Emergency Managers National Science On-FoundationLine Survey on Extratropical and Tropical Cyclone Forecast Information: Hurricane Forecast Improvement Program/Storm Surge Roadmap

January, 2013

Betty Hearn Morrow Jeffrey K. Lazo

Research Applications Laboratory Weather Systems and Assessment Program Societal Impacts Program

NCAR Technical Notes

National Center for Atmospheric Research P. O. Box 3000 Boulder, Colorado 80307-3000 www.ucar.edu NCAR/TN-497 +STR

NCAR TECHNICAL NOTES http://www.ucar.edu/library/collections/technotes/technotes.jsp

The Technical Notes series provides an outlet for a variety of NCAR Manuscripts that contribute in specialized ways to the body of scientific knowledge but that are not yet at a point of a formal journal, monograph or book publication. Reports in this series are issued by the NCAR scientific divisions, published by the NCAR Library. Designation symbols for the series include:

EDD – Engineering, Design, or Development Reports Equipment descriptions, test results, instrumentation, and operating and maintenance manuals.

IA – Instructional Aids Instruction manuals, bibliographies, film supplements, and other research or instructional aids.

PPR – Program Progress Reports Field program reports, interim and working reports, survey reports, and plans for experiments.

PROC – Proceedings Documentation or symposia, colloquia, conferences, workshops, and lectures. (Distribution maybe limited to attendees).

STR – Scientific and Technical Reports Data compilations, theoretical and numerical investigations, and experimental results.

The National Center for Atmospheric Research (NCAR) is operated by the nonprofit University Corporation for Atmospheric Research (UCAR) under the sponsorship of the National Science Foundation. Any opinions, findings, conclusions, or recommendations expressed in this publication are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.

National Center for Atmospheric Research P. O. Box 3000 Boulder, Colorado 80307-3000 ISSN Print Edition 2153-2397 ISSN Electronic Edition 2153-2400

Emergency Managers On-Line Survey on Extratropical and Tropical Cyclone Forecast Information: Hurricane Forecast Improvement Program/Storm Surge Roadmap

Betty Hearn Morrow SocResearch Miami, Miami, FL

Jeffrey K. Lazo National Center for Atmospheric Research, Boulder, CO

NCAR/TN-497+STR NCAR Technical Note Published By: NCAR Library January, 2013

EMERGENCY MANAGERS ON-LINE SURVEY ON EXTRATROPICAL AND TROPICAL CYCLONE FORECAST INFORMATION: HURRICANE FORECAST IMPROVEMENT PROGRAM/STORM SURGE ROADMAP

Betty Hearn Morrow SocResearch Miami

Jeffrey K. Lazo Societal Impacts Program National Center for Atmospheric Research PO Box 3000 Boulder, CO 80307

January 31, 2013

ETTC Emergency Manager Survey

This page intentionally left blank.

NCAR Societal Impacts Program ETTC Emergency Manager Survey

CONTENTS

TABLES ______ii FIGURES ______iii ACKNOWLEDGMENTS ______iv EXECUTIVE SUMMARY ______ES-1 Purpose ______ES-1 Survey Development, Implementation, and Sample ______ES-1 Summary of Results ______ES-2 1. BACKGROUND ______1 2. METHODS ______4 2.1 Survey Development ______4 2.2 Sample ______4 2.3 Data Collection and Analysis ______6 3. RESULTS ______7 3.1 Regional Hazards Risk and Public Preparation ______7 3.2 Opinions Regarding Storm Forecast Products ______9 3.2.1 Storm Surge Warning ______9 3.2.2 Storm Surge Warning Area Map ______12 3.2.3 Storm Surge Inundation Maps ______14 3.2.4 Forecast Track Cone Maps ______19 3.2.5 Wind Map ______22 3.2.6 Wind and Cone Maps ______23 3.2.7 Arrival of Tropical Storm Force Winds ______25 3.3 Use of Forecast Services ______27 4, SUMMARY ______32 REFERENCES ______34 Appendix A. Advance Letter and Invitation Email Sent To Broadcast Meteorologists __ App. A-1 Pre-Contact Letter______App. A-2 Invitation Email ______App. A-3 Appendix B. Survey Questions and Summary of Responses ______App. B-1

NCAR Societal Impacts Program i ETTC Emergency Manager Survey

TABLES

Table 2-1. Location of Respondents ______5 Table 3-1. Incidence of Activations for Storms in Last 10 Years (%) ______8 Table 3-2. Level of Concern About Hazards (%) ______9 Table 3-3. Opinion about Storm Surge Watches and Warnings (%) ______10 Table 3-4. Opinions about Effects of Storm Surge Warning (%) ______11 Table 3-5. Preference for Storm Surge Labels (%) ______15 Table 3-6. Assessment of Combined Wind and Cone Map (%) ______24 Table 3-7. Assessment of Arrival to TS Force Winds Map (%) ______26 Table 3-8. Reasons for Using Commercial Vendors (%) ______27 Table 3-9. How Often Use Public Weather Sources (%) ______28 Table 3-10. How Often Use Internet Websites (%) ______28

NCAR Societal Impacts Program ii ETTC Emergency Manager Survey

FIGURES Figure ES-1. (A) Potential storm surge area in one color; (B) Potential storm surge depths in graduations of one color; (C) Potential storm surge depths in different colors ______ES-3 Figure ES-2. (A) Current tropical cyclone forecast cone; (B) Transparent forecast error cone with dashed lines; (C) Transparent forecast error cone with no lines ______ES-3 Figure ES-3. Potential for damaging winds ______ES-4 Figure ES-4. (A) Combined wind and cone; (B) Combined wind and cone – land only ______ES-5 Figure ES-5. Arrival of tropical storm force winds ______ES-5 Figure 2-1. Locations of sample ______6 Figure 3-1. Comparison of hazard risk and public vulnerability ______7 Figure 3-2a. Vulnerability to tropical storms Figure 3-2b. Vulnerability to extratropical storms ______8 Figure 3-3. Experimental graphic for showing area under storm surge warning ______13 Figure 3-4. Perceived effectiveness of storm surge warning map ______13 Figure 3-5. Potential storm surge area in one color ______16 Figure 3-6. Potential storm surge depths in graduations of one color ______17 Figure 3-7. Potential storm surge depths in different colors ______18 Figure 3-8. Map preference ______19 Figure 3-9. Current tropical cyclone forecast cone ______20 Figure 3-10. Transparent forecast error cone with dashed lines ______21 Figure 3-11. Transparent forecast error cone with no lines ______21 Figure 3-12. Preference for tropical cyclone forecast cone ______22 Figure 3-13. Potential for damaging winds ______23 Figure 3-14. Combined wind and cone ______24 Figure 3-15. Combined wind and cone – land only ______24 Figure 3-16. Arrival of tropical storm force winds ______26

NCAR Societal Impacts Program iii ETTC Emergency Manager Survey

ACKNOWLEDGMENTS

Special thanks to Jesse Feyen, Jamie Rhome, Jennifer Sprague, and Robert Berg for their leadership on this project and to Ethan Gibney for his work on the prototype graphics. We are also indebted to Keelin Kuipers, Mary Erikson, Julie Demuth, Crystal Burghardt, Jennifer Boehnert, and Christina Thomas for their assistance with this research, analysis, and report. We appreciate numerous others from the National Oceanic and Atmospheric Administration (NOAA) who contributed through conference calls and reviews. Most of all we are indebted to the emergency managers who took considerable time out of their busy schedules to complete the survey. This work was carried out in part with funding under award numbers NA06OAR4310119 and NA06NWS4670013 from the U.S. NOAA, U.S. Department of Commerce, as well as EA133C-09-CQ-0034 through Eastern Research Group, Inc. (ERG).

