warmed and become more cloudy in spring and summer, but has cooled and become less cloudy in winter. The decadal rate of annual surface temperature change is 0.54oC for the area north of 60oN. The surface broadband albedo has decreased significantly in autumn, especially over the ocean, indicating a later freeze-up. The decadal rate of annual surface broadband albedo is -1.4% (absolute). Cloud amount has decreased at a rate of 5% (absolute) per decade in the winter, and increased at a rate of 2-4% per decade during the spring and summer, but on an annual time scale there is no trend. All of the trends are statistically significant at confidence levels of 92% or higher. During spring and summer, changes in sea ice albedo and extent that result from surface warming tend to modulate the radiative effect of increasing cloud cover resulting in no significant trend. On an annual scale, the net cloud forcing has decreased at a decadal rate of -3.56 W/m2 indicating an increased cooling by clouds. The combined effect of changes in these parameters is such that there is no significant trend in the net surface radiation flux except in winter when the net longwave radiation has decreased over the 18-year period as a result of decreasing cloud cover. There are large correlations between the Arctic Oscillation (AO) index and both surface temperature and cloud amount for some areas, though the sign of the relationship varies from region to region. Over Greenland, surface temperature and cloud fraction are negatively correlated with the AO index, while both parameters are positively correlated with the AO index over northern Europe. Through an eigenvector analysis (EOF) of the surface temperature anomaly as well as the cloud amount anomaly, it is found that the first two eigenvectors account for about half of the total variance in surface temperature. Eigenvector analysis also reveals spatial relationships such as an inverse surface temperature variation between Greenland and northern Europe in winter that may indicate an internal adjusting mechanism corresponding to the overall warming in the Arctic.

Keywords: Arctic climate change, Arctic Oscillation (AO) index, AVHRR, EOF,CASPR, APP Data Set, Trend study, Global Warming.