This study focuses on the hurricanes that developed and/or evolved in the Tropical Americas region, which encompasses the Gulf of Mexico and Caribbean, in 2005. These hurricanes are: Dennis (July 5-10), Emily (July 11-19), Katrina (August 2428), Rita (September 18-23), and Wilma (October 16-23). We perform our analysis on days when the hurricanes were over water only, which is when they were the strongest and the most structured.
In this study, we use a combination of continuous ground-based measurements and cloud-penetrating space-borne observations. Ground-based observations comprise of cloud-to-ground (CG) lightning flashes collected by the Long Range Lightning Detection Network (LLDN), which includes sensors from the U.S. and Canadian Lightning Detection Networks. The LLDN uses time of arrival and direction finding technology to determine the location of each CG flash (Cummins et al. 1998). Accurate flash placement depends on distance and location of the CG flash with respect to the network. In order to identify significantly misplaced CG flashes, we use hourly observations of 4 km horizontal resolution infrared brightness temperatures (IR Tb) from GOES-12 as a filter. We compare the location of lightning flashes within a +/—15 minute observation of IR Tb from GOES-12. Any flashes associated with co-located IR Tb warmer than 260 K are consequently excluded from the analysis.
Space-borne observations consist of water vapor measurements obtained from the Microwave Limb Sounder (MLS) instrument aboard the Aura satellite (Waters et al. 2006). This satellite was launched on July 15, 2004 and started reporting measurements in August 2004. The advantage of using this dataset is the capability of measuring in regions where ice clouds and aerosols are present. This cloud-penetrating capability allows for vertical profiling within convective systems, which makes it ideal for our study of hurricanes. Each orbit of the satellite covers from 82°N to 82°S with 14 orbits per day (^25° sampling in longitude). Measurements are obtained every 1.5° along the orbit track and cover a few kilometers across track. In this study, we use MLS version 1.51, Level 2 data, which covers the upper troposphere and lower stratosphere with a ^3 km vertical resolution.
We focus on measurements reported at 100, 147, and 215 hPa. At these altitudes, the single profile precision is less than 10%, and the averaging kernels are so sharply peaked that there is no instrument induced correlation among levels.
In order to elucidate the spatial and temporal evolution of the water vapor, lightning, and IR Tb fields over the Tropical Americas, we construct Hovmoller plots averaged over the 12°-30° North latitude band. For the cases of lightning and IR Tb, we calculate cumulative flashes and minimum IR Tb, respectively, in longitude increments of 2° on a daily basis. For the case of water vapor, we calculate averages in longitude increments of 8° every two days. The analysis focused on areas surrounding the center of the hurricane uses daily 2° x 2° grids of cumulative lightning flashes and co-located averaged water vapor.
Was this article helpful?