NCAR Societal Impacts Program iv ETTC Emergency Manager Survey

EXECUTIVE SUMMARY

Purpose

The National Weather Service (NWS), an agency of the National Oceanic and Atmospheric Administration (NOAA), issues suites of text and graphical products to communicate forecasts of severe storms. During tropical cyclone (TC) conditions, the National Hurricane Center (NHC) issues text and graphical products to communicate the forecasts. During both tropical and extratropical (ET) cyclones and other unique events resulting in storm surge, such as high astronomical tides, local Weather Forecast Offices (WFOs) issue Coastal Flood Advisories, Watches, and Warnings, which include detailed, localized information on expected storm surge and other hazards. This survey of emergency managers (EMs) is part of a larger body of work with the goal of better understanding how certain NOAA weather forecast products might improve public understanding and response to TC and ET events. The major focus is on storm surge products, but EMs from areas subject to TCs were also asked to comment about several forecast track and wind graphics. Major funding for this project came from the NOAA National Ocean Service project, “Assessing Current Storm Surge Information from the Public Perspective.” As an extension of that project, we also leveraged resources from the NOAA- funded “Hurricane Forecast Improvement Project (HFIP) Socio-Economic Impacts Assessment” (through Eastern Research Group, Inc. (ERG). To meet the objectives of this expanded project, we solicited opinions from EMs, broadcast meteorologists, NWS Warning Coordination Meteorologists, and the general public. Here we report the findings from the survey of coastal EMs.

Survey Development, Implementation, and Sample

Questions for this survey were developed in consultation with the NWS staff involved in both the Surge Roadmap Project and the HFIP Socioeconomic Working Group. Early exploration included one-on-one webinars and informal discussions with several EMs. This project built on information gathered in an earlier study of the hurricane forecast communication process funded by the National Science Foundation (Demuth et al. 2012) and earlier surveys on public preferences for hurricane information (Lazo and Waldman 2011; Lazo et al. 2010), and a recent

NCAR Societal Impacts Program ES-1 ETTC Emergency Manager Survey

survey of EMs in TC regions (Morrow and Lazo. Under Review). A crucial part of the survey was assessment of several graphic prototypes developed for communicating storm forecast information. Key NWS personnel reviewed the survey questions, which were approved by the Human Subjects Committee at the University Corporation for Atmospheric Research. The survey was web-based to allow respondents to view the graphics.

Because this survey covered both ET and TC forecast communication, the target population is directors of county/parish/territory emergency management in coastal regions of the Atlantic, Gulf, Pacific, and Alaska. About 90% report their jurisdiction to be vulnerable to tropical storms and 91% to extratropical storms. (Respondents were directed to sections of the survey based on the vulnerability of their jurisdiction to TCs, ETs, or both.) In all, 198 received invitations and 102 completed the survey for a response rate of 52%. This was an experienced sample of EMs, both in emergency management and in responses to severe storms. Thus, this was an appropriate sample for the survey.

Summary of Results

As an introduction to the survey, the EMs are asked how often their jurisdictions are affected by each of 12 different hazards. They are often impacted by heavy rain, coastal storms, and flash floods, and they do not think their public is well prepared for them. Only 10% report they are not vulnerable to TCs, and 9% are not vulnerable to ETs.

A key question is whether these EMs think the NWS should issue storm surge watches and warnings. Their answers reflect strong support for both—83% agree a watch should be issued and 87% support a storm surge warning. Many useful comments are made, most in support of the warnings. They believe these products will result in the public paying more attention, a better- informed response, and greater emphasis on coastal flooding in emergency management decision making. A proposed graphic for showing the area under a surge warning receives high marks on effectiveness, but many call for more localized information.

The next section of the survey deals with local surge inundation. As a prelude they are asked what label should be used to express the expected amount of water. Based on clarity, Height of Water Above Land is their first choice, closely followed by Depth of Water Above Land. Above

NCAR Societal Impacts Program ES-2 ETTC Emergency Manager Survey

Ground receives only moderate support and only 19% of the sample think the meaning of Above Elevation is clear. They are then given an opportunity to assess three prototypes showing local surge inundation forecasts (Figure ES-1) that have been developed through an iterative research process. One is solid blue and shows the entire area at risk for surge, one uses shades of blue to indicate varying depths, and one uses several colors to show Low, Moderate, High, and Extreme levels. They are asked to evaluate each on two criteria: ease of understanding and usefulness in communicating to the public. Assessments of all three are generally positive, but when asked to pick only one, 80% pick the multicolored map.

Figure ES-1. (A) Potential storm surge area in one color; (B) Potential storm surge depths in graduations of one color; (C) Potential storm surge depths in different colors

Moving beyond storm surge, EMs from TC regions are shown three versions of the Tropical Cyclone Forecast Cone (Figure ES-2), typically referred to as the Cone of Uncertainty. This set of questions is asked only of EMs from regions affected by TCs. The first map shows the current cone (white with a solid black line showing the edges), the second one is transparent gray and uses dashed lines for the edges as perhaps a better way to indicate uncertainty, and the third one is a transparent gray cone with no lines at the edges.

Figure ES-2. (A) Current tropical cyclone forecast cone; (B) Transparent forecast error cone with dashed lines; (C) Transparent forecast error cone with no lines

NCAR Societal Impacts Program ES-3 ETTC Emergency Manager Survey

Although only 11% rate the current cone as excellent, a total of 86% give it a positive rating (Excellent, Very Good, or Good) on ease of understanding and 84% on usefulness. In contrast, the transparent gray cone with dashed lines gets a 65% on ease of understanding and 66% on usefulness. The transparent cone with no lines receives scores of 70% and 72%. When asked to choose one, 64% prefer the current cone, 21% the cone with no lines, and 15% the cone with dashed lines. The most common comment is that the current cone has more information.1 A common complaint is the tendency for the public to think the total storm will fall within the confines of the cone. In fact, the public erroneously limiting its attention to the cone is the motivation behind the development of the next forecast products in the survey.

