Arctic Mixed-Phase Cloud Properties Using Ground-Based And Satellite Remote Sensing

Abstract

Under the influence of global climate change, the Arctic is warming at a rapid rate approximately twice the global average. Arctic clouds play important roles in modulating the radiation balance, and in precipitation and hydrological cycles. Among all cloud types, mixed-phase cloud, which is defined as the coexistence of liquid droplets and ice crystals in the cloud layer, is the dominant low-level cloud type over the Arctic. Different from mixed-phase cloud at lower latitudes, Arctic mixed-phase clouds (AMCs) often occur as stratiform type and can persist for hours to days and sometimes even weeks. The mechanism maintaining the AMC is not well understood; most models have problems representing the lifetime of an AMC or the phase partitioning in the cloud. There are two major objectives for this dissertation. The first one is to improve our understanding of the AMC as well as its interaction with the Arctic environment using the integrated observations available at the DOE Atmospheric Radiation Measurement (ARM) North Slope of Alaska (NSA) site. The second one is to evaluate the Clouds and the Earth’s Radiant Energy System (CERES)- the MODerate-resolution Imaging Spectroradiometer (MODIS) retrieved cloud properties at the NSA site based on the understanding of AMC properties at this site. Using the ARM merged sounding data, we find that temperature and humidity inversions exist five and eight times, respectively, more often above the AMC than below it. Furthermore, the occurrence frequency of the AMC increases with stronger humidity inversions in the cold season. This result helps to explain the persistence of AMCs even when the Arctic surface is covered by snow and ice. Then, we further investigate the interaction between different air masses with atmospheric thermodynamic properties as well as AMC properties. We find that the atmosphere is colder and drier, has lower relative humidity (RH) and is more stable under the influence of a continental type of air mass (when wind comes from south at the NSA site). AMC occurrence frequency is positively correlated with RH and negatively correlated with atmospheric stability. Therefore, the AMC occurrence frequency is 20-30% lower during a southerly wind than for other wind directions. Furthermore, AMC has a stronger precipitation process in a northerly wind than in a southerly wind. This is possibly due to the cleaner air masses from the ocean. To study Arctic cloud properties from satellite observations, we evaluate CERES- MODIS (CM) retrieved cloud properties using 10 years of ground-based observations at the ARM NSA site. The comparisons show that the CM Ed4 cloud fraction can represent the seasonal variation of cloud at the NSA site in a relative sense, but the monthly mean CF differences between CM Ed4 and ARM range from 3 to 10%. The cloud phase classification for CM Ed4 has an excellent agreement with the ground-based classification. During the daytime (solar zenith angle, SZA < 82o), the annual mean liquid (ice) CFs from ARM and CM are around 63% (16%), and 63% (12%), respectively. During the nighttime (SZA > 82o), the CM Ed4 monthly mean liquid cloud fractions are 10-18% lower than ARM ones from July to October, the differences are small from November to June. The CM Ed4 retrieved cloud-top temperature (Ttop) and height (Ztop), on annual mean, agree with ARM observations within 3 K and 400 m, respectively, for single-layer clouds. For multi-layer daytime clouds, the annual mean CM Ttop has a warm bias of 5.3 K, and Ztop has a negative bias of ~700 m than ARM observations. The CM retrieved cloud microphysical properties were also compared with ARM NSA observations and retrievals. The mean differences between CM and ARM retrieved cloud optical depth and liquid water path are less than 5% and 3%, respectively. The CERES-retrieved mean cloud droplet effective radius (re) is ~1.4-2 μm greater than ARM retrievals, which is mainly due to satellite and ground-based retrievals representing different levels within one cloud layer.