The Impacts Of Assimilating Retrieved Winds From Doppler Radar Observations On The Intensity And Track Forecasts Of Typhoon Jangmi (2008) And Its Mechanisms

Landfalling tropical cyclones (TCs) are among the most devastating natural disasters along coastal regions. The accuracy of TC track and intensity prediction is crucial in saving lives and properties. Studies show that the TC circulations play an important roll on the TC intensity and track forecasts. Ground-based Doppler radar has proved an important instrument to observe the landfalling TC internal circulation evolution with high temporal and spatial resolution. In this study, Typhoon Jangmi, which made landfall on the east coast of Taiwan on28September2008, is chosen as the test case; due to its sudden westward turn before landfall. During Jangmi’s landfalling, the typhoon is observed by four S-band WSR-88D radars located at Hualian, Wufenshan, Kenting, and Chigu, Taiwan, from0000UTC28to0600UTC28September. Firstly, new wind retrieval techniques are proposed to obtain the TC circulations by using the radar observations collected by the coastal radars in Taiwan. Secondly, the retrieved winds are assimilated into the cloud-permitted model to obtain better initialization of TC intensity and structure and improve TC forecasts. Finally, the impacts and mechanism of the asymmetric structure on the track deflection of Typhoon Jangmi are also investigated.Firstly, a gradient velocity track display (GrVTD) method is proposed using the gradient concept to obtain the TC circulation directly from aliased radial velocity. Ideal and real tests all show that the GrVTD method can avoid the impact of aliasing and retrieve the primary TC circulation accurately. The GrVTD-retrieved winds can be used as a first guess for dealiasing the radial velocity for subsequent use in other retrieval method or other applications. The coverage of the wind retrieval based on radial velocity is usually limited due to the limited radial velocity coverage. To obtain the accurate circulation with larger coverage, an extension of TREC (Tracking Radar Echo by Correlations) technique, named Typhoon Circulation TREC (T-TREC) is developed, in which the TREC analysis is performed on a polar grid centered on the tropical cyclone (TC) eye using arc-shaped correlation cells and arc-shaped search area with the shorter dimension in the radial direction and the longer dimension in the azimuthal direction in order to reflect the fact that the tangential component of typhoon circulation is usually much larger than its radial component. In addition to reflectivity data, the Doppler radar radial velocity is incorporated to estimate the search range and create a velocity correlation matrix in order to alleviate the uncertainty of tracking reflectivity only. Results shows that the method is robust and can retrieve the TC circulation with the retrieving errors less than4m s"1.Secondly, the T-TREC-retrieved winds (VTREC), for the first attempt, are assimilated into ARPS model for the analyses and forecasts of Typhoon Jangmi (2008) using an ensemble Kalman filter (EnKF), in order to examine the impact of circulations on the forecasts. In this study, the radar data are assimilated through30-minute assimilation cycles during different periods between0000UTC and0600UTC28before landfall. The relative impact of assimilating single radar radial velocity (VHL), or radial velocity from four radars (V4RD) or VTREC, or their combinations on the track, intensity, structure and precipitation are examined.The assimilation of VTREC data improves the intensity and structure of the typhoon significantly, while the assimilation of VHL data over the same time period yields a much smaller improvement due to limited VHL data coverage and poor retrieval of the cross-beam wind component. The improved analyses obtained by assimilating VTREC data allows for better track, intensity, structure, and precipitation forecasts. Sensitivity experiments show that VTREC assimilation can quickly build up a strong vortex, and thus have the potential to obtain a longer forecast leadtime, while a longer period of cycled data assimilation is required when using Vr data to spin up the vortex. Gradient wind balance diagnoses are performed to reveal the adjustment between pressure and wind fields, and results show that the pressure field can adjust quickly and efficiently to the enhanced wind fields. As a result, MSLP reduction can be achieved during the forecast steps of the EnKF, since strong vortex circulations have already been established by the radar data assimilation.The radial velocity has better accuracy, while VTREC data have larger coverage and include cross-beam wind component. Therefore, it is possible to combine two of them to further improve the TC forecasts. Five groups of experiments are designed to assimilate VHL, V4RD, VTREC, individually, or assimilate both V4RD and VTREC at the same time (marked as VTRECVr, or assimilate the VTREC at the first cycle then the V4RD for the rest of cycles (marked as VTFVR).For Vr assimilation, the more Vr observations in the inner core region, either a longer assimilation window or an assimilation window closer to landfall time, are assimilated, the better intensity and structure analyses can be obtained. Assimilating the combination of VTREC and V4RD can produce better analyses and forecasts than only assimilating individual of them. Quantitative verifications show that the assimilation of VTFVR produces the best track, intensity and precipitation forecast among all the groups of experiments, suggesting that assimilating VTREC at the first cycle then assimilating V4RD for the rest of cycles has the most significant impact on the TC forecasts. The precipitation verifications also suggest that the assimilating VTREC data has the ability to further improve the precipitation forecast compared to the assimilation of V4RD only. The full winds with large coverage retrieved from two reflectivity scans can well represent the motion of the rainbands, and thus have the potential to adjust the distribution of the rainbands and also improve the precipitation forecasts.Finally, based on the analyses from radar data assimilations, the mechanism influenced the westward track deflection of Typhoon Jangmi is investigated. Two initial conditions from two data assimilation experiments with significant track forecast differences are selected; and their environmental, axisymmetric, wavenumber1to3asymmetric fields are decomposed using the vortex separation and the Fourier decomposition methods. A series of experiments are conducted by superimposing different components to examine the impact of each component. Results show that the wavenumber1asymmetric structure dominates the westward turn of Typhoon Jangmi. A pair of cyclonic and anticyclonic gyres dominates the wavenumber1asymmetric circulation, and a nearly uniform ventilation flow through typhoon center is also induced, which advect the typhoon with the ventilation flow. The deep-layer-mean steering flow and ventilation flow are further calculated to quantitatively evaluate the impact of the ventilation flow. Results show that the steering flow and ventilation flow well match the typhoon motion, in which the ventilation flow provides most of the westward motion component, suggesting again that the ventilation flow controls the westward motion of Typhoon Jangmi. Overall, the wavenumber1asymmetric structure dominates the westward deflection of Typhoon Jangmi through the advection of ventilation flow induced by the wavenumber1gyres.