Emergency Manager Severe Weather Information Needs and Use of Experimental Warning Information Daphne LaDue, Ph.D. OU CAPS Sean Ernst OU SoM James Hocker Oklahoma Climatological Survey Christopher Karstens, Ph.D. OU CIMMS/NSSL James Correia, Jr., Ph.D. OU CIMMS/SPC Jonathan Wolfe NOAA NWS Charleston WV Forecast Office This work was funded by NOAA/NSSL and OU CIMMS
19
Embed
Emergency Manager Severe Weather Information Needs and Use of Experimental Warning Information Daphne LaDue, Ph.D. OU CAPS Sean Ernst OU SoM James Hocker.
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Emergency Manager Severe Weather Information Needs and Use of
• 1 state level, ESF 8• 1 military• 1 EM for school district
PART I FINDINGS:STORM HISTORY
CIT stories revealed that storm history gives EMs a better idea of what to expect in their community:
Need to know what storm has done, expected strength changes — EM5
“In real time I was able to redirect [medical response resources], ‘cause I have the latest, greatest that the National Weather Service is providing…” —EM8
Storm history info and track help give 1-1.5 hours heads up on inbound storms and potential impact —EM4
“Getting to know that storm a little better” —EM9
PART I FINDINGS:RELATIONSHIP1. Specific events build relationships —EM7
• Forecaster on duty didn’t realize impact of sub-severe storm and EM needs related to impacts
• EM used existing relationships to solve problem• Built longer-term understanding of EM information needs
2. Building open lines of communication
• EMs know they can call NWS when they need to• NWS might even reach out prior to event
3. Knowing NWS forecasters creates trust in forecast information
Knowing NWS forecasters builds trust in information —EM5
“If I see things that concern me, I’ll either start chatting or get on the phone with my local NWS and ask them if they’re going to put a warning out” —EM2
PART I FINDINGS:CONFIDENCE1. EM’s understand forecasts carry uncertainty, and they’d like
to hear about forecaster opinions on it
2. Example of good information
Want confidence on threat timing and likelihood to shelter large events well in advance —EM4
“[NWSChat] gives me a certain level of confidence, and I find out what they really think too” —EM10
“It was high, I think it was high confidence, of, supercell development, into individual supercell development. Moderate to high confidence, that it will affect the metro, and then they gave the eta, like between 7 and 9 pm.” —EM11
PART I: PRELIMINARY OUTCOMES
1. EMs looking for the NWS story of a weather event
• Want the narrative of the storm as it unfolds• Want to know how the forecasters perceive the event
2. Want to build and maintain strong relationship with forecasters to build trust in forecast
3. Want forecaster’s insights into inherent uncertainty
• EMs aware that no forecast is exact, want to know forecaster’s honest assessment of forecast
PART II:HAZARDOUS WEA. TESTBED
Purpose: bring key stakeholder group into PHI development early in the R&D to assure resulting work is useful, usable
Methodology:
• Pre-week survey to establish current views of uncertainty
• EMs viewed PHI generated by NWS forecasters and noted decision/action points
• Researcher observations and questions
• Joint debriefing discussions after each case or live event
• End-of-week EM-only and joint w/NWS discussions
Traditional SVR warning polygon
PHI Object and SVR Plume
vs.
Polygon extends beyond echo behind and to the sides of the storm.Polygon forward spreads out in width.One polygon for tornado, wind, hail.
Object tightly surrounds intense part of echo.Plume forward spreads out in width.Separate objects for tornado vs. wind/hail.
HWT PARTICIPANTS
10 EMs from 5 states:Alabama (1)Michigan (1)Minnesota (1)Oklahoma (6)Wyoming (1)
Each object has an associated set of information (yellow box):
Contents of the discussion box evolved each week as forecasters and EMs interacted.
Time of departure
Time of arrival
FINDINGS:HAZARD PLUMES VS. TRADITIONAL WARNINGS
Advantages:
• PHI Focus
• Still need trigger points; EMs cannot devote 100% attention to weather
• Polygon Confidence of NWS of EM’s of others, too. “Confidence is contagious.”
The main difference: “Uncertainty” —Co4
Gives you what areas to focus on, and what areas likely won’t be affected —Co2
Helps identify which cells in a line might do something —St1
“...if you’re confident enough to...warn [x number of] people...maybe I should be certain, too.” —Co4
FINDINGS:PROBABILITY IS USEFUL.IS CHANGE MEANINGFUL?
Liked seeing the increases, decreases in probabilities
Changes after warning issuance could be meaningful
EMs: A few percentage change probably not meaningful, and may fluctuate too much.
• 10% was suggested,
• or have the forecaster tell decide what was is a meaningful increase or decrease for that day
“I’m at an 85%, where maybe the warning came out at a 60%, and that’s like, boom. It gives me a lot more information.” —Co5
FINDINGS:THIS IS MANAGEABLE
Initially some concern about the increase in information
Iterated w/ forecasters on discussion box contents
EMs started injecting information to the NWS.
Need spot forecast:Hail hitting hazardous chemical tanker truck, can’t take much more. Will the hail get any bigger? When will it stop?
Need spot forecast:College football team on a bus with bow echo heading toward the highway. How strong will winds be?
Need spot forecast:Water loading on factory roof, might collapse. How much longer will rain last?
“I think we need to send some realistic injects back to them so that the scenario works just a little bit better. They can...see some of the challenges that we have” —Co2
CO-CREATION OF WHAT PHI SHOULD BEAmazing dynamic — participants became researchers, asking each other insightful questions to understand the others’ point of view.
Each week iterated toward the same things:
• Discussion box to contain:
• history, such as reports• forecaster thinking, but not bland warning-type statements
• Forecaster touch critical; did not trust automated forecast
• Did forecaster agree? Or that they changed numbers with his/her expertise & knowledge beyond radar
NWS: “I’m not, I’m not completely, I’m not sold enough to drop the probabilities in light...of the reports we’ve gotten [and] how strong it was there for awhile.”
EM: ”Why didn’t you write that in the box?”
CONCLUSIONS
Critical Incident StudyEMs want the narrative of storm as it unfolds
Relationships need to be built and maintained
EMs want forecasters’ ongoing assessment, including uncertainty
Hazardous Weather Testbed PHI is more specific, focused useful for EM decisions
Still need trigger points for action; confidence of forecaster
On our Research in the HWT:
Presence of EMs gave forecasters focus & rapid feedback