Anu Dudhia

Abstract:

Microwindows are the small subsets of the complete MIPAS spectrum which are used for the retrievals of the L2 products. Ideally, these microwindows represent regions of the spectrum with maximum information content on the target molecule and minimum contribution from systematic errors, such as interfering lines of other molecules. MIPAS originally generated spectra sampled at 0.025 cm-1 with a nominal scan pattern consisting of 17 tangent points from 68-6 km altitude with a minimum spacing of 3 km. Since January 2005 MIPAS has been operating at 40% reduced resolution (0.0625 cm-1), using a new nominal mode scan pattern consisting of 27 tangent altitudes with a minimum spacing of 1.5 km and a latitude-dependent altitude offset. The change in spectral resolution, the increase in vertical sampling and the floating altitude grid have all had significant impact on the microwindow selection. A new set of microwindows has now been selected for the reduced resolution operating mode and preliminary error estimates have been calculated for the L2 products.

Abstract:

This paper discusses the variation and validation of the precision (defined as the dispersion of an ensemble of retrievals obtained from measurements of the same atmospheric state) of the ESA L2 products from the Michelson Interferometer for Passive Atmospheric Sounding (MIPAS). The MIPAS L2 products contain estimates of random error derived from the propagation of the radiometric noise through the retrieval. The noise itself varies with time, steadily rising between decontamination events, but its contribution to the L2 random error also depends on the atmospheric temperature, which controls the total radiance received. Hence, for all species, the random error varies latitudinally/seasonally with atmospheric temperature, with a superimposed time dependence on decontamination events. The precision validation involves comparing two MIPAS retrievals at the intersections of ascending/descending orbits. For every 5 days per month of high resolution MIPAS operation, the standard deviation of the statistic of the matching profile pairs is computed and compared with the predicted random error given in the MIPAS Offline L2 data. Even taking into account the propagation of the pressure-temperature retrieval errors into the VMR retrieval, the observed scatter is usually a factor 1-2 larger than the predicted error. This is thought to be due to effects such as horizontal inhomogeneity of the atmosphere.

Abstract:

This paper discusses the variation and validation of the precision, or estimated random error, associated with the ESA Level 2 products from the Michelson Interferometer for Passive Atmospheric Sounding (MIPAS). This quantity represents the propagation of the radiometric noise from the spectra through the retrieval process into the Level 2 profile values. The noise itself varies with time, steadily rising between ice decontamination events, but the Level 2 precision has a greater variation due to the atmospheric temperature which controls the total radiance received. Hence, for all species, the precision varies latitudinally/seasonally with temperature, with a small superimposed temporal structure determined by the degree of ice contamination on the detectors. The precision validation involves comparing two MIPAS retrievals at the intersections of ascending/descending orbits. For 5 days per month of full resolution MIPAS operation, the standard deviation of the matching profile pairs is computed and compared with the precision given in the MIPAS Level 2 data, except for NO2 since it has a large diurnal variation between ascending/descending intersections. Even taking into account the propagation of the pressure-temperature retrieval errors into the VMR retrieval, the standard deviation of the matching pairs is usually a factor 1-2 larger than the precision. This is thought to be due to effects such as horizontal inhomogeneity of the atmosphere and instability of the retrieval.