Impact of poleward heat and moisture transports on Arctic clouds and climate simulation

Eun-Hyuk Baek,Baek-Min Kim,Joo-Hong Kim,Sungsu Park,Jee-Hoon Jeong

doi:10.5194/acp-20-2953-2020

Abstract

Abstract. Many general circulation models (GCMs) have difficulty simulating Arctic clouds and climate, causing substantial inter-model spread. To address this issue, two Atmospheric Model Intercomparison Project (AMIP) simulations from the Community Atmosphere Model version 5 (CAM5) and Seoul National University (SNU) Atmosphere Model version 0 (SAM0) with a unified convection scheme (UNICON) are employed to identify an effective mechanism for improving Arctic cloud and climate simulations. Over the Arctic, SAM0 produced a larger cloud fraction and cloud liquid mass than CAM5, reducing the negative Arctic cloud biases in CAM5. The analysis of cloud water condensation rates indicates that this improvement is associated with an enhanced net condensation rate of water vapor into the liquid condensate of Arctic low-level clouds, which in turn is driven by enhanced poleward transports of heat and moisture by the mean meridional circulation and transient eddies. The reduced Arctic cloud biases lead to improved simulations of surface radiation fluxes and near-surface air temperature over the Arctic throughout the year. The association between the enhanced poleward transports of heat and moisture and increase in liquid clouds over the Arctic is also evident not only in both models, but also in the multi-model analysis. Our study demonstrates that enhanced poleward heat and moisture transport in a model can improve simulations of Arctic clouds and climate.

Highlights

With the increasing amounts of greenhouse gases, the Arctic has undergone the most rapid warming of any location on Earth
Park et al (2019) showed that the global mean climate, 20th century global warming, and El Niño and Southern Oscillation simulated by SAM0 are roughly similar to those of Community Atmosphere Model version 5 (CAM5) and the Community Earth System Model version 1 (CESM1; Hurrell et al, 2013); SAM0 substantially improves the simulations of the Madden–Julian Oscillation (MJO) (Madden and Julian, 1971), diurnal cycle of precipitation, and tropical cyclones, all of which are known to be extremely difficult to simulate in general circulation models (GCMs)
Many GCMs suffer from cold bias over the Arctic, which has been speculated to be caused by radiation biases associated with cloud fraction and cloud liquid mass underestimation over the Arctic

Summary

Introduction

With the increasing amounts of greenhouse gases, the Arctic has undergone the most rapid warming of any location on Earth. The warming rate of the nearsurface air temperature over the Arctic has been 2 to 3 times that of the entire globe (Johannessen et al, 2016; Screen and Simmonds, 2010; Serreze and Barry, 2011). This pronounced Arctic temperature amplification, some of which is forced by the positive feedbacks among various climate components (e.g., sea ice–albedo feedback, Deser et al, 2000; water vapor and cloud feedback, Lu and Cai, 2009; and lapse-rate feedback, Pithan et al, 2014), is responsible for extreme weather and climate events over midlatitude continents (Kug et al, 2015; Screen and Simmonds, 2013; Wu and Smith, 2016). Many researchers have reported that the GCM-simulated cold biases over the Arctic are associated with the shortwave (SW) and longwave (LW) radiation biases at the surface, which are due to poor simulation of Arctic clouds (Barton et al, 2014; English et al, 2015; Karlsson and Svensson, 2013; Shupe and Intrieri, 2004)

Methods

Results

Conclusion