Tuesday, 24 January 2017: 4:45 PM
Conference Center: Skagit 5 (Washington State Convention Center )
At NOAA’s Hurricane Research Division (HRD), a variety of ensemble-based data assimilation activities are being pursued, primarily focusing on the high-resolution observations that are collected within tropical cyclones (TCs) during research and operational flight missions. For this purpose, the Hurricane Ensemble Data Assimilation System (HEDAS) was developed to combine a square-root ensemble Kalman filter with NOAA’s Hurricane Weather Research and Forecasting (HWRF) modeling system, and later augmented with a storm-relative observation processing capability. Observations that are routinely assimilated in HEDAS are obtained from NOAA P-3 and G-IV as well as Air Force Reserve C-130 research and reconnaissance flights (dropwindsonde, flight level, Stepped-Frequency Microwave Radiometer [SFMR], and tail Doppler radar [TDR]), NASA and NOAA unmanned aircraft system (UAS) platforms such as the Global Hawk and the experimental Coyote, satellite Atmospheric Motion Vectors (AMVs), and retrieved thermodynamic profiles from the Atmospheric InfraRed Sounder (AIRS) and Global Positioning System (GPS) Radio Occultation.
This talk will first introduce the general philosophy and approach adopted by HRD to tackle the problem of vortex-scale TC data assimilation and explain some of the details of HEDAS and its observation processing system, followed by an evaluation of how filter performance can be improved through alternative methods of humidity assimilation. Results from the implementation of an advanced quality control approach that aims to reduce the adverse impacts of filter nonlinearities and combines the ideas of adaptive observation error inflation, running-in-place assimilation, and observation recycling, all afforded by the use of storm-relative data assimilation approach, will also be presented. Finally, optimization considerations will be discussed with a holistic focus on assimilation frequency, covariance localization, and covariance inflation.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner