No abstract

No abstract

A technique for the computation of longwave radiative quantities using the line-by-line approach has been developed in the Soviet Union. The method has been applied to obtain fluxes and cooling rates for standard atmospheric profiles used in the Intercomparison of Radiation Codes Used in Climate Models (ICRCCM) sponsored by the World Meteorological Organization. The sensitivity of the result to changes in the vertical quadrature scheme, the angular integration, and the spectral line shape is evaluated. Fluxes and cooling rates in the troposphere are in general agreement with those obtained with different line-by-line models. Results from parameterized models, including a wideband statistical model and one employing the integral transmission function, have been compared to the line-by-line results. Flux errors in the simplified schemes are of the order of 10 W/m 2. The sensitivity of these models to changes in atmospheric profiles, or to an increase in CO2 amount, is similar to that of the line-by-line calculations. INTRODUCTIONRadiative heat exchange is a basic component of atmospheric energy transfer. The numerical models for weather forecasting, climate theory, cloud generation, and a number of mesoscale dynamic processes all invoke radiation codes to compute this quantity.The radiative flux calculation algorithms in these models differ appreciably from each other and are always an approximation to the exact solutions of the radiative transfer equations. These approximations are needed because (1) the radiative calculations have to be made quickly compared to the calculation time of the model as a whole; (2) there is insufficient information in dynamic models to calculate the fluxes, and (3) the spatial, particularly the vertical, resolutions of the models are not sufficiently detailed.The calculated fluxes in each model differ as a consequence of the distinctiveness of the specific algorithms, the initial data applied, the spectral resolutions, etc., of the model. We may thus ask how the results of the simulation of the dynamic processes in the atmosphere are affected by the differences in radiative flux calculation techniques in models. The second principal feature of this technique is the shift in frequency of the line centers to the nearest frequency points (nodal points) at which the absorption coefficients are evaluated. This approach permits the use of a uniformly spaced frequency grid for the calculation, and permits substantial improvement in both the accuracy and the computational efficiency of the numerical frequency integration. A deficiency in the direct application of (1)-(7) to flux computations is that, according to (2), a systematic underestimation of absorption occurs for "weak" lines, i.e., for the lines whose contributions to e(v) are less than e. In certain spectral regions, particularly the 8-to 13-/am "window" region, this underestimation can markedly affect the accuracy of the calculation. In order to compensate for this effect, the simple technique given below is used.If the ith line ...

No abstract

No abstract