HFIP Socio-Economic Working Group/NWS Storm Surge Roadmap/NOS

Social Science ResearchHurricane Forecast Communication

-- Report to 2012 NOAA Hurricane Conference

HFIP Socio-Economic Working Group/NWSStorm Surge Roadmap/NOS

National Center for Atmospheric Research (NCAR)Eastern Research Group, Inc. (ERG)

Jeff Lazo, NCARBetty Morrow, SocResearch Miami for ERG

Lou Nadeau, ERGLinda Girardi, ERG

OBJECTIVES

• To determine the concerns and needs of key stakeholders regarding storm surge and storm surge forecast communication

• To assess support for a storm surge warning

• To evaluate several prototype surge inundation graphics

RESEARCH PLANEXPLORATORY STAGE• Develop, test, and refine storm surge prototype graphics

– One-on-one webinars, interviews with key stakeholders– Discussions with groups of EMs and NWS personnel– Booths at AMS Weatherfest and National Hurricane Conference

• Develop survey questions for key stakeholders– OMB approval– Pre-testing

QUANTITATIVE STAGE• Collect data via 6 surveys (including

add-ons to other surveys)

SURVEYS• EMs TC On-Line Survey (NCAR)

– Population: All EM directors in coastal counties/parishes, NC to TX

– Sample: n = 53, 45% response rate– High vulnerability to storm surge – both land and population– Highly experienced (average time in current position was 12 years)

• EMs ET TC On-Line survey (NCAR)– Population: All EM directors in coastal counties/parishes/

territories on Atlantic, Gulf, Pacific coasts, Alaska, and Hawaii – Sample: n = 113, 55% response rate– 90% vulnerable to TCs; 91% to ETs– Experienced (average time in current position = 7 years, and

averaged 5-6 TS activations, 3-4 hurricanes, 1-2 ET storms)

SURVEYS (continued)• Public ET TC Telephone Survey (NCAR) (Lazo & Gladwin)

– Atlantic, Gulf, Pacific, Alaska, Hawaii– Census blocks with some portion within .5 mile of coast & any area less than

20 ft. above mean sea level– RDD of last digits of phone numbers– N = 900, 59% completion rate (of 1525 screened)– 38% ET, 30% TC, both 32%

• Public ET TC On-line Survey (ERG) (Nadeau)– Internet users living within 50 miles of Atlantic, Gulf, Pacific, Alaska, Hawaii

coasts (surge questions asked of those living within 10 miles of coast)– Total Sample: n = 459, 51,000 emails, 84% cooperation rate– Surge Sample: n = 177

SURVEYS (Continued)• Broadcast Media ET TC online survey (NCAR)

– 4 major local TV stations (ABC, CBS, Fox, NBC)in coastal markets

– Atlantic, Gulf, Pacific, Alaska, Hawaii regions– 82% vulnerable to TC, 90% vulnerable to ET– n = 51, 42% response rate– Average 19 years in meteorology, 12 years in current position

• NWS WCMs On-Line Survey (ERG)– Population: WCMs from Eastern, Southern, Western,

Pacific, Alaska regions– Sample: n = 54, 77% response rate

CONCERNS• Public– Concerned about severe coastal storms

• Most concern about wind and tornadoes– Often not aware of surge and flooding potential

• EMs– Very concerned about storm surge– Don’t believe public understands its

risk• Media– Very concerned about wind and storm surge– Believe only about 1/3 of public understands its risk

ASSESSMENT OF SUPPORT FOR STORM SURGE WARNING

QUESTION TO TC EMS: The Saffir-Simpson Scale and current hurricane warnings are based on wind speeds. The National Hurricane Center has removed storm surge information from the Saffir-Simpson Scale since often there is not a direct correlation between wind and surge. This has led to an investigation of how surge information should be communicated. Below are some suggestions. Please indicate how much you agree or disagree with each of these options.

