Abstract
Mineral dust is the most important natural source of atmospheric ice nuclei (IN) which may significantly mediate the properties of ice cloud through heterogeneous nucleation and lead to crucial impacts on hydrological and energy cycle. The potential dust IN effect on cloud top temperature (CTT) in a well-developed mesoscale convective system (MCS) was studied using both satellite observations and cloud resolving model (CRM) simulations. We combined satellite observations from passive spectrometer, active cloud radar, lidar, and wind field simulations from CRM to identify the place where ice cloud mixed with dust particles. For given ice water path, the CTT of dust-mixed cloud is warmer than that in relatively pristine cloud. The probability distribution function (PDF) of CTT for dust-mixed clouds shifted to the warmer end and showed two peaks at about −45 °C and −25 °C. The PDF for relatively pristine cloud only show one peak at −55 °C. Cloud simulations with different microphysical schemes agreed well with each other and showed better agreement with satellite observations in pristine clouds, but they showed large discrepancies in dust-mixed clouds. Some microphysical schemes failed to predict the warm peak of CTT related to heterogeneous ice formation.
Highlights
Since 1940s1 or earlier, laboratory experiments have been showing mineral dust particles, with varied mineralogy and size, can initialize ice nucleation at warmer temperatures and lower super saturations comparing to those required by pure water freezing[2]
High and thick deep convective clouds with widely spread ice clouds formed in this mid-latitude mesoscale convective system (MCS) with a clear comma structure indicating strong updraft and vorticity
Strong spatial gradient of dust loading with greater coarse mode aerosol optical depth (AOD) to its western area than that to its eastern area provided a good opportunity to investigate the potential impacts of dust aerosol on cloud properties
Summary
At 5:05 AM UTC on April 25, 2008, a typical case of dust-cloud interaction in Northeastern Asia was captured by Aqua, CALIPSO and CloudSat satellites in the A-train constellation[24,25,26]. The peaks of homogeneous ice formation in WRF simulations are about 5 °C colder than observations in sector 1 and 2, given the possible error of satellite retrieval (see later discussion), this difference is not significant This confirmed that under light dust-laden condition, most ice formation parameterizations in WRF model can correctly represent the CTT of ice clouds with tolerable errors. Compared to the satellite observations, at temperatures −20 to −40 °C, the WRF-MOR and WRF-Lin significantly underestimated the ice formation (very low PDF of CTT), while the WRF-WSM5, WRF-WSM6 and WRF-Goddard successfully simulated the peak of heterogeneous ice formation, but with some overestimations This result implies that the ice formation parametrizations in WRF can lead to significant variations in the prediction of ice cloud properties at warm temperatures. It demonstrated those microphysical parameterizations without taking into account the concentration of IN source aerosol (e.g. dust) may yield to large uncertainty in simulations and cannot explain the satellite observations
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
