Probabilistic Storm Forecasting for Traffic Flow Management: Progress and Challenges
7
Probabilistic Storm Forecasting for Traffic Flow Management: Progress and Challenges Friends and Partners of Aviation Weather NBAA Convention, October 9, 2008 Stephen S. Weygandt NOAA Earth System Research Laboratory
description
Probabilistic Storm Forecasting for Traffic Flow Management: Progress and Challenges. Stephen S. Weygandt NOAA Earth System Research Laboratory. Friends and Partners of Aviation Weather NBAA Convention, October 9, 2008. RCPF. Model-based storm guidance. Hourly updating model. RUC RR. - PowerPoint PPT Presentation
Transcript of Probabilistic Storm Forecasting for Traffic Flow Management: Progress and Challenges
Probabilistic Storm Forecasting for Traffic Flow Management:
Progress and Challenges
Friends and Partners of Aviation WeatherNBAA Convention, October 9, 2008
Stephen S. WeygandtNOAA Earth System Research Laboratory
Model-based storm guidance
ObservationsModel ensembles
Data Assimilation cycle
Model post-processing
RCPF
Probability products
HRRRHCPF
Storm-scale probabilityinformation
RUC RR
Hourly updating model
Radar Assim
13-km RUC & Rapid RefreshHourly update cycle gives accurate mesoscale environment, good convective initiation
RCPF (hourly 3-10 h outlook ) madefrom time-lagged ensemble of RUCRCPF used as guidance for CCFP
Mesoscale model guidance
6 hr fcst1 pm EDT
20 July 20089 am outlook
RCPF
4 hr fcst1 pm EDT
CCFP6-h CCFP
8-h RCPF
Outlook timeValid time
valid timestorm coverage
CS
I
4-km verificationagainst NCWD
(from RTVS)
Forecast
length
Verification: 1 June – 31 August 2008
Mesoscale model guidance
8 hr fcst3 pm EDT
RCPF
6 hr fcst3 pm EDT
CCFP
20 July 20089 am outlook
13-km RUC & Rapid RefreshHourly update cycle gives accurate mesoscale environment, good convective initiation
RCPF (hourly 3-10 h outlook ) madefrom time-lagged ensemble of RUCRCPF used as guidance for CCFP
6-h CCFP
8-h RCPFvalid time
storm coverage
CS
I
(from RTVS)
Outlook timeValid time
Forecast
length 4-km verificationagainst NCWD
Verification: 1 June – 31 August 2008
(3-km Hi-Res Rapid Refresh)High resolution needed for realistic storm structure (storm-types, gaps in lines, etc.)
Hourly 12-h forecast, 15-min VIL output
RUC radar assimilation improves HRRR
Storm-scale model guidance
radarHRRR
6 hr fcst8 pm EDT
Truth
8 pm EDT
20 July 20082 pm initial time
initial timestorm coverage
CS
I 3-km 20 dBZverification
Initial timeValid time
radarnoradForecast
length
6-h fcsts HRRR
Verification: 1 June – 31 August 2008
Storm-scale model guidance
noradHRRR
6 hr fcst8 pm EDT
Truth
8 pm EDT
initial timestorm coverage
CS
I 3-km 20 dBZverification
Initial timeValid time
radarnoradForecast
length
6-h fcsts HRRR
(3-km Hi-Res Rapid Refresh)High resolution needed for realistic storm structure (storm-types, gaps in lines, etc.)
Hourly 12-h forecast, 15-min VIL output
RUC radar assimilation improves HRRR
20 July 20082 pm initial time
Verification: 1 June – 31 August 2008
Model-based TFM guidance
HRRR ensembles ATM probabilities
HRRR
HRRR andHRRR-basedprobabilitiesImportant for
automatedATM
norad
HRRR forecastsradar
Radar assimilation improvesHRRR forecast through 12-h HRRR key component of CoSPAHRRR storm structure allowsextraction of ATM information C
SI
Forecast Length (hours)
3-km 30 dBZverification
1 June – 31 Aug 2008
