EWP Blog – Page 328 – Spring Experiments & Beyond

Week 2 CASA Summary: 19-23 April 2010

Posted in Weekly Summaries on April 23, 2010 by Greg Stumpf.

For the CASA Hazardous Weather Testbed during the week of April 19th – 23rd the emphasis was on continuing to improve our procedures and process while maintaining readiness for possible weather events. As it turned out we did have interesting weather enter the IP1 testbed area on the night of April 22nd – 23rd, but only marginally severe thunderstorms. No significant circulations were observed. We added some new tools to our repertoire including a screen capture to Twitter using a Twitpic procedure on the WDSII console (technical issues prevented it from working on the AWIPS terminal). We also built a blogging tool accessible at http://casaradar.blogspot.com that captures information on all emails sent to response@casa.umass.edu (such as forecasts) as well as all tweets sent out by the casaradar and casaalert accounts. The blogger can also be sent email which will post at blogger@casaradar.com. The CASA HWT Protocols document was updated based on these new tools and procedures as well as experience gained during the previous week of the HWT. On Tuesday we verified all the functionality during a test run using the April 2nd case study. Wednesday had the potential for storms to enter the testbed after 8PM so we worked on further refinements to tools and procedures during the afternoon. The weather didn’t cooperate so we called it quits around 10PM that day.

On Thursday while we waited for the weather to develop (there were possible testbed incursions in early evening that did not materialize) Don and Steve ran through one of Don’s cases. During the evening hours, Steve ran both the WDSII and AWIPS terminals with Don’s help on setting up and looking at some of the wind products (3DVar, dual-dopppler) and “warn-on-forecast” NWP on web browsers on one monitor of the WDSII terminal. Rachel was the communicator on NWSchat, and monitored spotter locations (including Cedar who was out chasing with EM James Nimrod), and took snapshots of the WDSII and also from Weatherscope for inclusion on Twitter using Twitpic. Westy was monitoring all the tools and CASA system status – everything important went smoothly, although we did lose wireless connectivity to half the laptops (thankfully not the communicator one) around midnight that was fixed by Don’s expert network hacking skills. Despite the storm never really producing anything resembling a tornado or interesting circulation or winds we stuck with it until almost 4AM. On Friday we debriefed and recovered. Don and Steve ran through more of Don’s case.

We appreciate all the volunteer work done by OU folks during the week.

Timeline

Monday: Revised HWT procedures for the week.
Tuesday: Continued work on HWT procedures; worked on new functionality for communicator and technical issues with hardware and software; built CASA blog, with connections to two CASA twitter accounts and an email portal at blogger@casaradar.com Ran through CASA Case Study with Steve and Rachel as communicator.
Wednesday: Continued work on new functionality for communicator. On standby from 3PM on in case there was weather in the testbed. Nothing happened.
Thursday: Don and Steve worked on Don’s case in the early evening. Storm event started around 10PM and ended at 4AM.
Friday: Debrief at noon, Don and Steve continued working on Don’s case.

People
CASA personell: David Westbrook (M-F), Rachel Butterworth (T-F), Brendan Hogan (M-W), Don Rude (Th- Fri), Cedar League (Th-F), Brenda Phillips (M), Ellen Bass (T), Jerry Brotzge (OU)
Forecaster: Steve Hodanish, WFO Pueblo CO (Steve.Hodanish@noaa.gov)
Additional OU support from: Patrick Marsh (NSSL), Greg Stumpf (MDL), Darrel Kingfield (WDTB)

Jerry Brotzge, Univ., Oklahoma CASA Project Scientist

Week 2 EWP Summary: 19-23 April 2010 (more later from CASA)

Posted in Weekly Summaries on April 23, 2010 by Greg Stumpf.

Week #2 of EWP2010 wrapped up with continued PARISE and CASA experimentation. During the week, we hosted the following National Weather Service participants:

CASA: Steve Hodanish (WFO Pueblo, CO)
PARISE: Doug Cain (WFO Midland/Odessa, TX), John Cockrell (WFO Amarillo, TX), Andrea Lammers (WFO Louisville, KY), Brian Montgomery (WFO Albany, NY)

We managed to get some real-time weather in the Central Oklahoma domain last week, but it was non-severe and occurred overnight (2-3am). CASA data collection was had by their participants, and more information will be provided in their project weekly summary.

