Albedo parametrization for models of intermediate complexity

Abstract

How much solar radiation a surface absorbs depends on the reflective properties of the specified surface, snow and ice are two surfaces on Earth that have shown to have high albedos due to their reflective properties. To date, albedo schemes are limited to simple parameterizations and to high complexity (also referred to as general circulation) models, the latter being computationally expensive options that require high spatial and temporal resolution, making their usage limited. Models of intermediate complexity (EMICs) aim to be as accurate as general circulation models over similar time scales, and with less computational power. However, for albedo there is as of yet, no appropriate well-tested representation in EMICs. The objective of this thesis was to develop, and test a ‘new’ parametrization scheme for Earth’s seasonal surface albedo; for all terrestrial ice, seasonal snow, and marine ice surfaces. Key variables (wavelength, temperature, snow depth and time) were determined using theoretical schemes and existing data sets were looked at in order to validate the parametrization. However, as data were limited, it was only possible to test a few snow schemes: linearalbedo, linear-albedo for VIS/NIR bands, and polynomial-albedo. It was possible to adjust the polynomial scheme to consider VIS and NIR bands; the results conclude that the polynomial scheme that considers band albedos was the most successful scheme achieved.