The next map (Figure ES-3) communicates the area with Potential for Damaging Winds in which Low, Moderate, High, and Extreme (each defined in terms of miles per hour winds) are depicted in different colors against a light gray U.S. map. This receives very positive reviews—91% on ease of understanding and 94% on usefulness.

Figure ES-3. Potential for damaging winds

The next two maps add a forecast track cone to this wind map (Figure ES-4). The impetus behind its development is to illustrate that the winds expected in a specific tropical cyclone can occur outside the cone. The first map shows the entire wind field, and the second one masks the portion over water with the intent to better emphasize land areas under threat. On both criteria, 75% give positive ratings to the one with the entire wind field and cone. The one showing the wind forecast over only land is rated positively by 76% on ease of understanding and 64% on

1 In retrospect, the amount of information should have been the same for all three in order to allow a valid comparison.

NCAR Societal Impacts Program ES-4 ETTC Emergency Manager Survey

usefulness in communicating with the public. Some comments reflect a lack of understanding as to why the wind forecast did not cover the entire cone area.

Figure ES-4. (A) Combined wind and cone; (B) Combined wind and cone – land only

The last map (Figure ES-5) was created in response to EMs’ requests for a product showing when they should expect the arrival of tropical storm force winds because this drives their preparation and evacuation timelines. The map uses several distinct gradations of one color (red) to illustrate the approximate time (early Friday morning, Saturday, etc.). There is strong support for this graphic, with 83% giving it a positive rating on ease of understanding and 81% on usefulness.

Figure ES-5. Arrival of tropical storm force winds

When asked about their relationships with local WFOs, all give positive responses. They are then afforded an opportunity to discuss any special issues or challenges they have in receiving and using NWS ET and TC products and what, if anything, they would like improved. Those mentioned most often are summarized first:

NCAR Societal Impacts Program ES-5 ETTC Emergency Manager Survey

• Changes in product formats, including more graphics and simpler templates. • More and better storm surge forecast products. • Dissemination of forecast products earlier in the storm cycle. • Use of well-known landmarks when describing areas under threat. • More information about the degree of certainty in the forecast. • More access to more complicated graphics such as model runs.

Other topics mentioned include more information on tsunami events, tide variations, and expected rain volume; fewer products overall; use of Twitter feeds; and more outreach to the public on where to locate the NHC website and its products.

When asked if their jurisdiction uses a commercial weather vendor, about 30% indicate yes. The more frequent reasons are more model information, timeliness, and data reliability. When asked how often they use public media sources for weather information, local TV is at the top of the list. Most also use Internet sources, with local WFO, NHC, and other NOAA websites getting the most use, followed by state emergency management websites. In closing, many express appreciation for the opportunity to provide input into NWS severe storm products and services and welcome continued collaboration.

NCAR Societal Impacts Program ES-6 ETTC Emergency Manager Survey

1. BACKGROUND

The National Weather Service (NWS), an agency of the National Oceanic and Atmospheric Administration (NOAA), issues suites of forecast text and graphical products to communicate forecasts of tropical cyclones (TCs) and extratropical (ET) storms. During TC conditions, the National Hurricane Center (NHC) uses text and graphical products to communicate the forecasts. These products include information about the position, movement, and characteristics of the storm as well as the threats it poses. The Public Advisory, Forecast Discussion, and Probabilistic Storm Surge products include detailed information about the storm surge threat, but this information is combined with all the other TC forecast information. Several products are used to communicate wind threat, including the categories based on the Saffir-Simpson Wind Scale, and probabilistic wind data and graphics. Projected storm track is typically displayed using the Cone of Uncertainty. The TC Watches and Warnings issued by the NHC currently are based on wind speed criteria only (i.e., for tropical storm and hurricane force wind speeds). During TCs, local NWS Weather Forecast Offices (WFOs) also issue detailed Hurricane Local Statements, which include localized information on the various hazards of a storm. Many of these WFOs are also issuing experimental TC Impacts Graphics, which include a graphic detailing expected storm impacts, including surge. During some TCs, storm surge can occur at coastal locations outside of the areas covered by NHC-issued TC Watches and Warnings. In these situations, local WFOs issue Coastal Flood Advisories, Watches, and Warnings, which include detailed, localized information on expected storm surge. During ETs and other events resulting in storm surge, such as high astronomical tides, WFOs again issue Coastal Flood Advisories, Watches, and Warnings with detailed, localized information on expected storm surge.

The serious threat to life and property posed by storm surge suggests that this hazard be specifically communicated in order to facilitate appropriate proactive and protective decisions. As most recently illustrated by Tropical Storm Debbie, Hurricane Isaac, and Tropical Storm/Hurricane Sandy, dangerous storm surge can occur in extratropical storms, tropical storms and in hurricanes classified at lower categories of the Saffir-Simpson Wind Scale. These latest storms brought renewed calls for clearer NWS forecast messaging and more modern dissemination systems (Norcross 2012). While the forecasts for Sandy, including the storm surge forecasts, were very good, many who should have evacuated did not. For example, eight

NCAR Societal Impacts Program 1 ETTC Emergency Manager Survey

residents of Midland Beach on Staten Island drowned in the surge from Tropical Storm/Hurricane Sandy in spite of a mandatory evacuation order issued at least one day before landfall (Semple and Goldstein 2012). At least 100 people died during Sandy, most from drowning. The fact that no hurricane warning was issued (due to the NWS communication protocols related to types of storms) has been called into question (Margusity 2012). Whatever the reasons, there is strong evidence that people do not pay sufficient attention to the dangers of storm surge.

Are new storm surge forecast communication approaches needed to improve decision making to protect life and property? The development of a storm surge index has been suggested, most recently after Hurricane Isaac (Schleifstein 2012), but NWS experts argue that providing the expected surge depth in feet is, in fact, a scale (NHC Public Affairs 2012). A separate Storm Surge Warning, accompanied by local inundation maps, is under consideration. Before making final decisions, however, the involved NOAA entities decided to support social science research aimed at soliciting opinions from key stakeholders, including the general public, emergency managers, broadcast media, and NWS Warning Coordination Meteorologists (WCMs). Several projects are gathering this information from across coastal areas of the United States and its territories, including the Atlantic, Gulf, and West coasts; Alaska; the Hawaiian Islands; and U.S. territories in the Pacific Ocean. The work reported here focuses on emergency managers (EMs) in both ET and TC locales. In addition to the surge-related questions, those from regions subject to TCs were also asked to assess several forecast track and wind graphics. Associated work under way with the public (conducted by Eastern Research Group Associates), broadcast meteorologists, and WCMs will be reported subsequently. Prior related work includes the Coastal Emergency Managers Internet Survey (Morrow and Lazo 2012), which examined the perceived value of various tropical cyclone forecast products, and the Extratropical –Tropical Cyclone Storm Surge Public Survey.

