This study investigates the response of the National Center for Environmental Prediction (NCEP) Regional Spectral Model (RSM) to the implementation of a new infrared radiation parameterization. The RSM can operate in hydrostatic and non-hydrostatic modes, and at a range of scales. The IR parameterization is from Chou et al. (1999), which has 10 IR spectral bands, and is based on a k-distribution approximation and random-maximum cloud overlapping treatment. Implementation of the Chou et al. parameterization in the NCEP Global Spectral Model (GSM) has been shown to yield improvements in that model.

This research seeks to determine the underlying factors governing the model response to the inclusion of the Chou et al. parameterization. To that end a set of sensitivity investigations are conducted to characterize the model response. The analysis concentrates on the tropical Pacific where the impact of the IR parameterization is most pronounced, and on interactions among cloud, radiative and dynamic responses. A radiative-convective model will also be employed to further probe the impact of the Chou et al. parameterization on these processes.

Observed data from the Atmospheric Radiation Measurement (ARM) site on Manus Island, Papua New Guinea in the tropical western Pacific (TWP) are used in this study. This data includes a comprehensive set of cloud, radiation, and meteorological data. Radiation and atmospheric measurements from ARM are used to evaluate the Chou et al. parameterization. The ARM measurements are also used to constrain the sensitivity analysis.