The NHC should:- issue storm surge information separate from wind information.- Issue a separate storm surge warning- Issue a separate storm surge watch

Agree Completely

Agree Somewhat Neutral

Disagree Somewhat

Disagree Completely

Issue storm surge information separate from wind

75.5% 11.3% 7.5% 3.8% 1.9%

Issue separate storm surge warning

67.9% 11.3% 3.8% 7.5% 9.4%

Issue separate storm surge watch

60.4% 17.0% 3.8% 7.5% 11.3%

n = 53

EM Survey – TC Regions (NC – TX)

Preference for surge depths expressed above GROUND LEVEL rather than SEA level

QUESTION TO ET-TC EMS: For tropical storms the Saffir-Simpson Hurricane Wind Scale and watches and warnings are based on wind speeds. Due to the lack of consistent correlation between wind and surge, storm surge information has been removed from the scale. This has led to an investigation of how surge information should be communicated for BOTH tropical and extratropical cyclones.

To what extent do you agree or disagree with these statements:The NWS should issue Storm Surge Watches.The NWS should issue Storm Surge Warnings.

Strongly Agree Agree Undecided Disagree

Strongly Disagree

Issue separate storm surge warning

50.0% 37.3% 2.7% 1.8% 8.2%

Issue separate storm surge watch

44.5% 38.2% 6.4% 2.7% 8.2%

n = 110

EM Survey – ET- TC Regions (Atlantic, Gulf, Pacific, Alaska, Hawaii)

QUESTION TO ET-TC PUBLIC ON TELEPHONE SURVEY: In addition to the current forecasts and warnings issued for severe coastal storms, the NWS is considering issuing a new coastal surge warning. One reason for it is that the current Coastal Flood Warning may not convey the potential danger adequately. One reason against it is that too many warnings can be confusing to the public.In your opinion should the NWS issue a separate surge warning for severe coastal flooding events?• 71.7% answered yes• 19.6% answered no• 8.8% don’t know

In Probit regression analysis, support was associated with being in TC region, home ownership, and having been previously impacted by coastal storm. Not associated with any other demographics.

QUESTION TO ET-TC PUBLIC ON-LINE SURVEY:

The category assigned to a hurricane (Categories 1 through 5) is based on its wind speed. The amount of surge, however, can vary among hurricanes with the same wind force. Therefore, NWS may decide to issue a SEPARATE warning for storm surge in addition to Hurricane or Tropical Storm Watches and Warnings.

Do you think it is a good idea for NWS to ISSUE A SEPARATE STORM SURGE WARNING for hurricanes or severe coastal flooding events?• 92.1% answered yes• 4.0% answered no• 4.0% not sure/don’t know

QUESTION TO ET-TC BROADCAST METS ON-LINE SURVEY:

For tropical storms the Saffir-Simpson Hurricane Wind Scale and watches and warnings are based on wind speeds. Due to the lack of consistent correlation between wind and surge, storm surge information has been removed from the scale. This has led to an investigation of how surge information should be communicated for BOTH tropical and extratropical hurricanes.

To what extent do you agree or disagree with the following options:• The NWS should issue Storm Surge Warnings• The NWS should issue Storm Surge Watches.

Strongly Agree Agree Undecided Disagree

Strongly Disagree

Issue separate storm surge warning

74.5% 19.6% 3.9% 2.0% 0

Issue separate storm surge watch

68.6% 21.6% 3.9% 5.9% 0

n = 51

EM Survey – ET- TC Regions (Atlantic, Gulf, Pacific, Alaska, Hawaii)

QUESTION TO ET-TC NWS WCMs ON-LINE SURVEY:

Storm surge forecasts have historically not included waves or tides. Given that they will now include tides and will likely include waves in the future, what should the warning product be called?