The NOAA National Ocean Service (NOS)-funded project “Assessing Current Storm Surge Information from the Public Perspective” defines two objectives:

NCAR Societal Impacts Program 2 ETTC Emergency Manager Survey

1. To explore and assess the public’s awareness and understanding, or lack thereof, concerning storm surge and currently available storm surge information, regardless of the meteorological cause; i.e., “Do they know what storm surge is?” 2. To assess whether the NWS should develop new storm surge informational approaches to improve the communication and decision-making with respect to extratropical and tropical cyclone storm surge risk.

As part of that project, we leveraged resources from the NOAA-funded “Hurricane Forecast Improvement Project (HFIP) Socio-Economic Impacts Assessment” to gather opinions of emergency managers related to the presentation of storm surge information, as well as certain track and wind forecast communication products. Resources from the two projects were pooled to support a larger body of work; the work reported here is only one part.

NCAR Societal Impacts Program 3 ETTC Emergency Manager Survey

2. METHODS

2.1 Survey Development Questions for this survey were developed in consultation with the NWS staff involved in both the Surge Roadmap Project and HFIP. This work built on information gathered in an earlier National Science Foundation-funded study of the hurricane forecast communication process (Demuth et al. 2012), earlier surveys on public preferences for hurricane information (Lazo and Waldman 2011; Lazo et al. 2010), qualitative exploratory interviews with stakeholders in hurricane vulnerable areas (Lazrus et al. 2012), exploratory one-on-one webinar and in-person interviews with several EMs, and a review of the literature on storm surge communication.2 Some questions from a prior survey of TC coastal EMs (Morrow and Lazo 2012) were expanded to include ET questions. The graphics tested in this survey were developed through an iterative process involving input from stakeholders and NWS personnel. Survey questions were reviewed by key NOAA personnel and approved by the Human Subjects Committee at the University Corporation for Atmospheric Research. Input was also gathered at several emergency management workshops held at the NHC. The survey instrument was programmed for online implementation, pilot tested, and administered by ResearchExec (http://www.researchexec.com/). The final survey consisted of 84 questions, including demographics. Respondents were directed to various sections of the survey based on the vulnerability of their jurisdictions to TCs, ETs, or both. The codebook in Appendix B includes the complete questionnaire and summary response data.3

2.2 Sample Because this survey covered both ET and TC forecast communication, the target population was directors of county/parish/territory emergency management in coastal areas of the Atlantic, Gulf, Pacific and Alaska.4 Email addresses were obtained and verified through directories, websites, phone calls, and a trial message. An introductory letter was emailed to each address (see Appendix A). ResearchExec then sent an email with details for participation. Twice during the

2 A separate report is forthcoming on the literature review. 3 We do not report responses to open-ended questions to maintain respondent confidentiality. Quotes used throughout the report have been chosen to be representative of open-ended responses while maintaining confidentiality. 4 In a few cases, city or state regional EMs fulfill this function.

NCAR Societal Impacts Program 4 ETTC Emergency Manager Survey

process reminders were sent to those who had not completed the survey. A total of 204 EMs were sent invitations and 113 EMs completed the survey, or at least most of it. If we assume all 204 received their invitations, the response rate for the survey is 55%. Of the total, 90% represented counties, parishes, or boroughs, 8% cities or towns, and 2% state agencies. About 57% were directors of their agencies, 28% were emergency managers, and 14% were other (usually employed by public safety agencies). They had been in emergency management for an average of 16 years and in their current position for an average of 7 years. The approximate populations of their jurisdictions vary from less than 50,000 to more than 1 million, with most having between 50,000 and 500,000 residents. Table 1 reports the number of respondents from each coastal state or territory. During sample development we were unable to identify appropriate contact information for EMs in the District of Columbia and New Hampshire, and these areas are not represented in the sample.

Table 3-1-1. Location of Respondents

Coastal State or Territory Number Contacted Number of Respondents Alaska 5 4 Alabama 2 0 California 15 11 Connecticut 3 1 Delaware 3 1 Florida 36 23 Georgia 6 4 Hawaii 4 4 Louisiana 19 9 Maine 8 3 Maryland 5 3 Massachusetts 5 2 Mississippi 3 1 North Carolina 16 9 New Jersey 8 3 New York 6 4 Oregon 7 6 Rhode Island 5 3 South Carolina 6 3 Texas 28 11 US Virgin Islands 1 0 Virginia 6 4 Washington 7 4 Totals 204 113

NCAR Societal Impacts Program 5 ETTC Emergency Manager Survey

Figure 2-1 shows where respondents and non-respondents are located. The counties corresponding to the jurisdictions of respondents are shaded blue and those of non- respondents are shaded red.

Figure 2-1. Locations of sample

2.3 Data Collection and Analysis The survey was designed to gain a better understanding of cyclone-related concerns of EMs and how they assess and use selected NWS products in their work. They were also asked to review prototypes of several graphics being considered for communicating track, wind, and surge information. Because the respondents needed to view graphics as they answered the questions, the survey was web-based. ResearchExec formatted and administered it in June and July 2012. Data were then analyzed and a codebook with questions and a summary of responses prepared at NCAR (see Appendix B).

NCAR Societal Impacts Program 6 ETTC Emergency Manager Survey

3. RESULTS 3.1 Regional Hazards Risk and Public Preparation As an introduction to the survey, the EMs are asked how often their jurisdiction area is affected by each of 12 different hazards. The choices are presented in a five-point scale: 1 Never, 2 Rarely, 3 Occasionally, 4 Often, and 5 Frequently. Then they are asked how prepared they think people are for each hazard. Again the choices are scaled: 1 Not at All Prepared, 2 Somewhat Prepared, 3 Prepared, 4 Very Prepared, and 5 Extremely Prepared. The median scores for each hazard are then compared. Figure 3-1 shows the relationship between hazard frequency and public preparation for each of the hazards.

Figure 3-1. Comparison of hazard risk and public vulnerability

Although the response scales are not directly comparable, looking at difference in the gaps across hazards, in all except tsunamis and mud slides the frequency is rated higher than public preparation, but this is especially true for heavy rain, coastal storms, and high waves. When

NCAR Societal Impacts Program 7 ETTC Emergency Manager Survey

asked about any other weather-related hazards, lightning, high wind events, fog, and cold weather events are most often mentioned.

After a short description of tropical and extratropical storms, they are asked to rate the vulnerability of their jurisdiction to TCs and ETs that produce at least tropical storm force winds (39–54 mph). The scale is from Extremely Vulnerable to Not At All Vulnerable, with Vulnerable as the middle choice. Figures 3a and 3b show the responses to these two questions.