With a new set of four NWS forecasters, the PARISE experiment repeated their exercise from Week 1. As in Week 1, there were no real-time PAR data collection opportunities, so all exercises were conducted with archive case data sets.

Greg Stumpf, EWP2010 Operations Coordinator

Forecaster Thoughts – Brian Montgomery (2010 Week 2 – PARISE)

Posted in Forecaster Thoughts on April 23, 2010 by Greg Stumpf.

I wanted to re-express my gratitude to you, Dr. Pam Heinselman and the remainder of the team for an amazing week with PAR! I am confident these experiments will prove the future of this technology will not only be revolutionary, but provide faster life saving information. Here are some additional thoughts:

1. The experiment was well coordinated with an exceptional set of enthusiastic graduate students to make this experience quite rewarding. Recommendation would be add one more full day of activities. This would allow for forecasters to become more familiar with the software, visit with SPC and Norman Forecast Office, and perhaps increase the potential to catch a severe weather episode (assuming another experiment during convective season).
2. Forecaster fatigue was discussed during the experiment since PAR offers a high frequency of updates. While we all strive to get as much lead time as possible, the human factor of mental breaks is a necessity. At the time of writing this, during VCP12, we have become accustomed to 4-5 minute update intervals. I would recommend an experiment of working a full event to test the “breaking point” when a forecaster would need to step away for that mental break.
3. While we had some software glitches during our visit, WDSS-II does offer a promise of where this display technology is going. I was delighted to see the enhancements from the original display and our current legacy D2D AWIPS software. Some integration of software technologies including GR2Analyst may provide additional flexibility with PAR.
4. I was pleased to see the emphasis on base data! While algorithms can aid in where forecasters should focus, this is usually “after the fact” and it does take away from the warn-on-forecast philosophy. In fact, with PAR, the concept of these algorithms will likely grow beyond TVS’s, MESO’s and Hail to the future “Warn On Forecast” as presented to us by David Stensrud (and Dustan Wheatley). This future research is rather exciting as meteorologists, Emergency Managers, media and decision makers can provide better prognosis in a rapidly changing environment. Now all we need is for computational Moore’s Law to catch up to provide this enhanced awareness.

Again, a phenomenal opportunity to work with everyone as I would be honored to continue where we left off back in April.

Brian Montgomery (Lead Forecaster, NWS Albany NY – 2010 Week 2 Evaluator)

Week 1 PARISE Summary: 13-16 April 2010

Posted in Weekly Summaries on April 16, 2010 by Greg Stumpf.

The goal of the 2010 PARISE is to gain an understanding of the impact of temporal sampling on warning decision making, and warning lead time. During the first week, 13–16 April 2010, four NWS forecasters from the Southern Region, Eastern Region, and Central Region helped us address this goal by applying their warning decision expertise to five different playback cases sampled by the National Weather Radar Testbed Phased Array Radar (NWRT PAR). The forecasters worked each case in teams of two.

Before each case, forecasters developed situational awareness of the forcing mechanisms and near-storm environment in which the storms developed. This situational awareness allowed them to form a conceptual model of the storm type and severe weather threats they anticipated. Then they applied their warning decision expertise to interrogate the NWRT PAR data using the Warning Decision Support System – Integrated Information, and to issue warnings using a WarnGen tool similar to the one in AWIPS. After each case, the teams discussed their warning decision process with a PAR scientist. The case culminated with an overview of any severe weather that occurred, so that forecasters could self-evaluate their warning process.

The experiment wrapped up with a group discussion on participant’s experiences with NWRT PAR data and the PARISE as a whole. Forecasters said that they had a good experience and will encourage others at their office to participate next year. They enjoyed getting to work with rapid update data and experiencing how those data may change their current warning decision process. Forecasters also enjoyed getting to work with researchers and people from other NWSFOs.

Following analysis of the data collected during PARISE 2010, findings will be shared initially via conference papers in Fall 2010.

Pamela Heinselman, NSSL, PARISE Project Scientist

Week 1 CASA Summary: 12-16 April 2010

Posted in Weekly Summaries on April 16, 2010 by Greg Stumpf.

