Investigating reflectivity-liquid water content relationships in mixed-phase clouds

Poster PDF

Description

Guo et al. showed a technique to partition Ka-band ARM zenith radar (KAZR) Doppler velocity spectra into reflectivity contributions from cloud (small fall-velocity) and precipitation particles. This technique effectively partitions the total reflectivity between liquid and ice contributions in mixed-phase clouds. Quantifying the liquid water content (LWC) from the reflectivity (liquid contribution) remains problematic in these mixed-phase clouds because of the complex physical processes that influence the drop size distribution (DSD). It is common to use a power law relationship, LWC = a*Zb, to derive LWCs from radar-measured reflectivities. For analytical DSDs, the coefficients a and b depend on the total number concentration of droplets and shape parameter of the DSD, which may vary widely depending on the different environmental conditions and physical processes in clouds. In particular, in mixed-phase clouds the ice particles grow at the expense of liquid droplets, impacting the liquid DSD in ways not present in liquid-only clouds. Therefore, one should anticipate larger uncertainties if only a single Z-LWC relationship is used to calculate LWC throughout the cloud. The accuracy of the retrieved LWC fields will be compromised without a corresponding understanding of the physical processes of the formation of Z-LWC relationships, even if the retrieved liquid contributed reflectivities are successfully separated from radar observations.

In our study, we combine ground-based observations and model simulations to investigate the variation of Z-LWC relationships in mixed-phase clouds. For each hour of data we define a single relationship, the a and b coefficients of which are chosen to provide the best fit to the microwave radiometer (MWR) liquid water path (LWP) over that period. We find that the radar-derived LWPs generally track the MWR LWPs well, with short periods of larger deviations occasionally. We use the DSD output from a cloud-resolving model to calculate the Z-LWC relationships at each model grid. The coefficients are found to vary depending on the microphysics and dynamics of the different stages of the cloud processes. Any improvement of our understanding of how the DSD varies with the mixed-phase cloud processes will improve LWC estimates.