Figure 3-2a. Vulnerability to tropical storms Figure 3-2b. Vulnerability to extratropical storms

The results indicate a great deal of vulnerability to both types of storms. Only 10% reported they are not at all vulnerable to tropical storms and 9% are not vulnerable to extratropical storms.

As shown in Table 3-1, these emergency managers have considerable experience dealing with tropical and extratropical cyclones.

Table 3-1. Incidence of Activations for Storms in Last 10 Years (%) 0 1–4 5–8 9 or more Tropical Storms 15 25 37 23 Hurricanes 20 37 26 16 Extratropical Storms 26 44 19 12 N = 110 (Rows may not add to 100% because of rounding.)

NCAR Societal Impacts Program 8 ETTC Emergency Manager Survey

The median number of activations for storms for the entire sample is five to six for tropical storms, three to four for hurricanes, and one to two for extratropical storms. Clearly, this is an experienced sample whose opinions regarding NWS severe storm forecast products are valuable.

The next question asks, “..if a severe storm were predicted for your area, to what extent would you be concerned about each of the following hazards?” As shown in Table 3-2, the listed hazards are storm surge, wind, tornadoes, inland flooding, and heavy snow.

Table 3-2. Level of Concern About Hazards (%) Does Extremely Very A little Not at all not Hazard concerned concerned Concerned concerned concerned apply Storm Surge 43 37 14 5 0 0 Wind 35 48 16 2 0 0 Inland 33 35 22 10 0 1 Flooding Tornadoes 32 30 17 11 4 6 Heavy Snow 9 14 19 14 9 35 N = 113

All express concern, to one degree or another, about storm surge, wind, and inland flooding. Concern for tornadoes and heavy snow (where applicable) is also high. One comment from an emergency manager concerning extratropical forecasts, “We’re very concerned with storm surge forecasts due to local coastal erosion and the presence of a nuclear power plant within our jurisdiction.”

When asked what portion of their population lives in areas vulnerable to storm surge or coastal flooding, the answers vary from zero to 100%, with 40% being the median. When asked “of those that are vulnerable, what portion (from 0% to 100%) do they think adequately understand their vulnerability,” the median answer is 40%. This suggests that these EMs perceive that most of those vulnerable to storm surge (e.g., 60%) do not adequately understand their vulnerability.

3.2 OPINIONS REGARDING STORM FORECAST PRODUCTS 3.2.1 Storm Surge Warning An introduction was provided at the beginning of this section of the 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

NCAR Societal Impacts Program 9 ETTC Emergency Manager Survey

scale. This has led to an investigation of how surge information should be communicated for BOTH tropical and extratropical cyclones.

Respondents are then asked to what extent they agree or disagree with the two statements regarding whether the NWS should issue storm surge watches and warnings (see Table 3-3).

Table 3-3. Opinion about Storm Surge Watches and Warnings (%) Strongly Strongly agree Agree Undecided Disagree disagree The NWS should issue Storm Surge 45 38 6 3 8 Watches The NWS should issue Storm Surge 50 37 3 2 8 Warnings N=110

As shown, 83% either Strongly Agree or Agree that storm surge watches should be issued and 87% feel the same way about warnings. They are asked to provide reasons for their answers. Most of the comments are supportive and provide useful insights. A few follow.

“We need to start letting the public know that the Weather Service is decoupling surge from wind speed. The surge threat needs to stand alone and be recognized as the killer it could be.” “Why not? These hazards kill the most people along coastal areas. Earlier warnings save lives.” “If the NWS issued storm surge watches and warnings perhaps the public would pay more attention to the evacuation orders issued.” “Because they have the tools available to appropriately assess the situation. In addition they have the tools available to provide immediate notification to the media and the public. As the subject matter experts, the public would consider it a reliable source. Most importantly, it would not require a novice to interpret what it means.” “I believe these can be used as a tool to help educate our population about flooding issues.” “My barrier island residents are educated about surge but citizens living along our rivers don’t understand surge will affect them too. Sooner watches and warnings issued may help to get citizens attention sooner and give them more time to make their evacuation plans.”

A number of concerns or reservations were expressed.

“This would be too confusing to have wind and surge watches and warnings. I think the surge threat can be conveyed in plain English. . . .’ Storm surge of up to X above ground level could be expected from ____ to _____ in the next X hours.” “We already have hurricane watches and warnings with established break points. Surge warnings would further confuse the public relative to evacuations. It would be possible for part of a county

NCAR Societal Impacts Program 10 ETTC Emergency Manager Survey

to be under each warning and which would be the most important? I see a minefield here unless done right. People might wait for a surge warning and then it might be too late to evacuate.” “Information regarding storm surge potential should be included in any watch or warning that is issued for a Tropical Storm or Hurricane . . . plus, there is already the Coastal Flood product that can be used for nor’easters.” “Our job is to notify the public of hazards. We only rely on NWS to give us the science in a timely manner.”

Some provide caveats and suggestions.

“Not sure if a watch would cause people to react any differently but a warning should tell people that they need to react to the forecast.” “The landfall location of tropical storms is much different than storm surge inundation areas…these watches/warnings need to include inland penetration depth.” “We have a coastal area with many fishing boats and intertidal harvesters. They should be recognized as a group affected by storm surge condition. The fishing boats work mainly inshore and could easily be pushed aground if they are unaware of the weather conditions.” “. . . we have frequent storms so it must be used judiciously to avoid over-saturating the public.”

Some comments, such as the one below may reflect confusion about the product.

“…however, they could continue to be incorporated into ongoing tropical watches and warnings. Storm surge watch/warnings by themselves would be extremely rare, but if condition warranted, I would support specific watches/warnings for them.”

The next three questions ask about the possible effects of a Storm Surge Warning, specifically whether it would result in the public paying more attention to surge, a better informed response, and greater emphasis on coastal flooding or storm threat in emergency management decision making. As shown in Table 3-4, the majority believes all three of these presumably positive outcomes will occur if a storm surge warning is issued.

Table 3-4. Opinions about Effects of Storm Surge Warning (%) Definitely not/ Definitely Probably Possibly probably not Public paying more attention 25 36 30 9 A better informed response 31 40 22 8 Greater emphasis on coastal flooding or storm threat in emergency management decision 38 32 21 9 making N = 107

NCAR Societal Impacts Program 11 ETTC Emergency Manager Survey

These EMs are then asked to evaluate several possible names for a surge warning product. These are not mutually exclusive. To some degree:

• Nearly 90% like Storm Surge Warning. • More than 70% like Extreme Coastal Flood Warning. • Nearly 60% like Storm Surge Risk. • 56% like Storm Surge Danger.

