The Physical Representation of Bare Ice Albedo in Radiative Transfer Models and the Implications on Greenland Ice Sheet Albedo and Surface Mass Balance

Abstract

Accurate modeling of cryospheric surface albedo is essential for our understanding of climate change as snow and ice surfaces regulate the global radiative energy budget and sea level through their albedo and mass balance. Although significant progress has been made using physical principles to represent the dynamic albedo of snow, models of land ice albedo tend to be heavily parameterized and not explicitly connected with physical properties that govern albedo, such as the number and size of air bubbles, specific surface area, and presence of abiotic and biotic light absorbing constituents (LACs). The lack of physically based and computationally efficient ice albedo models has led to unrealistic bare ice albedo representations in Earth System Models (ESMs). For example, many ESMs prescribe a constant albedo over ice surfaces. However, it is increasingly important that ESMs capture the spatially, temporally, and spectrally varying ice albedo as polar temperatures are rapidly increasing, and more bare ice is being exposed. The Greenland Ice Sheet (GrIS) is currently the largest cryospheric contributor to increasing sea levels, and a significant portion of GrIS surface melt is due to dark ice regions along the edge of the ice sheet, where solar absorption is influenced by the ice albedo. The work presented in this thesis (1) improves our ability to simulate bare ice albedo using physical and optical properties of ice surfaces, (2) incorporates those improvements in an ESM to quantify the contribution of exposed bare ice to the GrIS surface mass balance, and (3) quantifies the relative impact of three different LACs on the GrIS ablation zone melt rates. First, I introduce a single column cryospheric radiative transfer model that accurately simulates the albedo of snow and ice using their physical and optical properties (SNICAR-ADv4). SNICAR-ADv4 compares well to in-situ measurements of snow and ice albedo. SNICAR-ADv4 is applied to ice on the GrIS in the Exascale Earth System Model (E3SM) using MODIS observations to constrain the physical properties of bare ice surfaces. The SNICAR-ADv4 bare ice albedo scheme improves the representation of ice albedo in E3SM. Compared to MODIS albedo measurements, the original ice albedo in E3SM overestimates bare ice albedo by ~4% in the visible and ~7% in the near-infrared wavelengths, while the SNICAR-ADv4 enabled E3SM bare ice albedo is within an error of 0.7% of MODIS. SNICAR-ADv4 in E3SM also reduces simulated surface mass balance (by increasing surface melt) on the GrIS by ~145 Gt, or 0.4mm of sea-level equivalent, over a 20-year historical simulation, compared to the default E3SM albedo scheme. SNICAR-ADv4 in E3SM is then used with multiple satellite observation streams in a feasibility study focused on quantifying the impact of black carbon, mineral dust, and glacier algae on GrIS bare ice albedo and surface melt. This work provides a framework for utilizing hyperspectral imagery and ice modeling techniques to infer the impact of different LACs. The results indicate all three LACs have a non-negligible effect on ice albedo and surface mass balance. It also highlights key areas where further research and measurements are needed. The work presented in this dissertation has improved our ability to simulate bare ice albedo and quantify the implications of increasing bare ice exposure in our warming climate system. It will also motivate future work on developing modeling techniques to prognostically determine surface ice properties.