Use of second order statistics of observed and synthetic outgoing long-wave radiation spectra datasets for testing Global Circulation Models

A part of the study of future climate changes is based on the forecast provided by Global Circulation Models (GCM). Testing GCM output for the past is therefore a major issue for climate studies. Different approaches are possible, based on comparison between numerical output and atmospheric and oceanic measurements. More recently, another approach has been proposed, which makes use of direct observations, such as outgoing long-wave radiance, instead of retrieved products. In order to accomplish this goal, it is necessary to obtain an equivalent set of data from numerical models and observations. High-resolution spectrally resolved outgoing long-wave radiance from the Earth-Atmosphere system has been measured in the last decades by satellite-borne interferometers. On the model side, we built an equivalent synthetic dataset by processing the output of a GCM with a Radiative Transfer Model (RTM) code. As suggested by, by comparing first and second order statistics from the synthetic and measured datasets of outgoing long-wave radiance spectra we are able to test the performances of a GCM in describing not only the main behaviour of the Earth-Atmosphere system, but also its variability and climate sensitivity.