Some comments are:

“The terms warning and watches are already ingrained in the populations’ mindset. Most people understand that a watch means the event is likely and a warning means it is imminent.” “I believe you should use what you have done in the past and use the watch and warning. People are more attuned to using the same notices as with tornadoes and hurricanes. Why change?” “People in my area have become somewhat desensitized to CFWs.”

Some other points of view:

“Surge is too confusing for our public. If it is flooding, call it that.” “I would call it just what it is – warning, danger, risk, etc. Public doesn’t respond well to ‘soft’ terminology. If it poses a danger, let’s call it that.”

Other name suggestions include Storm Surge Threat, Extreme Coastal Storm Surge Warning, and Extreme Storm Surge Danger.

3.2.2 Storm Surge Warning Area Map Assessment of proposed graphics for illustrating surge-threatened areas is an important part of this survey. These graphics underwent considerable exploratory research with various stakeholders and were revised in an iterative process before arriving at those tested here.

If a Storm Surge Warning is issued, one question is how it will be displayed on the NWS public websites. Respondents were asked to assess the effectiveness of several graphics starting with the map in Figure 3-3.

NCAR Societal Impacts Program 12 ETTC Emergency Manager Survey

Figure 3-3. Experimental graphic for showing area under storm surge warning

As illustrated in Figure 3-4, the results are positive, with more than 65% rating this map in Figure 3-3 as Extremely or Very Effective and another 27% as Effective.

Figure 3-4. Perceived effectiveness of storm surge warning map

NCAR Societal Impacts Program 13 ETTC Emergency Manager Survey

The map elicited many comments, including numerous calls for more localized information. Note that at this point in the survey, respondents did not know about the planned surge inundation maps shown subsequently.

“As a visual it can be easily updated and copied over to my county emergency management website for dissemination to the public and for media to easily visit as they often help me get the message out during tropical weather events.” “The map would be most useful to our core group of emergency planners. We would then translate this map info into area emergency notifications.” “Because this area is so large, the entire west coast, less people may see this as a theat. Perhaps a smaller warning area with a minimum threshold, such as more than 3 feet.” “Must be done in conjunction with other warnings and same areas to avoid confusion.” “As long as its clear to the public which depicts areas under wind watch or warning and areas under storm surge watch or warning.” “Our coastline is rugged and rocky. Surge will be most felt in low lying areas but would not be a general area wide problem due to the high rocky bluffs and hilly terrain along the coast.” “A storm surge warning could cover the entire state but graphics should be local to at least the county level to provide more fidelity and a more meaningful visual to the public.” “It would only be effective to tell them they might be affected, which will elicit little to no response from the public because there is nothing to get them to personalize the threat – it is not specific enough.”

These comments were generally consistent with expectations and led into the next section of the survey and a range of additional potential graphics.

3.2.3 Storm Surge Inundation Maps A preamble to this section of the survey explains that the NHC is developing tropical cyclone storm-specific inundation maps for their website and that similar maps could be developed by NWS in the future for extratropical storms. It is explained that the maps show estimates of the amount of ocean water (including tides) expected on top of the land. The elevation above sea level of that general area will already have been subtracted. Before assessing the maps, respondents were asked about the best way to label this hazard for the public to understand. They are asked to assess five possible names as shown in Table 3-5.

NCAR Societal Impacts Program 14 ETTC Emergency Manager Survey

Table 3-5. Preference for Storm Surge Labels (%)

Extremely Not very Not at all Label clear Very clear Clear clear clear Height of Water Above Land 19 29 35 15 2 Depth of Water Above Land 11 28 39 17 5 Above Ground Level 10 21 36 24 10 Above Ground 4 12 28 46 11 Above Elevation 3 3 13 50 31 N= 105

The results are somewhat unexpected. If the three positive choices (Extremely Clear, Very Clear, and Clear) are added together, Height of Water Above Land received a positive assessment of 83%, Depth of Water Above Land of 78%, Above Ground Level of 67%, and Above Ground is considered clear by 44%. Most think Above Elevation is not clear. Perhaps the idea that it will be measured above ground, i.e., above where people will be standing, explains the preference for height over depth (which will be measured from the top down). Other suggested labels include Possible Surge Risk, Above Base Flood Elevation, Height or Depth of Water Over Land, and Overland Surge. Several suggestions indicate that some respondents did not realize it will include tide levels. Some concerns were voiced.

“I disagree with the idea of giving depth of water over ground. It implies that there may be safe places to be when in fact that may not be the case.” “These do not articulate wind driven wave threats versus inundation.”

A crucial part of this survey is an assessment of three different prototypal maps developed for illustrating storm surge inundation. The EMs are asked to assess each map on two criteria: ease of understanding, and usefulness for communicating to the public what they need to know about storm surge. After assessing each graphic individually, the EMs were then asked to pick the one they prefer.

The first map (Figure 3-5) depicts as one color (blue) the entire area forecast to have surge.5

5 The maps were presented in the same order to all respondents.

NCAR Societal Impacts Program 15 ETTC Emergency Manager Survey

Figure 3-5. Potential storm surge area in one color

About 82% rate this map as Excellent, Very Good, or Good in ease of understanding and 75% in usefulness. Comments include numerous calls for depth information such as, “What does the map represent? If we do not show the various levels, the perception will be that the map is worst- case and an exaggeration that will never happen.” Another comment, “It depends on the confidence of the information. What if people not indicated to be in the surge area stay around and you have underestimated the surge?”

The second surge inundation map (Figure 3-6) uses color graduation to illustrate expected levels of storm surge.

NCAR Societal Impacts Program 16 ETTC Emergency Manager Survey

Figure 3-6. Potential storm surge depths in graduations of one color

About 75% think this map is Excellent, Very Good, or Good in ease of understanding and 73% make the same choices for usefulness in communicating to the public. There are a few positive comments such as, “The more information the better. I believe that the general public has a greater ability to understand these products than maybe we give credit for.” However, considerable concerns are raised.

“I think height is a better descriptor than depth. Depth infers below.” “Too much detail. People may be tempted to ride it out if they see a map that shows water of 1 or 2 feet as opposed to a general single color map that shows their property being inundated.” “People will be confused as to where they live and what it would mean at their location.” “It is difficult to determine the different surge levels.”

The third map (Figure 3-7) uses colors to depict different expected depths of storm surge (Low, Moderate, High, Extreme—based on the feet above ground level as indicated in parentheses in the legend).

NCAR Societal Impacts Program 17 ETTC Emergency Manager Survey

Figure 3-7. Potential storm surge depths in different colors

This map receives the most positive assessments. In terms of ease of understanding, 86% rate it Excellent, Very Good, or Good and for usefulness in communication to the public, the total for these choices is 84%. Numerous positive comments are made, some quite enthusiastic.

