[17] Observational issues 2


[17-1] March 2, 10:50-11:20

[Invited]

The impact of Airborne Radio Occultation Observations on the simulation of Hurricane Karl (2010)

S.-H. Chen (University of California, Davis), X.M. Chen (University of California, Davis), J.S. Haase (Scripps, UCSD), B.J. Murphy (Purdue University), J.L. Garrison (Purdue University), K.N. Wang (Jet Propulsion Laboratory, California Institute of Technology), S.Y. Chen (National Central University), C.Y. Huang (National Central University), L. Adhikari (Texas A&M University-Corpus Christi), and F. Xie (Texas A&M University-Corpus Christi)

 
Abstract

This study evaluates, for the first time, the impact of airborne Global Positioning System (GPS) radio occultation (RO) observations on weather forecasts. The evaluation is based on the case study of Hurricane Karl during the Pre-Depression Investigation of Cloud-systems in the Tropics (PREDICT) field campaign in 2010. The assimilation of airborne RO (ARO) data was developed for a three-dimensional variational (3DVAR) analysis system. The impact of ARO data on Karl's forecasts was evaluated through data assimilation experiments of local refractivity and non-local excess phase (EPH), in which the latter accounts for the extended horizontal sampling of the signal ray path. The tangent point positions (closest point of a RO ray path to the Earth surface) drift horizontally and the drifting distance of ARO data is about two to three times of that of spaceborne RO. Thus, the assimilation of ARO data with drifting versus non-drifting tangent point positions was also tested. Results indicate that the assimilation of ARO observations improved model initial conditions and the forecast of Karl's rapid intensification. The assimilation of non-local EPH resulted in a greater improvement of the analysis than the assimilation of local refractivity. Later development of the storm indicates that forecasts were sensitive to the tangent point locations of assimilated ARO observations. The consideration of the ARO tangent point drift resulted in more accurate forecasts. Among all experiments, the best forecast was obtained by assimilating ARO data with the most accurate geometric representation, i.e., the use of non-local EPH operators with tangent point drift.

  Presentation file: 17_1_S.H.Chen.pdf