Tangent linear superparameterization of convection in a 10 layer global atmosphere with calibrated climatology. (24th April 2017)

• Record Type: Journal Article
• Authors: Kelly, Patrick; Mapes, Brian; Hu, I‐Kuan; Song, Siwon; Kuang, Zhiming
• Is Part Of: Journal of advances in modeling earth systems. Volume 9:Number 2(2017)
• Page Start: 932
• Page End: 948
• Publication Date: 2017-04-24
• Subjects: superparameterization -- convection -- matrix; Geological modeling -- Periodicals; Climatology -- Periodicals; Geochemical modeling -- Periodicals; 551.5011
• DOI: 10.1002/2016MS000871
• Languages: English
• ISSNs: 1942-2466

Abstract

Abstract: This paper describes a new intermediate global atmosphere model in which synoptic and planetary dynamics including the advection of water vapor are explicit in 10 layers, the time‐mean flow is centered near a realistic state through the use of carefully calibrated time‐independent 3‐D forcings, and temporal anomalies of convective tendencies of heat and moisture in each column are represented as a linear matrix acting on the anomalous temperature and moisture profiles. Currently, this matrix is Kuang's [2010 ] linear response function (LRF) of a cyclic convection‐permitting model (CCPM) in equilibrium with specified atmospheric cooling (i.e., without radiation or WISHE interactions, so it conserves column moist static energy exactly). The goal of this effort is to cleanly test the role of convection's free‐tropospheric moisture sensitivity in tropical waves, without incurring large changes of mean climate that confuse the interpretation of experiments with entrainment parameters in the convection schemes of full‐physics GCMs. When the sensitivity to free‐tropospheric moisture is multiplied by a factor ranging from 0 to 2, the model's variability ranges from: (1) moderately strong convectively coupled Kelvin waves with speeds near 20 m s −1 ; to (0) similar but much weaker waves; to (2) similar but stronger and slightly faster waves as the water vapor field plays an increasingly important role. Longitudinal structure in the model's time‐mean tropical flow is not fully realistic, and does change significantly with matrix‐coupled variability, but further work on editing the anomaly physics matrix and calibrating the mean state could improve this class of models. Key Points: A new intermediate global atmosphere model with tangent linear superparameterization of convection coupled to realistic 3‐D flow is described Increasing the free‐tropospheric moisture sensitivity in the matrix‐coupled model increases the amplitude of wave variability Despite strict linearity of the matrix, rectified time‐mean effects emerge in the GCM due to the coupling to nonlinear atmospheric dynamics