Investigation of Anvil Cloud Properties Associated with Mid Latitude Deep Convection by using Integrated Ground Radar and Satellite Observations

Anvil clouds associated with deep convection are important to atmospheric radiation and water budgets in the upper troposphere, but their properties are difficult to quantify due to the highly variable nature of deep convective clouds and the danger of sampling them in situ. To investigate the anvil cloud properties, we have developed a robust classification algorithm for mid latitude deep convective system by using data from the U.S. Next-Generation Radar (NEXRAD) and DOE Millimeter Cloud Radar. The algorithm can separate both non-precipitating (deep clouds and anvil clouds) and precipitating (convective and stratiform) portions from spring/summer time deep convective systems. It improves upon the classic convective and stratiform precipitation separation technique of Houze et al. (1995) by including vertical gradients of reflectivities, and multi-layer non-precipitating cloud classification (Frederick and Schumacher 2008). This classification algorithm can separate convective systems into convective, stratiform with or without brightband, deep cloud, mixed-phase anvil and ice anvil.

The radar classification of deep convection is then mapped onto GOES imager pixels. This integrated dataset is used to investigate three types of anvil cloud properties, namely mixed-phase anvil, ice anvil, and thin ice anvil. Anvil clouds generated from mesoscale convection (i.e. squall line – leading convective trailing stratiform) and non-organized convection (i.e. scatter storms) are compared separately. Anvil macro-properties such as area coverage and thickness, as well as microphysics retrieved from GOES radiances, such as optical thickness (τ), ice water path (IWP), and effective ice diameter will be examined. Preliminary results during May – Aug 2007 in Oklahoma show that the average thickness for mixed-phase anvil and ice anvil are 4.5 km and 2.8 km, respectively, and the variance of the mixed anvil thickness is larger than the ice anvil's due to greater variance in cloud top height. Probability density functions of GOES-retrieved optical depth show maximum frequencies at 20, 8, and 4 for mixed-phase, ice, and thin ice anvils, respectively; while the IWP for mixed anvils (500 g m^-2) is much larger than the two ice anvil types (100 g m^-2). The anvil characteristics in this study are different to those over tropics. This research demonstrates the potential application of the classification technique for studying convectively generated anvil clouds, particularly when integrated with satellite observations.