7.2
Lightning Data Assimilation using an Ensemble Kalman Filter
PAPER WITHDRAWN
Clifford Mass, Univ. of Washington, Seattle, WA; and G. J. Hakim, P. Regulski, and R. Torn
Lack of observational data over the eastern Pacific can lead to forecast failures over the western United States, with obvious implications for public safety and economic activities. With its wide coverage, lightning data is a promising dataset for filling gaps in observational data over the eastern Pacific and elsewhere. This paper describes the assimilation of Vaisala's National Lightning Detection Network (NLDN)/Long Range Lightning data stream into an Ensemble Kalman filter (EnKF) using the Weather Research and Forecasting (WRF) model, with the goal of improving analyses, initializations and forecasts.
Experiments include a control simulation that assimilates observations from surface stations, soundings, aircraft observations and cloud track winds during a December 2002 extratropical cyclone event over the eastern Pacific as well as experimental simulations that also include the assimilation of lightning data. Lightning rates are converted into a precipitation rate using a lightning rate-rainfall relationship, which then is assimilated into the model as convective rainfall rates. Data is assimilated every 6 hours over a two-week period with 48-h forecasts starting at 00 and 12 UTC.
Comparisons are made of the SLP, 500mb height and 850mb temperature fields between the control, experimental and GFS analysis during the same time period. Results show that the assimilation of lightning data into the EnKF improves the positioning and intensity of the cyclone in the analyses. 12 and 24 hour forecasts also show promising improvements over the control. Intensities and locations of the forecasted precipitation fields are also considered.
Session 7, Assimilation of Lightning Data into Numerical Models
Wednesday, 23 January 2008, 1:30 PM-2:30 PM, 222
Previous paper Next paper