Evaluation of Tropical Convection Regimes in the HadGEM2 Model Using Radio Occultation Observations

The evaluation of Global Climate Models (GCMs) requires high-quality observational data sets. Global Positioning System (GPS) Radio Occultation (RO) measurements, which are available continuously from various satellite missions since 2001, provide a global and independent data record for the upper troposphere and lower stratosphere (UTLS). The measurements are based on precise atomic clocks and feature long-term stability, homogeneity, and accuracy. Best data quality is achieved at about 5 km to 30 km altitude and observational errors are well characterized. Profiles of key climate variables, such as refractivity, pressure, geopotential height, and temperature, are retrieved at a high vertical resolution of about 0.5 km in the troposphere to 1.5 km in the stratosphere. Due to these characteristics, RO qualifies as climate benchmark data type, which can be used for the evaluation of climate models. In this presentation we focus on physical processes relevant to tropical water vapor/lapse rate feedback. Water vapor and lapse rate responses are closely coupled in the upper troposphere and models with larger negative lapse rate feedback also have a larger positive water vapor feedback. We show first results on the representation of UTLS temperature lapse rates and refractivity gradients in RO observations and in Met Office Hadley Centre HadGEM2 model data. To allow a consistent and process-oriented approach, we classify moist and dry tropical regimes through distinction between dynamical up- and downdraft regions. The pressure vertical velocity at 500 hPa (ω₅₀₀) and surface temperature (T) from ERA-Interim (for RO profiles) and from the HadGEM2 model itself (for the model) is used for the classification of these regimes. RO data and co-located HadGEM2 model profiles are sorted and sampled into vertical velocity classes and a systematic comparison is performed. A better insight into observation and model behavior in tropical convection regimes will contribute to the development of improved model parameterizations.