Simulated device: SEVIRI on MSG (other instruments would be possible): Visible and Near-Infrared Satellite Reflectances
Contact: Leonhard Scheck (firstname.lastname@example.org)
|Column or 3D?||3D or column based|
|Operation on model time step?||yes|
|Required libraries||NetCDF, and if grib files should be read or written grib_api|
|Programming language||Python + Fortran|
|If online, is it modular?||The part of the code required in an online version will be a Fortran module. We would prefer to use Python for the additional routines required for an offline version.|
|Ground-based or space-borne?||space-borne|
Minimum set of physical quantities:
- 3D: pressure, temperature, specific humidity, cloud water, cloud ice, cloud snow content, cloud cover
- 2D: base and top height of shallow convection
- The bare minimum required to compute reflectances would be pressure, temperature, specific humidity, and liquid water content. The ice / water particle radii are obtained from parametrizations.
- External: reflectance look-up table has to be read in once (~100MB).
Need for a special grid?
- Column based version: No.
- 3D version: For each pixel the data in a ray directed towards the satellite (and in future versions a second ray directed towards the sun) is required. At the moment implemented only for rectangular grids
Are parameters given via namelist or have to be coded?
- There is a standard command line interface. For the online version there will be a namelist.
Are there special requirements to the input quantities?
Models for application: COSMO (+ preliminary tests with UCLA-LES, ICON-LES)
Suitable for MESSy?: Column-based: should become available in the next months
Format of input data: NetCDF or grib
Format of output data: NetCDF
Computational costs: The pure computing time is less than a CPU minute for a COSMO-DE scene using the look-up table based version with parallax correction. Future version will take further 3D effects into account (e.g. cloud shadows) and will presumably be several times slower.