Reanalysis of western Pacific typhoons in 2004 using 4DVAR Data Assimilation Technique

Understanding rapid TC intensification, cyclogenesis, and their dynamics requires reliable, high temporal and spatial resolution data. With the latest satellite products (Aqua, QuikSCAT, TRMM, GOES, etc.) and some reconnaissance aircraft dropsonde data (DOTSTAR), we are now able to construct a comprehensive, high-resolution 3D TC reanalysis dataset using advanced 4DVAR data assimilation technique. This project is founded by NOAA. The 4DVAR system assimilates all available observational data into the forecast model outputs and retrieve dynamically balanced TC wind, pressure, moisture and temperature fields through mesoscale model dynamics (WRF or MM5) as adjoints. Our project focuses on analyses of TCs from May 1 to October 31, 2004 over the western Pacific and derives the data sets with a temporal resolution of 2-3 hours and a spatial resolution of 20-30 km with 30 vertical levels from the surface to 100mb, with each case reanalysis covering the entire life cycle of a TC from its genesis to dissipating stage.

As an example (Fig. 1), we show the 2004 Typhoon Tokage which was known as the deadliest storm to hit Japan in a decade. We compare NCEP analysis, observations and our reanalyzed intensities. The reanalyzed minimal sea level pressures have very good coincidence with the observation and reconstruct the progress of TC intensify and weaken.

To further show the performance of reanalysis, we select a satellite image from TRMM PR sensor at 200410182133Z when the TMI captured almost the whole typhoon (left panel of figure 2). The right panel of figure 2 is the reanalyzed simulated radar echo (DBZ) at the same time. Compared 2 images, we can see that the reanalyzed DBZ resemble the observed rain rate distribution very well.

The high-quality TC reanalysis effort may contribute to the NOAA Pacific Region Integrated Data Enterprise (PRIDE) objectives by providing new, integrated data products that are useful for risk management for US-affiliated Pacific islands. This database is utilized for a wide variety of purposes: setting of appropriate building codes for coastal zones, risk assessment for emergency managers, analysis of potential losses for insurance and business interests, intensity forecasting techniques, seasonal forecast-ing, and climatic change studies. The dataset will be posted in the Asian-Pacific Data-Research Center (APDRC) network (http://apdrc.soest.hawaii.edu/) and accessible via web tools to the public.