4th Ensemble Workshop Laurel MD May 2008s Overview of some operational issues related to EPS data Richard H. Grumm National Weather Service Office State College, PA 16803 Contributions from : Larry Struble, Pittsburgh Josh Korotky, Pittsburgh Chris Mellow Cleveland Justin Arnott Binghamton David Bright SPC 4th Ensemble Workshop Laurel MD May 2008s OVERVIEW • Training Issues, experiences and initiatives – NWS – Eastern Region – And beyond • The Future – new data and forecasting – NAEFS/GEFS in GFE – SREF GFE • Real world ensemble applications – Recent winter of 2008 – Real cases affecting real forecasters 4th Ensemble Workshop Laurel MD May 2008s Training Issues • COMAP-2008 – Training on ensembles for new SOO’s – 2 full days from chaos to spaghetti • State College-Pittsburgh – Sub-regional workshop in State College – 5 offices attended locally BUF-BGM-PHI- WBC-CTP – Remote offices Go-to-Meeting ER/SSD-BTV- OKX-LWX and CWSU Leesburg 4th Ensemble Workshop Laurel MD May 2008s Local Workshop 5-6 May 2008 • Requested by CTP operations team – Learn how to use the data better – Become more knowledgeable in these data • Topics – Chaos and uncertainty from Mary Baxter (Central Michigan), Jim Hansen (Monterey), and Josh Korotky (WFO-PIT) – Products and forecasting –David Bright (SPC) – Basic products and statistics –Richard Grumm (CTP) – Local modeling and ensemble efforts- Justin Arnott and Mike Evans (Binghamton) 4th Ensemble Workshop Laurel MD May 2008s Personal Experiences Training and use of EPS data • Effort Growing world wide – China, Africa, South America – Involvement in these activities provides cases and contacts (the high touch side of it all). – Local WFO’s and Universities • Could be short window of opportunity – Exploit these data now – Training in Universities is slowly evolving as I learned visit St Louis University April 2008 4th Ensemble Workshop Laurel MD May 2008s Future and Data NWS perspective • Importing NAEF/GEFS data into GFE (CTP) – 2m temperatures and winds – Add value to the forecast – Skill comparable to GMOS. – Data provided by NCEP (Yuejian Zhu) • SREF data into GFE (WFO-PIT) – Assists in complicated forecast problems – NAEFS/GEFS in GFE – SREF GFE 4th Ensemble Workshop Laurel MD May 2008s NAEFS Data 5km downscale EPS data aid in forecasting winds and temperatures 4th Ensemble Workshop Laurel MD May 2008s SREF Probability of Rain is copied into Potential grid (PotRainShowers grid) 4th Ensemble Workshop Laurel MD May 2008s 3 hr PoP grid from SREF • PoP input to a local PoP grid can be from forecaster created PoP or SREF PoP grids • This figure shows the SREF PoP grid 4th Ensemble Workshop Laurel MD May 2008s Winter Precipitation Type Grids Procedure is especially useful when creating complex winter weather grids SREF PoP and SREF precipitation types can be easily merged into resultant weather grid Weather grid created from SREF precipitation type and PoP are coherent in space and time 4th Ensemble Workshop Laurel MD May 2008s SREF Smart Init Probability Snow SPC Probability Snow 4th Ensemble Workshop Laurel MD May 2008s Probability ZR SPC Probability ZR 4th Ensemble Workshop Laurel MD May 2008s Real world ensemble applications • Proving the value of EPS data to users – Around the world • China cold outbreak in January 2008 • Korean floods August 2007 – Within the NWS • Cases for impacted offices • Recent winter of 2008 provided many good cases to facilitate training, cases studies, and applications. 