P2.16
Correcting for Precipitation Scattering Effects in Satellite-based Passive Microwave Tropical Cyclone Intensity Estimates
Robert S. Wacker, CIMSS/Univ. of Wisconsin, Madison, WI; and C. S. Velden
Despite recent advances in tropical cyclone (TC) track forecasting, intensity forecasting remains problematic, partly due to uncertainty in intensity observations. Aircraft reconnaissance provides high-quality in situ observations, but is limited to parts of the Atlantic and Eastern Pacific basins. For the rest of the world, forecasters rely on satellite intensity estimates. The Dvorak technique, a set of empirical rules using visible and enhanced infrared imagery, has been the standard satellite intensity estimation technique for 30 years, but is subjective and lags rapid TC intensity changes. Current Dvorak RMS errors are approximately 10mb. Early satellite microwave sounding instruments in the 1970's demonstrated potential to infer intensity from the upper tropospheric warming above TC centers, but lacked the spatial and temporal resolution to be operationally useful. That changed in 1998 with the introduction of the Advanced Microwave Sounding Unit (AMSU). Since then, two approaches have emerged for estimating TC intensity with AMSU: Demuth et al (2004) use AMSU temperature retrievals, along with hydrostatic equilibrium and gradient wind balance, to derive the pressure and wind field around a TC. They estimate TC central pressure statistically using predictors derived from these pressure and wind fields, and achieve 9mb RMSE. Brueske and Velden (2003) developed a simpler approach, estimating TC central pressure directly from the TC core's warm brightness temperature anomaly using a single AMSU-A upper-tropospheric sounding channel. After refinement to correct for scan geometry effects, this technique's RMS error has reached 6 mb. The primary remaining uncertainty in the Brueske and Velden technique is the extent to which liquid and frozen hydrometeors scatter the upwelling microwave radiances. Based on qualitative scattering considerations, numerical simulations, and principal component analysis of AMSU warm core observations, additional AMSU-A and -B channels are incorporated into the Brueske and Velden technique to correct for scattering effects and further improve microwave intensity estimates.
Poster Session 2, New and Future Sensors and Applications: Part 2
Monday, 20 September 2004, 3:00 PM-4:30 PM
Previous paper Next paper