Evaluation of the variation in albedo over snow-covered forest in Northern Sweden and Finland

Abstract

The effect of albedo variability induced by meteorological parameters in snow-covered forested areas at local scale is poorly known. Few meteorological stations measure albedo, even though the effect on albedo caused by land use changes is well known. The boreal forest is mainly comprised of coniferous tree species and represents the largest terrestrial biome which is characterised by long winters with seasonal snow cover. Since boreal forests alter the albedo through snow interception by the canopy, they are subjects of this investigation. This study evaluates the induced effect of meteorological variations and forest metric variations on albedo fluctuations at three different forested sites in the boreal zone during winter season. All study sites belong to the Integrated Carbon Observation System (ICOS) network where meteorological variables are measured from towers on the three individual study sites named Svarberget (SE-Svb), Norunda (SE-Nor) and Hyytiälä (FI-Hyy). SE-Svb and SE-Nor is located in the boreal forest of Sweden while FI-Hyy is located in the boreal forest of Finland. The influences of several meteorological parameters were investigated by use of correlation analysis first and then regression analysis. Forest metrics were derived from Light Detection And Ranging (LiDAR) data and further related to mean albedo. The influence of temperature on albedo was seen at every study site investigated in this study. When the temperature is below freezing and the albedo is initially high (0.3), a decrease in albedo to a value of around 0.1 is observed when the temperature increases to above freezing. This relation is ascribed to the temperature influence of snow properties and snow appearance which further is connected to albedo. Apart from this, precipitation was related to albedo but the strength of the relationship was difficult to interpret since the correlation analysis demonstrated both a positive and negative connection to albedo. This could possibly derive from the form of the precipitation, snow or rain, and further be temperature dependent. However, snow depth and increased snow amount were parameters that induced an increase in albedo, although the relationship was not that prominent (varied between insignificant up to correlation coefficient of 0.51 in the correlation analysis). The use of forest structural metrics, canopy density, tree height and Canopy Relief Ratio (CRR), did not show any clear relationship to mean albedo. However, the variable model output from the regression analysis highlights that there is a need for site-specific investigation to better understand local albedo variability in snow-covered forested areas.