Fast cloud parameter retrievals of MIPAS/Envisat

The infrared limb spectra of the Michelson Interferometer for Passive Atmospheric Sounding (MIPAS) on board the Envisat satellite include detailed information on tropospheric clouds and polar stratospheric clouds (PSC). However, no consolidated cloud product is available for the scientific community. Here we describe a fast prototype processor for cloud parameter retrieval from MIPAS (MIPclouds). Retrieval of parameters such as cloud top height, temperature, and extinction are implemented, as well as retrieval of microphysical parameters, e.g. effective radius and the integrated quantities over the limb path (surface area density and volume density). MIPclouds classifies clouds as either liquid or ice cloud in the upper troposphere and polar stratospheric clouds types in the stratosphere based on statistical combinations of colour ratios and brightness temperature differences. Comparison of limb measurements of clouds with model results or cloud parameters from nadir looking instruments is often difficult due to different observation geometries. We therefore introduce a new concept, the limb-integrated surface area density path (ADP). By means of validation and radiative transfer calculations of realistic 2-D cloud fields as input for a blind test retrieval (BTR), we demonstrate that ADP is an extremely valuable parameter for future comparison with model data of ice water content, when applying limb integration (ray tracing) through the model fields. In addition, ADP is used for a more objective definition of detection thresholds of the applied detection methods. Based on BTR, a detection threshold of ADP = 107 μm2 cm−2 and an ice water content of 10−5 g m−3 is estimated, depending on the horizontal and vertical extent of the cloud. Intensive validation of the cloud detection methods shows that the limb-sounding MIPAS instrument has a sensitivity in detecting stratospheric and tropospheric clouds similar to that of space- and ground-based lidars, with a tendency for higher cloud top heights and consequently higher sensitivity for some of the MIPAS detection methods. For the high cloud amount (HCA, pressure levels below 440 hPa) on global scales the sensitivity of MIPAS is significantly greater than that of passive nadir viewers. This means that the high cloud fraction will be underestimated in the ISCCP dataset compared to the amount of high clouds deduced by MIPAS. Good correspondence in seasonal variability and geographical distribution of cloud occurren