The representation of the TTL in a tropical channel version of the WRF model

Abstract

[1] In this study, the Weather Research Forecast (WRF) model is used to investigate key physical processes controlling the Tropical Tropopause Layer (TTL) temperature and water vapor distributions in December 2005 to January–February 2006. The model domain is configured as a tropical channel with a horizontal grid spacing of 36 km, a vertical grid spacing of 500 m, and a top at 0.1 hPa. Initial and boundary conditions are set using the ERA-Interim reanalysis data set. An ozone distribution computed from satellite and ozonesonde measurements is used for radiative forcing calculations. The model's ability to replicate observed TTL temperatures is evaluated via comparisons with radiosonde data and reanalyses (MERRA and ERA-Interim). The Microwave Limb Sounder (MLS) water vapor measurements are used to evaluate WRF-simulated water vapor in the TTL. Results of the simulations show that the model reproduces the mean temperature and its variability above 50 hPa as well as the tropical tropopause height. However, the model cold point tropopause temperature is colder than the reanalyses by ~1.2 K. The model captures the location of TTL water vapor minimum in the Western Pacific but is drier than the MLS observations in the TTL. To assess possible reasons for the tropopause temperature discrepancy, an additional WRF experiment was conducted using analysis nudging for water vapor. This experiment produces more tropical cirrus clouds in the upper troposphere and a warming of ~1.5 K of the cold point tropopause. This suggests that the radiative effects of cirrus clouds and water vapor must be considered for accurate temperature simulations in the TTL.