Airborne Doppler Wind Lidar Observations of the Tropical Cyclone Boundary Layer

Jun Zhang, Robert Atlas, G. Emmitt, Lisa Bucci, Kelly Ryan
2018 Remote Sensing  
This study presents a verification and an analysis of wind profile data collected during Tropical Storm Erika (2015) by a Doppler Wind Lidar (DWL) instrument aboard a P3 Hurricane Hunter aircraft of the National Oceanic and Atmospheric Administration (NOAA). DWL-measured winds are compared to those from nearly collocated GPS dropsondes, and show good agreement in terms of both the wind magnitude and asymmetric distribution of the wind field. A comparison of the DWL-measured wind speeds versus
more » ... ind speeds versus dropsonde-measured wind speeds yields a reasonably good correlation (r 2 = 0.95), with a root mean square error (RMSE) of 1.58 m s −1 and a bias of −0.023 m s −1 . Our analysis shows that the DWL complements the existing P3 Doppler radar, in that it collects wind data in rain-free and low-rain regions where Doppler radar is limited for wind observations. The DWL observations also complement dropsonde measurements by significantly enlarging the sampling size and spatial coverage of the boundary layer winds. An analysis of the DWL wind data shows that the boundary layer of Erika was much deeper than that of a typical hurricane-strength storm. Streamline and vorticity analyses based on DWL wind observations explain why Erika maintained intensity in a sheared environment. This study suggests that DWL wind data are valuable for real-time intensity forecasts, basic understanding of the boundary layer structure and dynamics, and offshore wind energy applications under tropical cyclone conditions. Remote Sens. 2018, 10, 825 2 of 15 in manned aircraft gathering direct wind and humidity measurements in this turbulent region of the storm. The use of unmanned aircraft is a promising tool for TC boundary layer observations, but the technology is not yet advanced enough to collect fast response wind and thermal data (e.g., the rate of data transfer through a satellite). Other in-situ observing platforms such as research buoys suffer the same plight as manned aircraft, due to the likelihood of damage to instrumentation. Even if a research buoy does occasionally survive in a strong TC [11] , it must be located in the eyewall to obtain hurricane-force wind measurements. The probability of this occurring is small, due to uncertainties in the track forecast at the time of the buoy's pre-storm deployment. This lack of observational data is believed to be one of the primary reasons why boundary layer processes remain poorly represented in operational TC models [12, 13] , which limits their ability to improve intensity forecasts. Our understanding of the mean boundary layer structure has improved since the advent of the Global Positioning System (GPS) dropsonde in 1997 [14] . Due to limited resources, however, the sampling size of GPS dropsondes in individual storms is generally quite small (<20 for a 12 h observational period). In rare cases where multiple research aircraft are flown simultaneously, such as in Hurricanes Earl (2010) and Edouard (2014), more than 40 dropsondes can be collected in a 12-h window. Such composite analyses can present a radius-height view of the boundary layer [15, 16] . Previous studies have used this type of composite method to analyze a large number of dropsonde data collected in multiple storms, in order to characterize the mean climatological boundary-layer structure [17, 18] . Of note, these studies focused on the boundary layer of TCs with hurricane-force (33 m s −1 ) and stronger-strength winds. The differences in the boundary layer structure between that of a tropical storm and a hurricane are not well documented. Doppler radar onboard research aircraft provides extensive wind observations in hurricanes, but its vertical resolution is generally too coarse for boundary layer studies. When a TC experiences strong environmental vertical wind shear, its convective structure is usually asymmetric, which makes the distribution of precipitation asymmetric. Under such a scenario, Doppler radar wind measurements are limited by the lack of backscattering from precipitation. Doppler Wind Lidar (DWL) observations complement Doppler radar observations in regions of little to no precipitation in TCs. Additionally, the DWL provides a much larger data coverage area for wind profiles than GPS dropsondes. Baker et al. [19] provided an excellent review of previous impact studies that used both simulated satellite and real-world, aircraft-based DWL data to demonstrate the DWL's ability to measure winds in TCs. Impact experiments with real data, termed Observing System Experiments, have been conducted with and without DWL data to show the positive impact of this observing system on TC track and intensity forecasts [20] . Similar experiments with simulated DWL data, termed Observing System Simulation Experiments, have also shown positive impacts on track forecasts [21] [22] [23] . The present study further illustrates the usefulness of the DWL for TC studies, with a focus on understanding TC structure by analyzing wind profiles collected in Tropical Storm (TS) Erika (2015). More recently, a coherent DWL was flown in 2017 on the National Aeronautics and Space Administration (NASA)'s DC8 aircraft during the Convective Processes Experiment (CPEX), which has provided a new perspective on tropical convective systems [24] . DWLs continue to mature as airborne systems, based on their ability to derive wind measurements from molecular motions through direct detection, providing wind data in aerosol-sparse areas.
doi:10.3390/rs10060825 fatcat:mv3vcrsdkvcthoth6gktu3feyy