A primary goal for CASA during 2010 is to document the improvement in warning operations through the use of enhanced forecaster-emergency management communications. The use of NWS personnel in the HWT allows us to test and refine how best to utilize CASA’s high-spatial (100 m) and temporal (1-minute) data, in coordination with field emergency managers (EMs). Several particular areas of interest include the concepts of data overload, multi-sensor products, and the development of warn-on-forecast products and display.

The lack of severe weather in Central Oklahoma during Week #1 allowed CASA to refine its warning and communications operations. Les Lemon (WDTB) spent much of the week with CASA PIs Brenda Philips and Ellen Bass refining the communication technologies to be used between the forecaster and EM. Twitter and NWS Chat will be the primary means for communication, with additional use of Skype by some EMs and spotters. Les spent Monday getting trained, learning about the latest products from CASA (e.g., new NWP forecasts available every 10 minutes!), and refined products (real-time 3DVAR analysis). Tuesday through Thursday was spent reviewing and revising the interaction among the CASA forecasters and EMs. Several EMs and spotters visited the HWT and practiced with various communication technologies. Les spent Wednesday morning with Jerry Brotzge, Keith Brewster, Brenda, and Ellen in discussions on ways to improve the presentation of output from the real-time forecasts – in essence, how do we present “Warn-on-Forecast” information to forecasters already overwhelmed with data? Friday ended with precipitation in the HWT – an opportunity to review CASA products one last time…but now with echoes!

Jerry Botzge, Univ. Oklahoma, EWP2010 CASA Scientist

Forecaster Thoughts – Michael Scotten (2010 Week 1 – PARISE)

Posted in Forecaster Thoughts on April 16, 2010 by Greg Stumpf.

I want to thank Pam, Greg, Daphne, Heather, and others for giving me this great opportunity to use and evaluate PAR data. This was an awesome experience! I truly believe this technology will help meteorologists make better warning decisions in the near future.

The higher temporal PAR data with 40 to 80 second updates of the lowest elevation slices appear to be the biggest advantage for NWS forecasters. This will definitely change future warning decisions. Meteorologists will be able to detect microscale evolution of hooks, bows, and appendages to quickly pinpoint tornado spinnups and microburst winds. The tropical case event with rapid tornadogenesis stood out as the most compelling argument for higher temporal data as the higher temporal PAR data caught each tornado occurrence, while the current much lower temporal WSR-88D data did not. As a result of using higher temporal data, meteorologists will likely uncover more small scale phenomena such as tornadoes, especially weak ones, and microburst winds. Not to mention, faster detection of tornadoes and damaging winds will occur as well. These benefits will help us better understand severe local storms.

There may be a few possible downfalls with the higher temporal PAR data. For some radar operators, the higher temporal data may be overwhelming, particularly in quickly evolving weather situations. Also, the uncovering of more small scale phenomena may skew tornado and damaging wind climatologies upwards. When higher temporal data first arrives to local forecast offices, false alarms may increase with more warnings issued as more phenomena get discovered. However down the road, possibly fewer warnings and much smaller spatial areas for warnings will likely unfold, leading to better customer service.

I also enjoyed working with new and creative PAR sampling adaptive strategies. In particular, I preferred the Oversampled_VCP_within_120km_only scanning strategy with the quickest update times of lowest elevation slices. New scanning strategies will lead to better sampling of varying weather phenomena. For example, perhaps in the near future, differing scanning strategies could be used for detecting mesocyclones with discrete supercells that specialize on interrogating one or two cells compared to a wider sampling for several mesoscyclones embedded within a large scale QLCS. This will only help the NWS further fulfill its mission of saving lives.

We need this technology now!

Michael Scotten (Lead Forecaster, NWS Memphis TN – 2010 Week 1 Evaluator)

EWP2010 Week 1 Underway

Posted in General on April 15, 2010 by Greg Stumpf.

Week 1 of the EWP is underway. PARISE is chugging along with their four forecasters and collecting data on archive cases. Because each week the participants will be reviewing the same cases, we won’t be providing any case information here (or will spoil the secrets!). The CASA project has been in a shakedown this week, and the PIs have been adding a few new details to the project, which will soon be revealed in the CASA prject plan and briefing to be posted soon on the EWP2010 page. Les Lemon has been very helpful during the week, and will probably carry over into Week 2 to serve as additional help.