• Storm Surge Warning• Storm Tide Warning• Storm Surge Inundation Warning• Extreme Coastal Flood Warning

NWS WCMs On-Line SurveyExcellent Very Good Good Fair Poor

Storm Surge Warning

26.7% 20.0% 30.0% 6.7% 16.7%

Storm Tide Warning

6.7% 10.0% 13.3% 50.0% 20.0%

Storm Surge Inundation Warning

16.7% 23.3% 13.3% 13.3% 13.3%

Extreme Coastal Flood Warning

6.7% 10.0% 16.7% 36.7% 30.0%

SUMMARY OF POSITIVE SUPPORT FOR STORM SURGE WARNING• 76% EMs TC survey• 87% EMs ET-TC survey• 72% Public telephone survey• 94% Public on-line survey• 95% Broadcast media• 77% NWS WCMs (in favor of name)

Most support the label “Storm Surge Warning”

REASONS FOR STORM SURGE WARNING• Serious lack of public understanding of surge (all)• Too much focus by public on categories (EMs, Bmets)• Would increase public attention to surge (EMs)• Would result in greater emphasis in EM decisions• Would lead to better communication of surge threat in weathercasts (Bmets)

COUNTER ARGUMENTS:• Public familiar with current coastal flood warnings• Already too many warnings• Concern about NWS’s ability to provide reliable, timely information at necessary spatial and temporal

scale

ASSESSMENT OF PROTOTYPES

• Storm Surge Warning Area Map• Storm Surge Inundation Maps

STORM SURGE WARNING AREA Effectiveness:*

• 92% by EMs• 96% by Public• 96% by Media• 63% by WCMs

*Total of first three choices

If a warning for storm surge is issued, a map will show the AREA INCLUDED IN THE WARNING, something that is currently done for Hurricane and Tropical Storm Warnings. More detailed maps will be provided for localities. The color purple is used in the following example in order to avoid colors used for other NWs watches and Warnings.

How effective do you think a map like this would be in communicating to the public the area under storm surge or coastal flood warning?(Extremely effective, Very Effective, Effective, Somewhat Effective, Not Effective at All)

ASSESSMENT OF STORM SURGE INUNDATION MAPS

STORM SURGEINUNDATION MAP

STORM SURGEINUNDATION MAP

STORM SURGEINUNDATION MAPCriteria:Easy to understand*• 86% by EMs• 96% by Media• 77% by Public• 90% by WCMs

Provides useful information*• 84% by EMs• 94% by Media• 98% by Public• 83% by WCMs

*Total Excellent, Very Good, GoodThe majority in all surveys preferred this mapProblems with using “low” to describe storm surge hazardMay affect evacuation decisions at both extremes

INTERPRETATION BY PUBLICThe rates for evacuation intent from Cypress Lake (located in a High area) based on each of the three maps were:

• 68% with the solid blue map• 84% with the shades of blue map• 95% with the multicolored map

Using the multicolored map, they were then asked if they would evacuate from the Villas (located near border between Moderate and Low) and 54% said they would evacuate.

Providing potential depth information makes a difference, with those located at higher levels being more likely to evacuate than those at lower levels, even though the surge can be life-threatening.

ADDITIONAL FINDINGS• Maps should be:– High resolution– Interactive• Allow zooming (to some degree)•Mouse-overs with explanations

– Usable on smart phones, etc.

ADDITIONAL FINDINGS (continued)

• NWS websites need to be more user friendly

• WFO websites are under-utilized by public

• Too many NWS products and too much text

• Timing of storm surge information is too late for EMs

• Timing of forecast products too close to media broadcast times

SUMMARY• Strong support for Storm Surge Warning from

all stakeholder groups• Positive assessment of Storm Surge Warning

Area Map• Positive assessment of Storm Surge

Inundation Map using multiple colors to show categories of depth

After review, all six reports will be posted on websites, as well as one-page summaries.

This Powerpoint is available at:http://www.nws.noaa.gov/om/hurricane/hurr_conf-12.html

Please send comments or questions to:[email protected]

[email protected]@erg.com