A New Historical Database of Tropical Cyclone Position, Intensity, and Size Parameters Optimized for Wind Risk Modeling

Striking advances have occurred in the realm of tropical cyclone (TC) wind observations over the past quarter century. These advances include the advent of highly accurate GPS-based navigation for aircraft-based observing platforms, airborne Doppler radar, GPS dropsondes, improved wind retrievals from satellite scatterometers, and the Stepped Frequency Microwave Radiometer (SFMR) instrument for measuring surface wind speeds. In the near future, there is potential for additional novel observing platforms and instruments to come online. Wind risk modeling may benefit substantially if such observations can be combined in a consistent and transparent manner across the modern era in a way that provides greater precision and characterization of the surface wind field.

The present work aims to construct a new historical database that is optimized for parametric wind risk modeling applications. Toward this goal, three new research-grade datasets have been constructed by systematically gathering and standardizing all available observations. These datasets include: (i) the Enhanced Vortex Data Message (VDM+) Dataset (1989-2015), the QuikSCAT Tropical Cyclone Radial Structure (QSCAT-R) Dataset (1999-2009), and the Extended Flight Level (FLIGHT+) Dataset (1997-2013). Each dataset is being made available to the research community to support a wide variety of research uses beyond the present work. The characteristics of each dataset will be briefly described at the beginning of this talk.

The remainder of the talk will focus on the effort to build the new historical database from the three source datasets. The new database will feature higher temporal and spatial precision by using the high resolution wind centers obtained from the Willoughby-Rahn center finding method. Center positions will be provided at 0.01 degree of precision and will attempt to characterize the actual wind center of the TC. Data for all metrics will be provided at 1-hr intervals. The new database will include time-dependent uncertainty bounds based on the inherent uncertainty of the available observing platforms/instruments that were available over the characteristic influence period for each time point. The database will also include alternative metrics beyond the traditional metrics of intensity and 34-, 50-, and 64-kt wind radii. Such metrics may include: the azimuthal mean wind speed, the probability of multiple wind maxima, the radial structure of the probability of exceeding various wind thresholds by quadrant, and other metrics related to the spatial and temporal coherence of the location of wind maxima. If time permits, we hope to validate the new database against a catalog of historical losses from well-known landfall events.