5-Day-Wave Interactions with Tropical Precipitation in CMIP5 Models

• Published in: Journal of climate 2016-12, Vol.29 (23), p.8611-8624
• Main Authors: King, Malcolm J., Wheeler, Matthew C., Lane, Todd P.
• Format: Article
• Language: English
• Subjects: Anomalies; Atmospheric precipitations; Climate; Climate models; Computer simulation; Convection; Dynamics; Earth science; Geographical distribution; Gravitational waves; Hydrologic data; Interactions; Intercomparison; Loads (forces); Modes; Organizations; Precipitation; Precipitation anomalies; Precipitation data; Precipitation variations; Rain; Rainfall; Rainfall anomalies; Rossby waves; Rossby-Haurwitz waves; Simulation; Spatial distribution; Studies; Tropical climate; Tropical convection; Tropical rainfall; Tropospheric winds; Variance; Wave dynamics; Wave interactions; Wave propagation; Wind; Winds; Zonal winds
• ISSN: 0894-8755; 1520-0442
• DOI: 10.1175/JCLI-D-16-0190.1

The 5-day Rossby–Haurwitz wave is unlike other large-scale wave modes that interact with tropical rainfall in that associated rainfall presents as a modulation of localized areas of rainfall instead of propagating with the wave. This form of wave-modulated convective organization in climate models has received little attention. This study investigates the simulation of interactions between the 5-day wave and tropical convection in 30 models from phase 5 of the Coupled Model Intercomparison Project (CMIP5) and compares these with the interaction diagnosed from ERA-Interim and TRMM precipitation data. Models simulate the dry dynamics of the 5-day wave well, with realistic coherences between upper- and lower-tropospheric winds, as well as magnitudes and geographic distribution of wave wind anomalies being close to observations. The models consistently display significant coherences between 5-day-wave zonal winds and precipitation but perform less well at simulating the spatial distribution and magnitude of precipitation anomalies. For example, a third of the models do not reproduce significant observed anomalies near the Andes, and the best-performing model simulates only 38% of the observed variance over the tropical Andes and 24% of the observed variance over the Gulf of Guinea. Models with higher resolution perform better in simulating the magnitude of the Andean rainfall anomalies, but there is no similar relationship over the Gulf of Guinea. The evidence therefore suggests that the simulated interaction is mostly one way only, with the wave dynamics forcing the precipitation variations on the 5-day time scale.