A fast radiative transfer model for limb radiances: Accounting for horizontal gradients.

A fast radiative transfer model for limb radiances: Accounting for horizontal gradients.
Technical memorandum
Date Published
Secondary Title
ECMWF Technical Memoranda
Sean Healy
Abstract We investigate the influence of horizontal structure in the atmosphere on the simulation of emitted infrared limb radiances from the Envisat-MIPAS instrument, and examine a pragmatic extension of the regression-based RTMIPAS fast radiative transfer model to deal with horizontal atmospheric structure. The pragmatic extension of RTMIPAS takes into account the horizontal structure at the ray-tracing step, but uses regression models derived from a set of horizontally homogeneous atmospheres in the transmittance parameterisation. This reduces the computational effort required for the derivation of the regression models. The effects of horizontal gradients in the atmosphere on MIPAS radiances are examined by comparing radiance spectra simulated with RTMIPAS to line-by-line equivalents and to real MIPAS observations. It is shown that neglecting horizontal structure in the atmosphere can introduce errors in the radiance simulation that exceed by far the instrument noise, particularly for lower tangent heights and for strongly absorbing spectral regions. The pragmatic extension of the fast radiative transfer model can substantially reduce this error, and for tangent pressures less than 350 hPa (tangent heights greater than about 8 km) almost all error due to horizontal gradients is eliminated. In agreement with this, fast-model radiances simulated from cross-sections of the ECMWF model compare significantly better to observed MIPAS limb radiances than radiances simulated under the assumption of horizontal homogeneity. This further confirms a considerable reduction in forward-model error when horizontal atmospheric structure is taken into account.
URL https://www.ecmwf.int/en/elibrary/73751-fast-radiative-transfer-model-limb-radiances-accounting-horizontal-gradients
DOI 10.21957/7o491z