Increasing the spatial resolution of cloud property retrievals from Meteosat SEVIRI by use of its high-resolution visible channel: implementation and examples

The modification of an existing cloud property retrieval
scheme for the Spinning Enhanced Visible and Infrared
Imager (SEVIRI) instrument on board the geostationary
Meteosat satellites is described to utilize its highresolution
visible (HRV) channel for increasing the spatial
resolution of its physical outputs. This results in products
with a nadir spatial resolution of 1x1 km2 compared to
the standard 3x3 km2 resolution offered by the narrowband
channels. This improvement thus greatly reduces the resolution
gap between current geostationary and polar-orbiting
meteorological satellite imagers. In the first processing step,
cloudiness is determined from the HRV observations by a
threshold-based cloud masking algorithm. Subsequently, a
linear model that links the 0.6 μm, 0.8 μm, and HRV reflectances
provides a physical constraint to incorporate the
spatial high-frequency component of the HRV observations
into the retrieval of cloud optical depth. The implementation
of the method is described, including the ancillary datasets
used. It is demonstrated that the omission of high-frequency
variations in the cloud-absorbing 1.6 μm channel results in
comparatively large uncertainties in the retrieved cloud effective
radius, likely due to the mismatch in channel resolutions.
A newly developed downscaling scheme for the 1.6 μm
reflectance is therefore applied to mitigate the effects of this
scale mismatch. Benefits of the increased spatial resolution
of the resulting SEVIRI products are demonstrated for three
example applications: (i) for a convective cloud field, it is
shown that significantly better agreement between the distributions
of cloud optical depth retrieved from SEVIRI and
from collocated MODIS observations is achieved. (ii) The
temporal evolution of cloud properties for a growing convective
storm at standard and HRV spatial resolutions are compared,
illustrating an improved contrast in growth signatures
resulting from the use of the HRV channel. (iii) An example
of surface solar irradiance, determined from the retrieved
cloud properties, is shown, for which the HRV channel helps
to better capture the large spatiotemporal variability induced
by convective clouds. These results suggest that incorporating
the HRV channel into the retrieval has potential for improving
Meteosat-based cloud products for several application
domains.