March 26, 2021 - 3:30 pm
March 26, 2021 - 4:30 pm
CategoriesConvective Meteorology (Mesoscale Dynamics)
Convective Meteorology (Mesoscale Dynamics) Seminar
Frequent Assimilation of Enhanced AMVs Using Hybrid EnVar to Improve Hurricane Prediction
Ben J. Davis
Friday, March 26th
Join Google Meet
Novel high frequency, high resolution enhanced atmospheric motion vector (AMV) data is assimilated in Hurricane Irma (2017). The enhanced AMV data set has sub-hourly frequency, 6 km spatial resolution, and is available directly above the tropical cyclone. In contrast, operational atmospheric motion vector (AMV) data is available at 6-Hourly frequency with >25 km resolution, and is not available directly above tropical cyclones due to data processing limitations. 6-Hourly data assimilation (DA) may struggle with the more frequent data when the background error is rapidly evolving. Therefore, this study evaluates the impact of using 1-Hourly Hybrid EnVar DA with the enhanced AMVs, which may better resolve rapidly evolving background error covariance. Further, this study looks to evaluate the impact of assimilating the AMV data directly above the TC.
The observation error profile and frequency of vortex relocation are found to be important considerations for optimizing 1-Hourly DA. Further, assimilating enhanced AMVs with 1-Hourly frequency in found to improve track forecasts compared to assimilating the same data with 6-Hourly frequency. However, intensity forecasts are not significantly improved.