Impact of multi-mode and multi-species aerosols on 1D snow simulation at observational sites distributed at different latitudes.

Snow plays a critical role in energy budget by reflecting a significant portion of incoming solar radiation, thereby influencing local and global climate dynamics. However, the state-of-the-art climate models still face challenges to simulating global snow amount partly due to inadequate representation of snow albedo. Current models predominately parameterize snow albedo as an age-dependent, exponentially decaying function, which oversimplify its complexity. Also, most of these models neglect the deposition of aerosols (such as dust, black and organic carbons) and their ability of absorbing visible part of solar radiation, leading to reduced albedo and accelerated snowmelt. This “snow darkening effect” process is essential for improving the transient simulation of snow for climate and enhancing our understanding of climate feedback mechanism. To incorporate this phenomenon in ORCHIDEE, the land surface component of IPSL’s Earth System Model, we have implemented a comprehensive tracer framework that simulate the deposition and vertical transport of four log-normal modes of dusts, hydrophobic and hydrophilic black and organic carbons within snowpack. In order to enhance the snow aging processes, a snow metamorphism approach has been used that explicitly simulates the physical evaluation of snow optical diameter and sphericity, rather than relying on a simple chronological aging parametrization. To replace the empirically decaying albedo parametrization with a physics-based impure snow albedo, we have employed unique combination of Warren-Wiscombe’s uni-directional snow radiative transfer scheme with online optical property calculations of snow using Khokhanovsky’s scheme and mie-theory based offline aerosol optical properties. This enhanced physical representation of snow albedo dynamics. For validation against observation, offline ORCHIDEE simulations are conducted using in-situ meteorological forcing and MERRA-2 reanalysis aerosol deposition data across observation sites localized in different climatic areas over the Earth. These sites are selected to represent different aerosols regimes, each characterized by distinct dominant aerosol species. In these simulations, as snowpack develops seasonally, it harnesses aerosols deposited on the surface which are subsequently buried by additional snowfall and redistributed during melt-refreeze cycles. Consequently, snow albedo fluctuates, starting at high values following fresh snowfall and decreasing gradually due to increase in snow optical diameter (metamorphism) and accumulation of impurities, influenced by snow liquid content, vertical temperature gradient, aerosol species and deposition rate. The buried aerosols act as a memory and re-emerge at the surface in high concentration during the melting season. This re-exposure further reduces snow albedo, thereby accelerating melt rates. This simulated behavior is validated against in-situ observation of surface aerosol concentration and snow albedo. Through sensitivity experiments isolating the effects of different modes of dusts and other species, we further identified non-linear dynamics that critically influence the timing of snow melt and the end of the snow season.