Creating a consistent climatology of tropical cyclone structure as observed by satellite microwave sensors

Microwave imagers provide multiple benefits over visible and infrared imagers in discerning tropical cyclone (TC) structure, but compiling a research quality data set is problematic due to several inherent technical and logistical issues. The simple task of compiling the digital data sets from multiple operational and R&D sensors over the past 25 years (since the 1st operational sensor) has been one of the greatest limiting factors, but recent efforts by several projects has now facilitated major forward progress.

The HURSAT-MW (Knapp 2008) has provided the only resource to date for TC-centered microwave imagery data, though it is limited to the Special Sensor Microwave/Imager (SSM/I) satellite sensors. While there are multiple other satellite sensors that incorporate microwave frequencies, inter-comparison between such sensors is limited by the different available channels and calibration metrics between satellites. In addition, varying spatial resolutions provide data set inconsistencies in resolving TC structural features. To remedy these difficulties, a global archive of TCs as measured by all available US satellite microwave sensors is compiled and standardized, incorporating multiple large data sets available from Colorado State University (CSU), Jet Propulsion Laboratory (NASA/JPL), and the Naval Research Laboratory (NRL-DC).

TCs in microwave imagery are found and isolated using best track data from the National Hurricane Center, Central Pacific Hurricane Center, and the Joint Typhoon Warning Center for the Defense Meteorological Satellite Program (DMSP) series (including all SSM/I and SSMIS sensors), TMI, AMSR-E, and WindSat. For each TC overpass, data from the “ice scattering” channels (i.e. 85GHz, 89GHz, and 91GHz) are recalibrated to a common frequency (89GHz), and then resampled at a high resolution using the Backus-Gilbert technique (Backus and Gilbert 1967). Using this higher resolution image, the TC center is then calculated using the ARCHER technique (Wimmers and Velden 2010), which analyzes the storm's rainband and eyewall organization. Ultimately, this project develops a consistent climatology of research-quality historical TC satellite microwave observations. Such data sets are needed to facilitate automated TC intensity estimates, more accurately study TC structural evolution, and provide methods to enhance current operational and research products, such as at the NRL TC webpage (http://www.nrlmry.navy.mil/TC.html).