We enhance the Community Land Model (CLM) snow albedo modeling by implementing several new features with more realistic and physical representations of snow-aerosol-radiation interactions. Specifically, we incorporate the following model enhancements: (1) updating ice and aerosol optical properties with more realistic and accurate datasets, (2) adding multiple dust types, (3) adding multiple surface downward solar spectra to account for different atmospheric conditions, (4) incorporating a more accurate adding-doubling radiative transfer solver, (5) adding nonspherical snow grain representation, (6) adding black carbon-snow and dust-snow internal mixing representations, and (7) adding a hyperspectral (480-band versus the default 5-band) modeling capability. These model features/enhancements are included as new CLM physics/namelist options, which allows for quantification of model sensitivity to snow albedo processes and for multi-physics model ensemble analyses for uncertainty assessment. The model updates will be included in the next CLM version release. Sensitivity analyses reveal stronger impacts of using the new adding-doubling solver, nonspherical snow grains, and aerosol-snow internal mixing than the other new features/enhancements. These enhanced snow albedo representations improve the CLM simulated global snowpack evolution and land surface conditions, with reduced biases in simulated snow surface albedo, snow cover, snow water equivalent, snow depth, and surface temperature, particularly over northern mid-latitude mountainous regions and polar regions.

Snow and ice albedo reduction due to deposition of absorbing particles (i.e., snow darkening effect (SDE)) warms the Earth system and is largely attributed to black carbon (BC) and dust. Absorbing organic aerosol (BrC) also contributes to SDE but has received less attention due to uncertainty and challenges in model representation. This work incorporates the SDE of absorbing organic aerosol (BrC) from biomass burning and biofuel sources into the Snow Ice and Aerosol Radiative (SNICAR) model within a variant of the Community Earth System Model (CESM). Additionally, 12 different emission regions of BrC and BC from biomass burning and biofuel sources are tagged to quantify the relative contribution to global and regional SDE. BrC global SDE (0.021–0.056 Wm-2) is larger than other model estimates, corresponding to 37%–98% of the SDE from BC. When compared to observations, BrC simulations have a range in median bias (-2.5%–+21%), with better agreement in the simulations that include BrC photochemical bleaching. The largest relative contributions to global BrC SDE are traced to Northern Asia (23%–31%), Southeast Asia (16%–21%), and South Africa (13%–17%). Transport from Southeast Asia contributes nearly half of the regional BrC SDE in Antarctica (0.084–0.3 Wm-2), which is the largest regional input to global BrC SDE. Lower latitude BrC SDE is correlated with snowmelt, in-snow BrC concentrations, and snow cover fraction, while polar BrC SDE is correlated with surface insolation and snowmelt. This indicates the importance of in-snow processes and snow feedbacks on modeled BrC SDE.

The Ice, Cloud, and Land Elevation Satellite-2 (ICESat-2) mission has collected global surface elevation measurements for over three years. ICESat-2 carries the Advanced Topographic Laser Altimeter (ATLAS) instrument, which emits laser light at 532 nm, and ice and snow absorb weakly at this wavelength. Previous modeling studies found that melting snow could induce significant bias to altimetry signals, but there is no formal assessment on ICESat-2 acquisitions during the Northern Hemisphere melting season. In this work, we performed two case studies over the Greenland Ice Sheet to quantify volumetric scattering in ICESat-2 signals over snow. Elevation data from ICESat-2 was compared to Airborne Topographic Mapper (ATM) data to quantify bias. We used snow optical grain sizes derived from ATM and the Next Generation Airborne Visible/Infrared Imaging Spectrometer (AVIRIS-NG) to attribute altimetry bias to snowpack properties. For the first case study, the mean optical grain sizes were 340±65 µm (AVIRIS-NG) and 670±420 µm (ATM), which corresponded with a mean altimetry bias of 4.81±1.76 cm in ATM. We observed larger grain sizes for the second case study, with a mean grain size of 910±381 µm and biases of 6.42±1.77 cm (ICESat-2) and 9.82±0.97 cm (ATM). Although these altimetry biases are within the accuracy requirements of the ICESat-2 mission, we cannot rule out more significant errors over coarse-grained snow, particularly during the Northern Hemisphere melting season.