Project Overview:
This was first week of our four-week spring experiment of the 2014 NSSL-NWS Experimental Warning Program (EWP2014) in the NOAA Hazardous Weather Testbed at the National Weather Center in Norman, OK. “The Big Experiment” or “Spring Experiment” had three components: (1) an evaluation of multiple CONUS GOES-R convective applications, including satellite and lightning; (2) an evaluation of the model performance and forecast utility of two convection-allowing models (the variational Local Analysis Prediction System and the Norman WRF); (3) and an evaluation of a new feature tracking tool from NASA SPORT. Additionally we coordinated daily with Experimental Forecast Program, participating in briefings and evaluating the probabilistic severe weather outlooks produced by their forecasters as guidance for our warning operations.
Participants:
The NWS guests were Jared Maples (WFO Grand Rapids, MI), Scott Rudge (WFO Rapid City, SD), and Bruce Thoren (WFO Norman, OK). Additionally, for the first time, the EWP is hosting a weather broadcaster to work with the NWS forecasters at the forecast desk. This week, the guest was Daniel Bickford of WSPA-TV (Greenville, SC). The GOES-R program office, the NOAA Global Systems Divisions (GSD), and the National Severe Storms Laboratory provided travel stipends for our participants from NWS forecast offices and television stations nationwide.
Visiting scientists this week included Hongli Jiang (GSD), Paola Salio (Ciudad Universitaria, Buenos Aires, Argentina), Justin Sieglaff (CIMSS), and Kris White (WFO Huntsville, AL).
Kristin Calhoun was the weekly coordinator, Gabe Garfield the back-up coordinator and Lance VandenBoogart (WDTB) was our “Tales from the Testbed” Webinar facilitator. Our support team also included Darrel Kingfield, Bill Line, Chris Karstens, Greg Stumpf, Karen Cooper, Vicki Farmer, Lans Rothfusz, Travis Smith, Aaron Anderson, and David Andra.
Daily Operations Summary:
Feedback on Experimental Products:
- Easily showed the lack of handle on the current situation of the model: if it didn’t match current observations would disregard that solution
- Useful product for visualization of moisture and instability
- Easily track of movement of moisture and instability gradients.
- Believe that more frequent analyses could be better, but current forecast frequency is adequate.
- Suggest filling in gaps in coverage with other model data, but highly believe there should be an option to turn on / off model interpolation so forecasters can see what is observations vs model data.
- Gain more knowledge / understanding of the product with increased use.
- “It’s ok.” In some regimes it was fantastic, in others not so great. (Surprised it did not work well on 8 May 2014 in Kansas without cirrus coverage).
- Need to better understand what when it works well, and when it doesn’t/
- Foresee utilization outside of severe convection (e.g., updating near-term POPs)
- Good tool for short-term use and initial storm development.
- Used tool as assistant to further interrogate the storm via all-tilts radar, etc.
- Watched trends in probabilities; self-calibrated by watching control (or first) storm to develop in environment and applied to other storms that day.
- Look forward to having different probabilities for tornado / wind / hail.
- See limitations for linear systems as it will grab long line and produce probability and max values for entire line. Would prefer to see see lines segmented more (possibly could be done once they get above a certain size, e.g., 50 or 100 km).
- Would like to see strong/ mod colorization of text in metadata via mouseover (e.g., MESH text move from white to yellow at 1 in).
- High probabilities failed on Thurs 8 May due to influence of high growth rates (strong glaciation) though MESH remained was low. This provided low confidence in values during the event.
- highlights features in imagery – useful where radar coverage is sparse or radar is down
- “Very cool” ; “Eye candy”
- Could easily pick out boundaries and associated movement
- Able to view the overshooting tops form and see shadows move (utilized more than CI or overshooting tops products)
- Provided a view of updraft strengthening
- Flash extent density picked out which storms growing the quickest and which were the strongest.
- Could see intensification and decay of multiple storms
- Not highly utilized this week due to day-to-day changes in product.
- Could see it being a useful tool to examine trends multiple products including gate-to-gate values and storm-top divergence.
- Much difficulty with initial runs immediately after re-centering domain.
- Could easily interrogated max reflectivity in model and compare w/current observations
- Visible artifacts (hot spots in humidity) from possible assimilation of bad observations from AWOS sites (more comparison QC needed?).
- For enhanced operational utility, suggest extending forecast to 6 hrs window. (1st hour forecast is generally “now,” thus only currently used as 2 hr forecast)
- For similar computational usage suggest updating only 30 min for longer forecast window, but still begin a new forecast every hour.
Contributors:
-Kristin Calhoun, Week 1 Coordinator
-Gabe Garfield, EWP Spring Experiment Coordinator