Weather wise in Central Oklahoma has been very quiet, although some convection is anticipated this afternoon and evening. Severe weather is unlikely. Due to PARISE’s unique schedule and experiment requirements, they will not be observing real-time data this evening. However, the CASA folks will be observing the data in a “low key” sense and using the opportunity to continue to shakedown the technology and concept of operations.

We will post end-of-week summaries for both PARISE and CASA on Friday or Saturday.

Greg Stumpf, EWP2010 Operations Coordinator

Almost ready for EWP2010 “Phase I”

Posted in General on April 8, 2010 by Greg Stumpf.

We are only 4 days away from beginning “Phase I” of the EWP2010 spring experiment. Phase I consists of our PAR and CASA experiments. We’ve had some successes over the past 2 weeks:

1. The PARISE shakedown on 4/6 went well with guest evaluators from the WFO OUN and the WDTB.
2. The WDSSII version of WarnGen is almost ready to go. It looks and feels a lot like AWIPS WarnGen.
3. The HWT AWIPS and ORPG have been upgraded. Bonus: CASA data are now displayable in AWIPS in the HWT!
4. The participant selections for all 9 weeks have been made and will be entered into the EWP Google Calendar.

We will be welcoming our first set of evaluators next week. They are:

CASA: Les Lemon (WDTB)
PAR: Mark Bacon (ILM); Jim Caruso (ICT); Jeff Cupo (FAATC-OKC); Mike Scotten (MEG)

One note: do not expect to see many details on the blog during Phase I of EWP2010, unless there are real-time events in Central Oklahoma. The evaluators will be going through a number of archive cases, and we don’t want to reveal details about those cases to participants who will be coming in future weeks. In addition, the PAR experiment is being conducted as a social science experiment, and much of the data are considered confidential.

Greg Stumpf, EWP2010 Operations Coordinator

2010 Planning Underway

Posted in General on March 9, 2010 by Greg Stumpf.

With the first significant tornado of 2010 hitting Oklahoma yesterday, it’s time to resurrect the EWP Blog. We are underway getting the EWP spring experiment going for 2010. This year we will be looking at a lot of new data sets, and the experiment period is longer than ever (9 weeks). We are also planning double the participation of NWS folks this year. Here is the experiment schedule (by week):

Phase I:

12 Apr – 16 Apr: PAR, CASA
19 Apr – 23 Apr: PAR, CASA
26 Apr – 30 Apr: PAR, CASA
3 May – 7 May: CASA
10 May – 14 May: CASA

Phase II:

17 May – 21 May: GOES-R, LMA, MRMS
24 May – 28 May: GOES-R, LMA, MRMS
31 May – 4 June: No operations (Memorial Day week)
7 June – 11 June: GOES-R, LMA, MRMS
14 June – 18 June: GOES-R, LMA, MRMS

In addition, during the latter half of the experiment, we may be ready to introduce the participants to some of the early radar data assimilation work being done for the Warn-On-Forecast program.

More information about each project is available here: http://ewp.nssl.noaa.gov/2010plans.pdf

The invitation for participants has been sent to all six NWS Regions and we expect to make our decisions by 29 March.

Check back here for more updates.

Greg Stumpf, EWP 2010 Operations Coordinator

The EWP2009 Thank You Post

Posted in General on June 23, 2009 by Greg Stumpf.

After we wrapped up daily operations in the HWT on 12 June 2009, Norman got hit by a weak tornado. While only minor damage occurred, and there were no injuries and deaths, what a way to end our 6 week experiment! This is this year’s EWP Thank You post, expressing our gratitude to the many participants of the Experimental Warning Program’s 2009 spring experiment. This year’s experiment was just as successful as the 2007 and 2008 experiments, and it could not have been carried out without the hard work and long hours of our team of participants.

The biggest expression of thanks goes to our IT Coordinator, Kevin Manross, who put in more hours than anyone else to pull off the experiment. As you will see below, Kevin wore many hats again this year.

Next, we’d like to thank our primary- and co-Weekly Coordinators for keeping operations on track each week: Kiel Ortega, Dale Morris, Jim LaDue, Patrick Burke, Travis Smith, Liz Quoetone, Paul Schlatter, Greg Stumpf, and Kevin Manross.

