ALBERT

Description: Albedo – the reflectivity of a surface - is an important component in the energy budget, impacting the local to global climate. Data from nadir-viewing satellites can be combined with bidirectional reflectance distribution function (BRDF) data from multi-angular observation platforms to achieve realistic albedo values that acknowledge anisotropy. In my thesis, I evaluated how the land surface albedo varied on spatial and temporal scales during the snow-free period on Disko Island, Greenland. I examined how the albedo differed among the vegetation classes. Concerning the methodology, I assessed how the combination of MODIS BRDF data with Landsat 8 (L8) or Sentinel-2 (S2) influenced the albedo. The study area was located at the southern tip of Disko Island (69.27 °N, -53.47 °E) in West Greenland and covered a wetland and a range of tundra vegetation. I analysed automatic weather station (AWS) data from 2013 to 2022 and conducted mobile albedo measurements in August and September 2022 to examine the temporal and spatial variability. For the period from June to September 2022, I derived the L8 and S2 based albedo with inclusion of MODIS BRDF and narrow to broadband conversion and analysed their variability with regard to vegetation classes. In the snow-free period, the albedo increased from a monthly mean of 0.16 in June to 0.19 in September in the AWS data. The mobile measurements ranged from 〈 0.10 above bare soil and water to 〉 0.23 above areas dominated by lichen, Salix glauca or Equisetum arvense. The satellite-based albedo revealed temporally variable, significant correlations to normalised difference vegetation and moisture indices that reached values 〉 0.5 in the fen and wet heath class on several days. The albedo of shrubs was not notably smaller than other vegetation types but partly 0.01-0.05 above them in both the mobile measurements and the satellite-derived albedo. This finding challenges the assumption that shrubification causes climate forcing in all circumstances. The albedo of L8 and S2 differed to each other and the local data (root-mean-square error 0.04-0.14). The BRDF correction increased the albedo by 0.01 on average compared to nadir reflectance. L8 was better in reproducing the expected temporal and spatial variability of albedo than S2, which displayed less variability. S2 seemed to be more sensitive to atmospheric effects of haze and clouds influencing albedo. Thus, L8 seemed more suitable to calculate albedo in the study area. Though there were some methodological limitations, this thesis highlights aspects that should be considered when analysing albedo or jointly using L8 and S2 in high latitude regions.