“This map better explains graphically the depth of water above ground level so most people can understand it.” “Best visual. The media would be more willing to use this graphic to help emergency management officials warn the public.” “This is by far the best graphic for potential storm surge for dissemination to the public.”

However, there are some concerns.

“Too many choices. You are either in a coastal flood zone or not.” “This map now encourages people not to evacuate a barrier island… because parts of it will only have 3’ or less of surge. This is suicide for many who will stay when no help can reach them— and you’re shooting the EM agencies and first responders in the foot…I cannot stress enough, YOU CANNOT ISSUE MAPS THAT WILL NOT AGREE WITH THE EVCUATION ZONES DEVELOPED BASED ON SLOSH MOM’s and local policy.”

The last comment here is potentially important because some risk communication literature emphasizes consistency in risk messaging.

NCAR Societal Impacts Program 18 ETTC Emergency Manager Survey

Several suggestions for improvement are made, including darkening the land portion outside the surge area and making sure any graphics show up well on smart phones and iPads.

Assessments of all three maps are somewhat positive. Respondents were asked to indicate a preference: “If the NHC uses only one of these inundation maps on its website, which is best?” The results as shown in Figure 3-8 are unequivocally in favor of the multicolored map.

Figure 3-8. Map preference

When asked why they think their choice is best, there are numerous statements related to providing more information, more detail, more specifics, being more eye-catching. It is also suggested that the blue be changed to green to match the colors used in SLOSH. Some other reasons for picking the multicolored graphic include:

“Basic understanding that red = worst, yellow = moderate…basic education training. Use it!” “This multi-colored map better explains to people who do not know their elevation.” “I think it better defines the ‘kill zone.’’’

3.2.4 Forecast Track Cone Maps Emergency managers who indicated their jurisdiction is not impacted by tropical cyclones skipped this section, reducing the sample size to 81. The section begins by asking for an assessment of the current Tropical Cyclone Forecast Cone as presented in Figure 3-9.

NCAR Societal Impacts Program 19 ETTC Emergency Manager Survey

Figure 3-9. Current tropical cyclone forecast cone

When asked to take the public perspective, these EMs are generally positive about the current cone. Although only 11% rate it as excellent, combining those who thought it was Excellent, Very Good, and Good results in a 86% positive rating on ease of understanding and 84% on usefulness.

Positive and negative comments are about equal in number.

“This is a very useful map…majority of public is getting familiar with this particular visual information.” “Gives the public a sense of danger without a lot of extra information.” “Does not represent threat potential very well.” “Most laypeople misinterpret it.” “Public will not see themselves in the widest part of the cone.” “The program being used to demonstrate your example should be improved…better utilization of graphics is available and we are shooting ourselves and those that make decisions by not using better tools.”

In the second map ((Figure 3-10) the track forecast cone is presented as transparent with dashed border lines in an effort to signify uncertainty.

NCAR Societal Impacts Program 20 ETTC Emergency Manager Survey

Figure 3-10. Transparent forecast error cone with dashed lines

Most like this one less than the current cone. Only 5% think it is Excellent, but when Excellent, Very Good, and Good are combined, the score for ease of understanding is 65% and 66% for usefulness to the public. Some think the dashed lines do not help get the uncertainty message across. Other comments deal with the fact that the same amount of information is not included on this one nor is it in color. So there are important differences between the current cone and this one beyond the dashed lines. In retrospect, in order to get a fair assessment, the only thing different on this map should have been the way the cone was displayed.6

On the third map (Figure 3-11) in this series, the cone is transparent gray with no lines.

Figure 3-11. Transparent forecast error cone with no lines

6 These responses also reveal the value of eliciting input on potential changes in products, because well-intentioned changes can actually cause unexpected or unanticipated interpretations.

NCAR Societal Impacts Program 21 ETTC Emergency Manager Survey

Again, although only 3% rate it as Excellent, the positive choices together add to 70% on ease of understanding and 72% on usefulness. Comments are fewer and are concerned mostly with lack of color and the fact that this map does not provide as much information as the current cone. An interesting comment was, “. . . laying the image of an actual storm onto this graphic will help since the size of each storm differs.”

Next, respondents are asked to make a choice between the three cone options.

Figure 3-12. Preference for tropical cyclone forecast cone

The results in Figure 3-12 indicate that 64% prefer the current cone, 21% the cone with no lines, and 15% the one with dashed lines. When asked to explain their choices, many commented that the first cone provides more information (watches, warnings, etc.) and is in color. Even some who like the dashed lines or no lines cones mention they need color and more information. As previously noted, this would have been a better assessment if all three maps had shared the same features except for the line differences.

3.2.5 Wind Map The next map (Figure 3-13) shows the varying potential for damaging winds for the same storm illustrated in the previous cone examples.

NCAR Societal Impacts Program 22 ETTC Emergency Manager Survey

Figure 3-13. Potential for damaging winds

Respondents are very positive about this graphic, with 91% giving it a Excellent, Very Good, or Good rating on ease of understanding and 94% on usefulness. In general they like the presentation, i.e., strong colors against a plain background. Some positive comments:

“Tremendous amount of information. Easy to understand.” “Great map.” “The contrast makes a big difference in providing impact information concerning the storm.”

One concern:

“I think the 10 to 20 percent chance is not real helpful. The low, moderate, high, extreme is the increase in wind, not potential for wind. That is what could be misunderstood by the public.”

3.2.6 Wind and Cone Maps The next map adds the dashed cone to the Potential Damaging Winds map (Figure 3-14) in an effort in part to show that winds can occur outside of the cone track area. Two versions are presented, with the second one (Figure 3-15) showing the wind graphic over land only, masking the water portion. We discuss them together with results summarized in Table 3-6.

NCAR Societal Impacts Program 23 ETTC Emergency Manager Survey

Figure 3-14. Combined wind and cone

Figure 3-15. Combined wind and cone – land only

Table 3-6. Assessment of Combined Wind and Cone Map (%) Excellent Very good Good Fair Poor Combined Cone and Wind – Total (Figure 3-14) Ease of Understanding 11 40 24 16 9 Usefulness 10 43 23 16 9 Combined Cone and Wind – Land Only (Figure 3-15) Ease of Understanding 10 35 21 24 10 Usefulness 10 36 18 26 10 N = 80

Three-fourths give positive (Excellent, Very Good, Good) ratings to the first map on both criteria. The second map rates 76% on ease of understanding, but only 64% on usefulness in communicating with the public.

NCAR Societal Impacts Program 24 ETTC Emergency Manager Survey

Comments on the first one mostly have to do with the lack of congruence between the wind and cone. Some of the EM respondents did not understand how that could occur.

“Will confuse public because some tracking areas have no wind issues.” “It gets confusing when the high wind probabilities are outside the cone.”

Opinions about showing the wind information only over land are mixed—some are enthusiastic.

