Sensitivity of Warm Rain Rates to Lower PBL Structure as Observed by Synthetic Aperture Radar over the Subtropical Ocean.

Synthetic Aperture Radar (SAR) can detect wind and rain signatures at meters-scale resolution on the ocean surface without interference from overlying cloud cover. Here, gusts and lulls in the surface winds induce recognizable patterns on the spatial distributions of capillary waves. These distinct wind signatures vary based on the 3-dimensional structure of the lower boundary layer and fall into two dominant categories: mesoscale convective (MC), and wind streaks (WS). MC boundary layers tend to be associated with gentler motions and lighter winds while WS are associated with stronger overturning rolls and stronger winds. SAR can detect rain columns because drops disturb the ocean surface and can also discern cold pools and atmospheric boundaries associated with precipitation. A machine learning routine has been developed to classify SAR images based on these signatures. We use Sentinel-1 WV images (20x20 km, 5 m pixels), which are routinely collected every 100 km over the world’s oceans.

SAR images classified as MC, WS, or containing boundaries or rain columns are compared to a variety of satellite cloud data in order to independently verify the classification system, and to gain insight into whether these classified PBL states have an effect on cloud and precipitation processes. SAR images, collected at sunrise and sunset during the year 2018, are linked to polar orbiting A-Train satellite observations. Observations are linked using Lagrangian PBL trajectories, following the cloud-level winds forward and backward from the SAR image to the 1:30 and 13:30 A-Train observation times. This “brackets” the SAR image with satellite data observed 12-hours apart, or in the case of daytime-only data, 24-hours apart. Comparisons are made in four marine subtropical stratocumulus regions.

Results show that SAR-observed rain columns and cold pools are associated with higher rain rates as seen by AMSR/2 89 GHz rain rate estimates trained by CloudSat. PBLs with wind streaks are associated with stronger rain rates and greater cloud liquid water path compared with mesoscale convective PBLs, even after controlling for wind speed. Further analysis shows that WS PBLs tend to be cloudier, shallower, and contain fewer cloud drops. This work highlights the importance of small-scale turbulent boundary layer processes in controlling cloud processes on sub-daily timescales, and motivates investment in future SAR observations over the ocean.