Vicarious calibration monitoring for MWI and ICI using NWP fields

Vicarious calibration monitoring for MWI and ICI using NWP fields
Date Published
Eumetsat Contract Report
Abstract The MicroWave Imager (MWI) and Ice Cloud Imager (ICI) on board Metop-SG-B will provide a wealth of information to constrain our knowledge of the global hydrological cycle, from ice particles in the upper troposphere to the heaviest tropical precipitation. These instruments will represent the first microwave imagers launched by Europe, building upon designs that have underpinned global estimates of precipitation since the 1980s, but crucially adding coverage of the currently unexploited “sub-mm” part of the spectrum by observing above 300 GHz. These new frequency bands will provide much greater sensitivity to atmospheric ice than ever before. Beyond observing hydrometeors in all forms, the combined capabilities of MWI and ICI span sensitivities to surface properties, water vapour, atmospheric temperature, and even ozone. Because of this highly complementary information across the spectrum, the two sensors can be combined into a single super-sensor designated “MWIICI” and this reflects how the data will be used at ECMWF, both scientifically and technically. The large range of physical sensitivities and cross-spectral sensitivity is the great strength of MWIICI, but it represents a significant challenge to traditional calibration and validation activities (“cal/val”) for new instruments.
In this study, a method is developed for analysing the in-orbit performance of MWIICI in terms of biases. It is based on comparing the observations against modelled radiances from a state-of-the-art numerical weather prediction (NWP) model. By examining departures (observations minus model background, or O-B) against the IFS as a reference, we can analyse sensor performance everywhere on the globe—millions of observations each day—and readily compare to similar sensors, as well as potentially other observations. The monitoring is underpinned by the “all-sky” approach for assimilating microwave radiances that has been pioneered at ECMWF and uses the RTTOV-SCATT radiative transfer model to include the radiative effects of clouds and precipitation. The latest version of RTTOV-SCATT is used, capitalising on numerous radiative transfer developments in recent years such as the SURFEM-Ocean emissivity model, updated gas spectroscopy for the sub-mm, and improved scattering properties of ice particles. By including hydrometeors in the forward model and considering each channel’s surface and cloud sensitivity for every scene, this method retains observations that might be thrown away by a traditional “clear-sky” sampling approach whilst yielding a more balanced data sample for cal/val. The all-sky sampling approach seeks to maximise the data available for cal/val whilst removing most scenes with known model biases, namely thicker clouds and surfaces such as sea-ice and most land.
The sampling method relies on a symmetric approach for screening out radiatively significant cloud signals, checking for the presence of cloud in both the model and the observation. This is accomplished using the cloud impact (CI) parameter, defined by the difference in brightness temperature (TB) between a clear scene and the TBs that are observed and modelled. Additionally, a channel-specific surface-to-space transmittance is checked to determine if an observation has too much surface sensitivity to be included in the cal/val sample. Because CI and surface-to-space transmittance are location- and channel-specific, the main cal/val sample (“stringent” data selection) is maximised based on each channel’s specific sensitivities to the surface and hydrometeors; a “dynamic” sample adds data over land for surface-sensitive channels to better capture the dynamic range of MWIICI. In addition to these channel-based samples for cal/val, a “unified” sample is defined using CI at key wavelengths that considers the sensitivities of all channels together to define a common sample. Thus three different data samples are defined for cal/val, applied equally to MWIICI and microwave imagers currently assimilated at ECMWF like GMI, SSMIS, and AMSR2.
Results are split into two sections: application of the method to two test orbits of MWIICI and application to current microwave imagers. The former clarifies the expected sensitivities of MWIICI and shows the data flow through the IFS. By applying the method to current sensors, we can assess its ability to detect EUMETSAT Contract Report 3 MWI and ICI Calibration Monitoring known sensor biases. For example, the analysis clearly shows orbital biases for F17 SSMIS, positive overall biases for AMSR2 channels, and generally small biases for GMI, each of which is well known in the cal/val literature. In addition, the method points to some bias structures that were not previously clear such as apparent scene-dependent biases for several channels, most notably 150 GHz on SSMIS. These results and the consistency of geographic patterns between similar channels on different sensors give confidence in the method, in that the cal/val sample exhibits similar patterns of atmospheric variability despite distinct orbits and differing central frequencies of matched channels. Between the three data samples defined here, between 20 and 70% of the total data are used for cal/val analysis, depending on the channel and the sample chosen. For the channel-specific samples, more data are retained for purely sounding frequencies such as the 50 and 118 GHz sounding complexes, with less data retained for frequencies like 89 GHz that have strong surface and cloud sensitivity.
Performance of the MWI and ICI instruments will be monitored in near-real-time after launch, leveraging the statistical and graphical tools developed at ECMWF for monitoring the global observing system. This will be the first time that such NWP-based monitoring is tailor-made to comprehensively evaluate instrument specifications, included as an integral part of the cal/val activities for MWI and ICI. Thus cal/val requirements such as inter-channel and intra-scan bias characteristics can be assessed almost immediately via an ECMWF-hosted website. Of particular interest is the direct comparison with equivalent channels on the reference-quality instrument GMI, permitting a type of “double difference” analysis of channel biases using the ECMWF model as a transfer standard. This facility will provide EUMETSAT and the global meteorological community with valuable information for assessing radiances from the MWI and ICI instruments. The technical developments of this project lay the groundwork for assimilation of MWI and ICI, aiding early operational exploitation of these missions.
DOI 10.21957/7c2d18d2e1