Premium
Towards more realistic hypotheses for the information content analysis of cloudy/precipitating situations – Application to a hyperspectral instrument in the microwave
Author(s) -
Aires Filipe,
Prigent Catherine,
Buehler Stefan A.,
Eriksson Patrick,
Milz Mathias,
Crewell Susanne
Publication year - 2018
Publication title -
quarterly journal of the royal meteorological society
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.744
H-Index - 143
eISSN - 1477-870X
pISSN - 0035-9009
DOI - 10.1002/qj.3315
Subject(s) - computer science , satellite , numerical weather prediction , remote sensing , data assimilation , gaussian , focus (optics) , sensitivity (control systems) , radiative transfer , algorithm , environmental science , meteorology , geology , physics , optics , engineering , quantum mechanics , astronomy , electronic engineering
Information Content (IC) analysis can be used before an instrument is built to estimate its retrieval uncertainties and analyse their sensitivity to several factors. It is a very useful method to define/optimize satellite instruments. IC has shown its potential to compare instrument concepts in the infrared or the microwave. IC is based on some hypotheses such as the the Gaussian character of the radiative transfer (RT) and instrument errors, the first‐guess errors (Gaussian character, std and correlation structure), or the linearization of the RT around a first guess. These hypotheses are easier to define for simple atmospheric situations. However, even in the clear‐sky case, their complexity has never ceased to increase towards more realism, to optimize the assimilation of satellite measurements in numerical weather prediction (NWP) systems. In the cloudy/precipitating case, these hypotheses are even more difficult to define in a realistic way as many factors are still very difficult to quantify. In this study, several tools are introduced to specify more realistic IC hypotheses than the current practice. We focus on microwave observations as they are more pertinent for clouds and precipitation. Although not perfect, the proposed solutions are a new step towards more realistic IC assumptions of cloudy/precipitating scenes. A state‐dependence of the RT errors is introduced, the first‐guess errors have a more complex vertical structure, the IC is performed simultaneously on all the hydrometeors to take into account the contamination effect of the RT input uncertainties, and the IC is performed on a diversified set of cloudy/precipitating scenes with well‐defined hydrometeor assumptions. The method presented in this study is illustrated using the HYperspectral Microwave Sensor (HYMS) instrument concept with channels between 6.9 and 874 GHz (millimetre and sub‐millimetre regions). HYMS is considered as a potential next generation microwave sounder.