An Evaluation of the Observational Capabilities of A Scanning 95-GHz Radar in Studying the 3D Structures of Marine Stratocumulus Clouds.

摘要

Marine stratocumulus clouds play a critical role in Earth's radiative balance primarily due to the role of their high albedo reflecting incoming solar radiation, causing a cooling effect, while weakly reflecting outgoing infrared radiation. Characterization of the 3-Dimensional (3D) structure of these cloud systems over scales of 20-40 km is required to accurately account for the role of cloud inhomogeneity and structure on their shortwave forcing and lifetime, which has important applications for Global Climate Models. For first time, such 3D measurements in clouds were made available from a scanning cloud radar during the U.S. Department of Energy (DOE) Atmospheric Radiation Measurement (ARM) program's Clouds, Aerosol, and Precipitation in the Marine Boundary Layer (CAP-MBL) field campaign in the Azores Islands. The scanning radar observations were complemented by a suite of zenith-pointing active and passive remote sensors that were deployed to provide a detailed description of marine stratus over a long-term observation period in the ideal marine environment commonly found at the Azores. The scanning cloud radar observations present a shift from a multi-instrument, vertically pointing 'soda-straw' observation technique to a radar-only, 'radar-centric' observation technique. The scanning radar observations were gridded using a nearest-neighbor type scheme devised to take the natural variability of the observed field into account. The ability of the scheme to capture primary cloud properties (cloud fraction, cloud boundaries, drizzle detection) was assessed using measurements from the vertically pointing sensors. Despite the great sensitivity of the scanning cloud radar (-42.5 dBZ at 1 km range), the drop in sensitivity with range resulted in an artificial thinning of clouds with range from the radar. Drizzle-free cloud structures were undetectable beyond 5 km from the radar. Cloud fields containing drizzle were generally detectable to ranges exceeding 10 km from the radar. Well-defined streaking patterns in the drizzle field (reflectivity greater than -15 dBZ) at cloud base were concluded to be concomitant with the formation of boundary layer rolls. Sounding data for these well-defined (unbroken) rolls revealed a mean sub-cloud layer wind exceeding 3.9 ms −1, sub-cloud layer shear exceeding 7.5 x 10−3 s−1, and a majority of streaks oriented within 20° of the mean sub-cloud layer wind, satisfying many boundary layer roll criteria proposed in past studies. Attempts to reconstruct the 3D cloud liquid water content and 2D column liquid water path across the scanning radar domain using Z (Reflectivity) vs. LWC (Liquid Water Content) regressions trained using the zenith measurements were proved ineffective due to the overall extent of drizzle at Graciosa, and errors associated with sensitivity loss at range. Despite some difficulties, the SWACR satisfied ARM metrics for success by proving effective at detecting weak clouds for extended time periods across a 10 km plane, and drizzle across a 20 km range, at high spatial resolutions. Difficulties in resolving accurate vertical velocity patterns also suggest the need for an adaptive sampling strategy to most effectively remove horizontal wind components.