The cognizant scientists brought their expertise to the experiment to help guide live operations and playback of archive cases for each of the experiments.

For the Multi-Radar/Multi-Sensor (MRMS) application experiment, they included NSSL/CIMMS principle investigators Travis Smith and Greg Stumpf, with additional help from NSSL scientists Arthur Witt and Kevin Manross.

For the Lightning Mapping Array (LMA) experiment, they included NSSL/CIMMS scientist Kristin Kuhlman who was the principle investigator, along with visiting scientists Geoffrey Stano (NASA-Huntsville) and Eric Bruning (Univ. Maryland/NESDIS).

For the Phased Array Radar (PAR) experiment, Dr. Pamela Heinselman captained the ship, along with these folks from NSSL, CIMMS, and OU: Dave Preignitz, Ric Adams, Arthur Witt, Rick Hluchan, Adam Smith, and Jennifer Newman.

For the Collaborative Adaptive Sensing of the Atmosphere (CASA) experiment was again led by Brenda Phillips (U. Mass.), Jerry Brotzge (OU), and Ellen Bass (U. VA). In addition, we had help from Don Rude (U. VA), David Westbrook (U. Mass.), Cedar League (Univ. Colorado – Colorado Springs), Rachel Butterworth (OU), Corey Potvin (OU), and Vivek Mahale (OU).

We had NSSL IT help from Jeff Brogden, Robert Toomey, Charles Kerr, Villiappa Lakshmanan, Vicki Farmer, Karen Cooper, Paul Griffin, Brad Sagowitz, Brian Schmidt, and Joe Young.

We were also graciously provided AWIPS help from NWS Warning Decision Training Branch scientists Ben Baranowski and Darrel Kingfield.

There were a number of guest evaluators from the NWC that provided expertise: From WDTB, Les Lemon and Veronica Davis, from SPC/CIMMS (and the GOES-R Proving Ground), Chris Siewart, and from Florida State University, Scott Rudlosky.

Undergraduate students who supported our SHAVE efforts were: James Miller (coordinator), Anthony Bain, Jessica Erlingis, Steve Irwin, Erika Kohler, Tiffany Meyer, Corey Mottice, Nicole Ramsey, and Brandon Smith.

The EWP leadership team of Travis Smith and David Andra, along with the other HWT management committee members (Steve Weiss, Jack Kain, Mike Foster, Joe Schaefer, and Jeff Kimpel), and Dr. Stephan Smith, chief of the MDL Decision Assistance Branch, were instrumental in providing the necessary resources to make the EWP spring experiment happen.

Finally, we express a multitude of thanks to our National Weather Service and international operational meteorologists who traveled to Norman to participate as evaluators in this experiment (and we also thank their local and regional management for providing the personnel). They are:

Steve Cobb (WFO Lubbock, TX)

Suzanne Fortin (WFO Pleasant Hill, MO)

Gino Izzi (WFO Chicago, IL)

Jeff Michalski (WFO Seattle, WA)

Tom Ainsworth (WFO Juneau, AK)

Chris Wielki (Environment Canada, Edmonton, AB)

Rebecca Schneider (Environment Canada, Montreal, QC)

John Billet (WFO Wakefield, VA)

Kevin Brown (WFO Norman, OK)

Steve Hodanish (WFO Pueblo, CO)

James Cummine (Environment Canada, Winnipeg, MB)

Sarah Wong (Environment Canada, Toronto, ON)

Matthew Kramar (WFO Sterling, VA)

Mike Vescio (WFO Pendleton, OR)

Rob Handel (WFO Peachtree City, GA)

Bill Martin (WFO Glasgow, MT)

Pete Wolf (WFO Jacksonville, FL)

Bill Ward (NWS Pacific Region HQ, Honolulu, HI)

Daniel Nietfeld (WFO Omaha, NE)

Gail Hartfield (WFO Raleigh, NC)

Steve Kieghton (WFO Blacksburg, VA)

Dan Miller (WFO Duluth, MN)

Jenni Rauhala (Finnish Meteorological Institute, Helsinki)

Many thanks to everyone, including those we may have inadvertently left off this list. Please let us know if we missed anyone. We can certainly edit this post and include their names later.

Greg Stumpf (EWP Operations Coordinator)