Publications

Ahn, E; Huang, Y; Siems, ST; Manton, MJ (2018). A Comparison of Cloud Microphysical Properties Derived From MODIS and CALIPSO With In Situ Measurements Over the Wintertime Southern Ocean. JOURNAL OF GEOPHYSICAL RESEARCH-ATMOSPHERES, 123(19), 11120-11140.

Abstract
In situ observations of cloud effective radius (r(eff)), droplet number concentration (N-d), and thermodynamic phase from 11 wintertime flights over the Southern Ocean (43-45 degrees S, 145-148 degrees E) are compared to products from MODerate-resolution Imaging Spectroradiometer (MODIS) and Cloud-Aerosol Lidar with Orthogonal Polarization. The in situ observations were in close alignment with A-train overpasses for a 30-min window. For open mesoscale cellular convection, which was predominantly observed, clouds were commonly found to be intermittently drizzling, patchy, and mixed phase. Compared to the in situ observations of the cloud thermodynamic phase, the Cloud-Aerosol Lidar with Orthogonal Polarization and MODIS cloud phase optical property products consistently underestimated the occurrence of mixed-phase clouds, whereas the MODIS infrared-based phase product showed a better qualitative agreement despite a frequent classification of uncertainty. The MODIS r(eff_2.1) overestimated the in situ r(eff) for nondrizzling clouds (by similar to 13 mu m on average) and, to a lesser extent, for lightly drizzling cases. Conversely, MODIS r(eff_2.1) underestimated the in situ r(eff) for heavily drizzling cases by similar to 10 mu m on average. The overestimation of r(eff) is much greater than that for the stratocumulus over the Southeast Pacific shown in other studies. An examination on subpixel heterogeneity, droplet size variability, a bimodal distribution, and solar zenith angle suggests that all of these factors have measurable impacts on the MODIS r(eff) bias. The MODIS N-d is largely consistent with the in situ observations. However, the N-d of the two high N-d cases (closed mesoscale cellular convection) are highly underestimated. An error analysis suggests that the N-d biases are likely a result of a compensating error effect.

DOI:
10.1029/2018JD028535

ISSN:
2169-897X