Dissertation Defense: Lili Manzo

Title: Accurate Treatment of Surface Longwave Radiation in Earth System Models

Abstract: Surface cooling and atmospheric warming are driven by an exchange of longwave (LW) radiation. Thus, it is crucial that Earth system models (ESMs) accurately simulate LW fluxes. This dissertation develops more accurate algorithms to treat LW emission and reflectance in ESMs, resolves physical inconsistencies between ESM model components, and demonstrates the climate impacts resulting from these improvements.

Many ESMs approximate surface emissivity as constant, or broadband, across wavelengths. However, both surface emissivity and atmospheric longwave absorption vary strongly with wavelength. The first study of this dissertation quantifies the instantaneous bias in LW flux and heating rate produced by the common broadband emissivity approximations over snow, ice, and water surfaces. I find that surface flux bias increases with surface temperature and decreases with atmospheric opacity, up to 1.60 W/m2 (0.52% of the total). I also devised a novel method of representing surface emissivity that can reduce flux and heating rate bias by up to 99%.

My second study tested this novel emissivity representation in a fully coupled ESM (E3SM). I implemented this method over the ocean surface and generated a 30-year climatology to quantify the resulting climate impacts. Global mean upwelling surface LW flux is decreased by 1.38 W/m2, which then modifies LW heating rates. This method eliminates a 0.24 K temperature bias, is more physically realistic than current broadband methods, and is more computationally efficient than the commonly proposed high-spectral solution.

My third study resolved an energy leak present in many ESMs between their surface and atmosphere components. I found that this inconsistency stems from an imprecise approximation of the Stefan-Boltzmann constant σ, which controls emitted flux. By updating σ, I reduced an energy leak of 0.024 W/m2 by 97%. This leads to a decrease in ocean heat content, which declines at a rate of ~54 petajoules per century. Improving σ has implications for model spin-up simulations and future Coupled Model Intercomparison Projects. Overall, this thesis provides more accurate treatments of LW radiation for ESMs, and thereby provides a demonstrated path to improve climate simulations and future projections.