A comparison of the MSG-CPP High IWC mask with 2015 Cayenne HAIC field campaign RASTA radar data shows a good probability of detection (~80%) of high IWC (>1 g/m3), and a rapid decrease in the probability of detection from a maximum IWC of 1 g/m3 down to 0.1 g/m3, consistent with results presented in de Laat et al.  and HAIC deliverable D32.2. The parallax correction and differences in spatial resolution of RASTA and SEVIRI have a limited effect on the verification. Large solar zenith angles, and large viewing angles do reduce the performance somewhat, but can be accounted for.
The usefulness of comparing the High IWC mask against in situ measurements of IWC by the ROBUST probe clearly is limited, even though the High IWC mask probability of detection of in situ IWC > 1 g/m3 is better than 80%.
Comparison of the High IWC mask with on qualitative board weather radar data is hampered by translating the radar data to more quantitative values, but confirms the common and frequent presence of potentially hazardous atmospheric conditions like deep convection or High IWC environments in absence of strong weather radar signals.
Finally, a brief investigation of a limited set of 18 MOZAIC/IAGOS inter-continental flights showed that commercial airlines generally tend to avoid hazardous atmospheric conditions, as for these 18 inter-continental flights combined only three measurements of in situ IWC > 1 g/m3 were found in the data.
ATJ de Laat, JF Meirink. Evaluation of SEVIRI measurements of high ice water content environments with aircraft data (HAIC Deliverable 32.4)