Assessment of MODIS cloud effective radius and optical thickness retrievals over the Southeast Pacific with VOCALS-REx in situ measurements

[1] Cloud microphysical observations collected in situ during the VAMOS Ocean-Cloud-Atmosphere-Land Study Regional Experiment within the Chile-Peru stratocumulus cloud deck during October–November 2008 were used to assess MODIS Level 2 cloud property retrievals. The in situ aircraft-derived cloud property values were constructed from the drop size distributions measured by the Cloud Droplet Probe (drop diameter <52 micron) and Two-Dimensional Cloud Probe (drop diameters up to 1600 micron) during 20 vertical profiles. Almost all of the MODIS cloud scenes were highly homogeneous. MODIS cloud optical thickness correlated well with the aircraft-derived value with a slight offset within instrumental/retrieval uncertainties. In contrast, the standard 2.1 micron-derived MODIS effective radius (re) systematically exceeded the in situ cloud top re by 15%–20%, for an absolute error that increased with droplet size. The individual effective radius retrievals at 1.6, 2.1, and 3.7 micron did not provide additional information on cloud vertical structure for our data sample. The secondarily derived MODIS liquid water path also exceeded the in situ value. A MODIS-derived cloud droplet number concentration (Nd) estimate agreed the best of the four MODIS variables with the aircraft observations. The analysis also highlighted a lack of agreement in published satellite-derived Nd values, despite drawing on the same sources. A best a priori formula choice for Nd is likely to vary regionally. Four sources of errors within the MODIS re retrieval were investigated further: the cloud mode droplet size distribution breadth, the presence of a drizzle mode, above-cloud water vapor absorption, and sensor viewing angles. These processes combined conspired to explain most of the observed bias. The above-cloud water vapor paths were poorly specified, primarily because the cloud top heights are placed too high, and secondarily because the water vapor paths are unrealistic. Improvement of the above-cloud water vapor path specification can most easily and systematically improve the MODIS effective radius and liquid water path retrievals.