Some important LBLRTM attributes are as follows:

the Voigt line shape is used at all atmospheric levels with an algorithm based on a linear combination of approximating functions;

it has been and continues to be extensively validated against atmospheric radiance spectra from the ultra-violet to the sub-millimeter;

it incorporates the self- and foreign-broadened water vapor continuum model, [Mlawer et al., 2003; MT_CKD_1.0] as well as continua for carbon dioxide, and for the collision induced bands of oxygen at 1600 cm^-1 and nitrogen at 2350 cm^-1.;

all parameters on the HITRAN line database are used including the pressure shift coefficient, the halfwidth temperature dependence and the coefficient for the self-broadening of water vapor;

a version of the Total Internal Partition Function (TIPS) program is used for the temperature dependence of the line intensities [Gamache at al., 1990];

the effects of line coupling are treated to second order with the coefficients for carbon dioxide in the 600 - 800 cm^-1 region [Hoke et al., 1989] and first order with the coefficients for carbon dioxide in 1932, 2080, 2093, and 2193 Q-branch regions [Strow et. al, 1994];

temperature dependent cross section data such as those available with the HITRAN database may be used to treat the absorption due to heavy molecules, e.g. the halocarbons;

an algorithm is implemented for the treatment of the variation of the Planck function within a vertically inhomogeneous layer as discussed in [Clough et al., 1992];

algorithmic accuracy of LBLRTM is approximately 0.5% and the errors associated with the computational procedures are of the order of five times less than those associated with the line parameters so that the limiting error is that attributable to the line parameters and the line shape;

its computational efficiency mitigates the computational burden of the line-by-line flux and cooling rate calculation [Clough et al., 1992], for example linear algebraic operations are used extensively in the computationally intensive parts of LBLRTM so that vectorization is particularly effective with a typical vectorized acceleration of 20;

FFT instrument function with a choice of 9 apodization functions;

includes a realistic spectral sea surface emissivity model in the infrared [Masuda, et. al., 1988, Wu and Smith, 1997];

input atmospheric profiles in either altitude or pressure coordinates;

interfaces with other radiative transfer models (like RRTM), and as the forward model for inversion algorithms (like the Tropospheric Emission Spectrometer (TES) retrieval algorithm.

These attributes provide spectral radiance calculations with accuracies consistent with the measurements against which they are validated and with computational times that greatly facilitate the application of the line-by-line approach to current radiative transfer applications.