Application of Consistent Quality Control Procedures to Provide 17-year (1996-2012) High Resolution GPS Dropsonde Data

The primary application of dropsonde data is use in studying and helping predict the path and intensity of hurricanes. Since 1996 the GPS dropsonde has served well for both the operational and research communities by making high quality, high vertical resolution atmospheric profile observations in regions that are otherwise very difficult to observe. A project has been underway in the Earth Observing Laboratory (EOL) at the National Center for Atmospheric Research (NCAR), to supply the community with a research quality dropsonde dataset that may be used for long-term studies of hurricane and tropical storm environments. Initial steps have included compiling seventeen years (1996-2013) of high resolution dropsonde data, from NOAA's National Hurricane Center and Hurricane Research Division and applying consistent quality-control measures in order to produce the highest research quality dropsonde climatology available.

Dropsondes are deployed from aircraft, most commonly over oceans. The NCAR GPS dropsonde system is currently installed on over 20 aircraft around the world. The utilization of these instruments has allowed researchers to obtain comprehensive observations of the structure, and evolution of tropical storms. As they descend, dropsondes measure pressure, temperature, and relative humidity at a half second rate and GPS winds at a quarter second rate. Between 1996 and 2012, 121 storms were sampled by the dropsondes in the Atlantic and Eastern Pacific Oceans, ranging in size and strength from tropical depressions to hurricanes, and over 15,000 dropsondes were deployed by NOAA alone.

Processing of these data includes subjecting them to an established set of quality control procedures, developed at NCAR/EOL, that conservatively remove erroneous data while retaining finer features of the thermodynamic and wind profiles. The quality control (QC) steps broadly include, examination of the individual raw data profiles for obvious problems; automated processing through NCAR's Atmospheric Sounding Processing Environment (ASPEN) software, which performs smoothing, and removes suspect data points; generation of histograms to evaluate the range and distribution of each parameter; examination of time series plots of each parameter to examine the consistency of soundings launched during each flight, and to show the variability of soundings from different missions; and finally, examination of the each individual quality controlled sounding profile. Care has also been taken to address errors related to fast falling dropsondes and those that do not transmit useful data to the surface.

This data set will serve as an invaluable tool for hurricane researchers, enabling them to better define characteristics unique to different regions within a storm, to document common large and small scale structures, and to improve understanding of the thermodynamic and wind structures that influence tropical cyclone tracks, intensity and precipitation forecasts.