Underrepresentation of Ubiquitous Opaque and Transmissive Arctic Atmospheric States in Modern Reanalyses

Abstract The high latitudes exhibit two distinct and recurring radiative states, often referred to as “transmissive” and “opaque.” These states, characterized by bimodality in surface longwave flux distributions, have a large influence on the surface energy balance and are ubiquitous across the Arctic. Since reanalyses are frequently used to supply surface radiative fluxes for Arctic applications, it is important to assess the extent to which reanalyses represent these states. This study uses satellite observations to evaluate how well the Modern-Era Retrospective Analysis for Research and Applications, version 2 (MERRA-2), the fifth major global reanalysis produced by the European Centre for Medium-Range Weather Forecasts (ERA5), and the Arctic System Reanalysis (ASR), version 2, capture these two modes. Each reanalysis demonstrates deficiencies in reproducing observed distributions of cloud radiative effect, often resulting in a noticeable absence of bimodality in surface longwave flux distributions. The ASR exhibits some improvement over the two global reanalyses, underscoring the value of regional reanalyses that more accurately resolve cloud processes. Idealized cloud forcing calculations provide insight into why these states are prevalent in the Arctic and indicate that too much ice at the expense of liquid water largely explains the absence of the opaque mode in ASR compared to observations. Given the non-Gaussian nature of polar longwave flux distributions, some traditional model evaluation methods, especially those relying on monthly means, likely miss errors in the representation of these states entirely. Thus, we urge model evaluators to consider the existence of distinct states in their assessments and, more generally, to adopt evaluation methods that can assess non-Gaussian distributions. Significance Statement Radiation plays an important role in the Arctic climate by modulating the surface temperature and the rate at which ice forms. Observations show that the Arctic alternates between periods with a large energy deficit (transmissive) and a small energy deficit (opaque). Due to the limited number of observation stations in the Arctic, models are often used to fill gaps where direct measurements are unavailable. These observation-assisted models are called “reanalyses.” This study assesses how well reanalyses capture the two dominant Arctic states. Understanding whether these states are accurately represented in well-constrained models may provide insights into our ability to predict the two states’ future evolution.