4th Ensemble Workshop Laurel MD May 2008s Severe Weather 5-6 Feb 2008 Figure Storm Prediction Center (SPC) storm report by type for the 24 hour periods ending 0800 CDT on 6 and 7 February 2008. Reports are color coded by severe weather event type. http://www.spc.noaa.gov/climo/reports/ Courtesy of the SPC. 4th Ensemble Workshop Laurel MD May 2008s Snow fall Image 4th Ensemble Workshop Laurel MD May 2008s Surface and QPF 4th Ensemble Workshop Laurel MD May 2008s SREF 04/0300 UTC has the big rain too 4th Ensemble Workshop Laurel MD May 2008s Korean Flood August 2007 • Ideal application of ensembles – Precipitation amounts (Probabilities) and timing provided excellent guidance of a significant event. – The probability of extreme amounts was high a clear sign to be cautious and aware of a big if not record event. – Meteorological setting • Intensity of key features associated with heavy rainfall. • Key features associated with heavy rains events • Climate anomalies of key features adding confidence to the forecasts and put the event into a meteorological context thus closing the loop. 4th Ensemble Workshop Laurel MD May 2008s Estimate Rainfall valid 0000 UTC 10-14 August 2007 North Korean Flood GEFS forecast up to 14 inches of rainfall! 4th Ensemble Workshop Laurel MD May 2008s PROB 100mm 48 hrs 4th Ensemble Workshop Laurel MD May 2008s Pattern Change in China • Xu Xuan Jia Shenyang Central Meteorological Observatory, 11001 • Examination of cold and snow in China January 2008 – Warm pattern turned very cold – Well forecast by Global Ensembles. – Good demonstration Case studies 4th Ensemble Workshop Laurel MD May 2008s Comparative Forecasts 4th Ensemble Workshop Laurel MD May 2008s Midwest Floods 18-19 March 2008 • Big rainfall event – Most of it fell over 36 hours – Over the central Mississippi and Ohio Valleys • Was particularly well forecast – Begs the question why? – Some extreme rain amounts for a cold season event. – Some large anomaly signals in the pattern too. 4th Ensemble Workshop Laurel MD May 2008s Rainfall with the event Figure 1. Observed Precipitation (mm) from the unified precipitation data set showing a) storm total precipitation from 1200 UTC 17 March through 1200 UTC 19 March 2008 and b) 24 hour precipitation for the period ending at 1200 UTC 19 March 2008. 4th Ensemble Workshop Laurel MD May 2008s Short range QPF for 3 inches Figure 14. As in Figure 12 except GEFS initialized at (left) 1200 UTC 17 March and (right) 1800 UTC 17 March 2008. 4th Ensemble Workshop Laurel MD May 2008s The Pattern as forecast Figure 1. As in Figure 9 except showing PW forecasts. Upper panels show each members 25, 12.5 and 6.75 mm contour and the spread about the mean. Lower panels show the ensemble mean and the standardized anomalies of the ensemble mean 4th Ensemble Workshop Laurel MD May 2008s Case Studies • Bring ensembles impact to users – Prove they can work any where then they can work here – Local cases get local offices interested • Louisville is doing a case Study on the Heavy rain events • LaCrosse on the Snow events – Western Region on the Big Jan 2008 winter storm (Salt Lake and Reno). • Case studies – Great form of training and learning – Help users see the value and increase interest in using these data – And you get to interact with fun people – David Bright offers to help anyone do severe cases at the end of all of his talks! 4th Ensemble Workshop Laurel MD May 2008s Review • Training – There is a lot going on and it is getting better – Interest is growing – Case studies are a great training tool and incentive. • The Future – NWS getting data into GFE to help with forecast problems • NAEFS/GEFS in GFE • SREF GFE winter weather problem is but one – Will evolve more probabilistic forecast outcomes and products it is inevitable and it will be unstoppable. • Real world ensemble applications – Case studies of big and significant events sparks users interest and will grow the use of ensembles 4th Ensemble Workshop Laurel MD May 2008s Acknowledgements • Mentors Yuejian Zhu and Jun Du • Input and ideas – ER/SSD – All those listed in the beginning who made contributions Larry Struble, Pittsburgh, Josh Korotky, Pittsburgh, Chris Mellow Cleveland, Justin Arnott Binghamton, David Bright SPC • Others for great support and encouragement Zoltan Toth, Louis Uccellini, Robert Hart, and several individuals in the WMO and my local MIC Bruce Budd.
Pages to are hidden for
"Ensemble Prediction System Products_ Interpretation_ and forecast "Please download to view full document