Computing Atmospheric Profiles From Spectral Measurements

The SEASCRAPE computer program estimates the composition, temperature, and pressure of the atmosphere as a function of position by fitting remote-sounding spectroscopic data to a mathematical model of the transfer of radiation through the atmosphere. SEASCRAPE consists of two integrated parts: (1) A forward mathematical model that is used to generate synthetic spectra and (2) an inversion algorithm that is used to estimate parameters. The forward model evaluates the emission and absorption of radiation on a line-by-line basis; in itscurrent form, it includes a one-dimensional (spherically symmetrical) atmospheric submodel with homogeneous layers of arbitrary thickness. A future version of the code will include a multidimensional submodel with inhomogeneous cells. The inversion algorithm uses a sequential parameter-estimation method known as the square-root information filter. The program incorporates some flexibility for handling large sets of data; the user can control the level of accuracy in the model equations for spectral lines to trade accuracy against speed. The program can run on a single processor but is written for a parallel-processing environment; if a parallel processor is available, it can simultaneously retrieve multiple distributions and/or estimate a single set of parameters from multiple spectra.