Version 2.2 released 26th November 2015
What's new in version 2.2:
The aerosol element of the software has been upgraded to use a new algorithm that works independently from the cloud detection and is focussed on the identification of desert dust (the most important and significant source of contamination in infrared spectra). The new algorithm examines differences between the observed brightness temperature at two spectral locations in the long-wave window. Several channels are averaged around each location to reduce noise effects and if the resulting difference exceeds a pre-defined threshold the entire spectrum is flagged as contaminated.
Version 2.1 released 19th February 2014
What's new in version 2.1:
A new screening step is introduced to make use of clustered AVHRR radiance data within a collocated IASI field-of-view. If the collocated AVHRR cluster data are available, an initial AVHRR-based cloud flag is derived and used as additional input to the main cloud detection algorithm. Additionally, default processing parameters for CrIS are updated on the basis of experience gathered from assimilation experiments using real radiance data at ECMWF.
This cloud and aerosol detection software is based on a pattern recognition algorithm developed for the detection of clouds in AIRS spectra.
The cloud detection algorithm works by taking the first guess departures (i.e. the difference between the observed brightness temperatures and brightness temperatues calculated from a good estimate of the atmospheric state - typically a 6-hour forecast from an NWP model) and looking for the signature of opacity that is not included in the clear-sky calculation (i.e. cloud or aerosol). The aerosol code decides whether an observation, initially defined as being cloud affected, is in fact mainly affected by aerosol. Identification of aerosol contamination is based on first guess departures of window channels in the 8µm region.
The software package is NWP-dependent (i.e., requires an estimate of the atmospheric state vector), but is sufficiently modular to "plug-in" to most NWP systems.
Input data: For each channel selected in the FOV, the algorithm requires the background (model) brightness temperature, the observed brightness temperature and a height assignment for each channel in units defined by the user (e.g., NWP model level, pressure level).
Output data: Output file is produced containing flags indicating clear, cloud-contaminated and aerosol-contaminated channels in each input satellite sounding.
- Further information
- McNally, A.P. and P.D. Watts, 2003. A cloud detection algorithm for high-spectral-resolution infrared sounders, Q J Roy Meteorol Soc, 129, 3411-3423.
- Eresmaa, R., 2014. Imager-assisted cloud detection for assimilation of Infrared Atmospheric Sounding Interferometer radiances, Q J Roy Meteorol Soc, 140, 2342-2352.
- User Manual
- Bug reports and fixes
- Brief outline of plans
- Continue improving the algorithm, releasing updates as appropriate.
- Feasibility of making extended use of different parts of the infrared spectrum for improved identification of clear / cloudy channels in the presence of background error is under investigation.
- Separate the aerosol detection part from the cloud detection part to allow diagnosing presence of aerosol in scenes where the aerosol radiative forcing is too weak to be detected through the cloud detection algorithm.
The Aerosol and Cloud Detection package can be requested using this form.