Considering the radiative effects of snow on tropical Pacific Ocean radiative heating profiles in contemporary GCMs using A‐Train observations

Abstract

This study characterizes biases in water vapor, dynamics, shortwave (SW) and longwave (LW) radiative properties in contemporary global climate models (GCMs) against observations over tropical Pacific Ocean. The observations are based on Atmospheric Infrared Sounder for water vapor, CloudSat 2B‐FLXHR‐LIDAR for LW and SW radiative heating profiles, and radiative flux from Clouds and the Earth's Radiant Energy System products. The model radiative heating profiles are adopted from the coupled and uncoupled National Center for Atmospheric Research (NCAR) Community Earth System Model version 1 (CESM1) and joint Year of Tropical Convection (YOTC)/Madden Julian Oscillation (MJO) Task Force‐Global Energy and Water Cycle Experiment Atmospheric System Studies (GASS) Multi‐Model Physical Processes Experiment (YOTC‐GASS). The results from the model evaluation for YOTC‐GASS and NCAR CESM1 demonstrate a number of systematic radiative biases. These biases include excessive outgoing LW radiation and excessive SW surface radiative fluxes, in conjunction with a radiatively unstable atmosphere with excessive LW cooling in the upper troposphere over convectively active areas, such as the Intertropical Convergence Zone/South Pacific Convergence Zone (ITCZ/SPCZ) and warm pool. Using sensitivity experiments with the NCAR‐uncoupled/NCAR‐coupled CESM1, we infer that these biases partly result from the interactions between falling snow and radiation that are missing in most contemporary GCMs (e.g., YOTC‐GASS, Coupled Model Intercomparison Project 3 (CMIP)3, and Atmospheric Model Intercomparison Project 5 (AMIP5)/CMIP5). A number of biases in the YOTC‐GASS model simulations are consistent with model biases in CMIP3, AMIP5/CMIP5, and NCAR‐uncoupled/NCAR‐coupled model simulation without snow‐radiation interactions. These include excessive upper level convection and low level downward motion with outflow from ITCZ/SPCZ. This generates weaker low‐level trade winds and excessive precipitation in the Central Pacific Trade wind regions. The excessive LW radiative cooling in NCAR‐coupled/NCAR‐uncoupled GCM simulations is reduced by 10–20% with snow‐radiative effects considered.

Improving climate projections by understanding how cloud phase affects radiation

“…atmosphere (TOA). Similarly, a warmer atmosphere produced by increased greenhouse gas concentrations would contain more liquid and less ice, and therefore induce changes in radiation fluxes that would provide a negative feedback to warming [Li andLe Treut, 1992, Senior andMitchell, 1993]. The strength of this feedback, however, depends on the distribution of cloud phase in the present atmosphere, which was, until recently, poorly constrained.…”

Impact of Cloud Ice Particle Size Uncertainty in a Climate Model and Implications for Future Satellite Missions

“…The radiative effect of falling snow is explicitly considered in CAM5. Some previous studies suggested the paramount importance of the snow radiative effect in reproducing the present-day radiative heating profiles and even polar sea ice in GCM (Li et al, 2016(Li et al, , 2017. Therefore, we repeat all the parameter perturbation experiments for the particle size of falling snow (R es ) in the same way with the those for R ei .…”

Falling Snow Radiative Effects Enhance the Global Warming Response of the Tropical Pacific Atmosphere

“…Thanks to the steady improvement of model physics and resolution during recent decades, current state‐of‐the‐art coupled general circulation models (CGCMs) are increasingly capable of reproducing the fundamental features of the tropical atmosphere over the Pacific Ocean with fidelity (e.g., AchutaRao & Sperber, ; Dai, ; Flato et al, ; Guilyardi et al, ; Li, Lee, Waliser, David Neelin, et al, ; Li, Lee, Waliser, Stachnik, et al, ; Meehl et al, ; Pincus et al, ; Randall et al, ). However, some biases remain, including an excessive westward extension of the equatorial Pacific cold tongue in most Coupled Model Intercomparison Project phase 3 (CMIP3) and CMIP5 models (e.g., Dai, ; Bernie et al, ; Ham et al, ; Li & Xie, ; Li & Xie, ; Zheng et al, ), with relatively warmer sea surface temperatures (SSTs) on the flanks of the Intertropical Convergence Zone (ITCZ), meaning a weaker zonal SST gradient over the equatorial Pacific (Li, Lee, Waliser, David Neelin, et al, ; Li, Lee, Waliser, Stachnik, et al, ; Li et al, , ; Li & Xie, ; Lin, ; Zhang et al, ). This is linked to weaker surface wind strength and low‐level winds over the trade‐wind regions and an inaccurate precipitation spatial pattern‐the so‐called double ITCZ, with excessive precipitation off the equator, insufficient rainfall over the equatorial Pacific, and an overly zonal orientation of the South‐Pacific Convergence Zone (SPCZ; e.g., Brown et al, ; Li, Lee, Waliser, David Neelin, et al, ; Li & Xie, ; Lin, ).…”

“…One of the persistent systematic biases commonly found over the tropical Pacific in most CMIP3 and CMIP5 CGCMs is related to radiative fluxes (Li et al, ; Li, Lee, Waliser, David Neelin, et al, ; Li et al, ; Waliser et al, ). Namely, over convective regions such as the western Pacific warm pool, ITCZ, and SPCZ, there tends to be excessive outgoing longwave radiation (OLR) and underestimated upward shortwave radiation (RSUT) at the top of the atmosphere, with overestimated downward shortwave radiation at the surface.…”

“…In most CMIP5 models, only radiative interactions with suspended cloud mass are considered while the precipitating and convective‐core cloud mass is excluded. However, snow radiative effects contribute considerably to the observed radiative balance and heating rates (Li et al, , ).…”