“Helps the public focus on their area, not the whole ocean.” “Good because it shows only relevant information.” “This looks cleaner.”

Several express concern about marine interests.

Focusing respondents on the potential impacts in North Carolina and Virginia, respondents are asked which of these same two maps best communicates the urgency at each location. Two- thirds prefer the first one—over both land and water. Comments are about evenly divided. Those in support of the first one mention marine interests and that it shows the size of the storm. Positive comments about the second one are mostly that it better highlights the danger.

3.2.7 Arrival of Tropical Storm Force Winds Emergency managers have been asking NWS to provide information about when to expect the arrival of tropical storm force winds, as they indicate that all preparations should be completed by then. The graphic shown in Figure 3-16 was developed to address that need and Table 3-7 presents responses.

NCAR Societal Impacts Program 25 ETTC Emergency Manager Survey

Figure 3-16. Arrival of tropical storm force winds

Table 3-7. Assessment of Arrival to TS Force Winds Map (%) Excellent Very good Good Fair Poor Ease of Understanding 11 36 35 13 5 Usefulness 11 39 30 14 6 N = 80

This graphic is well supported; about 83% give it a positive rating on ease of understanding and 81% on usefulness in communicating to the public when they should be ready.

Some positive comments:

“Easy to understand.” “Easy to follow.” “In our area we calculate that evacuations to be effective must be 36 hours before the onset of tropical winds. Hurricane evacuation routes are vulnerable to gridlock. This map helps communicate that info.”

Given that this product is being developed at the request of EMs, some of their concerns are presented in detail below.

“You better be confident on arrival times, or give a message that it will be updated every so many hours.” “Must extrapolate the time and location of Savannah and many will be unable to do that unless hit in the head with a stick and individually explained to them. May be misleading without that explanation. May also cost me my ‘hidden’ six hours before the arrival of TS force winds that I plan for in applying my clearance time.”

NCAR Societal Impacts Program 26 ETTC Emergency Manager Survey

“As an emergency manager I’d like to know what you are using to make these determinations because 25+ years of experience have taught us some rules of thumb about when to plan for the arrival of TS force winds. . . . If you issue something that the locals don’t understand the methodology, then we can’t adjust and the public again things the locals don’t know what they’re doing because it doesn’t match what the NHC is putting out.” “I think some folks would think this is when they need to finish boarding up their houses and get on the road. This also confuses evacuation zones. I look at this and think everyone in the shaded area needs to run! From an emergency management perspective it would be beneficial but not something I would want disseminated to the general public.” “Additional text, ‘All storm preparations should be completed prior to the arrival of tropical storm force winds’ would be helpful. More direct and specific instructions (not just inferred from the graphic title).”

Even with these comments and caveats, there appears to be substantial support for this graphic. Given concerns with respect to how the public would respond to this information and that this was designed in response to EMs’ decision-making, it seems reasonable that the NHC consider providing this to EMs only.

3.3 Use of Forecast Services When asked if they use a paid commercial vendor for weather forecast information, nearly 30% answer yes. The reasons are summarized in Table 3-8 in order of average importance.7

Table 3-8. Reasons for Using Commercial Vendors (%) Extremely Very Not very Not at all Reason Important important important important important Timeliness 21 24 38 7 13 Tailored to Market 10 28 38 17 7 More Model Information 7 28 41 17 7 Data Reliability 10 7 66 10 7 Interpretation of NWS Data 14 17 38 14 17 Better Graphics 10 14 38 21 17 N = 29

Most consider all as important reasons for using commercial vendors. The most-cited reason, however, is timeliness, closely followed by tailored to local market or area, getting more model information, and data reliability.

7 When calculating “mean” responses, we numbered response options 1 through 5 and calculated simply averages (as in the summary statistics reported in the Appendix B). We recognize that the scales are ordinal and not truly interval, but treating the response options as interval provides for simple rankings such as presented here.

NCAR Societal Impacts Program 27 ETTC Emergency Manager Survey

The EMs were asked to what extent they go to each of the following public media sources for forecast information when a hurricane or severe coastal storm threatens their areas. The sources are listed in order of average preference in Table 3-9.

Table 3-9. How Often Use Public Weather Sources (%) Source Frequently Often Occasionally Rarely Never Local TV News 23 37 23 14 4 NOAA Weather Radio 30 21 22 20 8 The Weather Channel on TV 23 22 34 16 6 National TV (e.g., ABC, CBS, NBC, CNN, or 16 23 33 21 8 Fox) Weather Underground 14 19 33 18 17 Local Radio 10 17 40 26 8 N = 102

Local TV is used Frequently or Often by 60% of the sample. The most frequent comment is that they rely on the NWS information.

This question also asked their frequency of using the Internet, and this was by far the most frequent source of forecast information for these EMs—more than any other public media source, reasonably supporting the idea that EMs need more detailed information than the general public for decision making given their responsibilities. Asked about their use of the Internet, 89% say they use it Frequently or Often. The following question then asks about their use of specific Internet websites, and Table 3-10 lists these in order of preference based on Frequently and Often response rates.

Table 3-10. How Often Use Internet Websites (%) Source Frequently Often Occasionally Rarely Never Local NWS Office 68 23 8 0 1 National Hurricane Center 61 18 5 2 14 Other NOAA Website 38 33 19 8 2 State Emergency Management 27 28 29 13 3 Weather Underground 16 17 25 23 19 Other Weather Website 14 18 48 10 10 Local TV 15 16 37 24 8 Weather Channel 12 15 40 22 11 National TV 9 12 35 29 15 N = 101

The most often used website is that of their local NWS office, followed by the NHC and other NOAA websites. It should be remembered that this question was asked of those in both tropical

NCAR Societal Impacts Program 28 ETTC Emergency Manager Survey and extratropical cyclone regions. In comments, they mention other sites such as Crown Weather, Intellicast, Accuweather, Vuetoo, and University of Hawaii Manoa.

When asked to rate their relationship with the NWS office(s) serving their area, the results are extremely positive. All rate their relationship as Excellent, Very Good, or Good. In fact, 76% of the total say they have an excellent relationship with their local WFO(s). The numerous compliments made about local offices can be summed up by “NWS does an excellent job with a personal touch here in _____. We are on a first name basis.”

The survey explores how this sample of EMs use the products and services of the NWS. Two open-ended questions are asked:

• What, if any, special issues or challenges do you have in receiving and using NWS tropical and extratropical storm forecast and warning products in your work? • What, if anything is the single most important change the NWS could make to improve its severe storm forecast and warning products and services?

About 50 emergency managers wrote answers to each open-ended question. Considerable overlap exists between the two questions, so the comments are merged by topic and presented next, with those mentioned most often summarized first. Examples of comments are provided under each.

The most common remarks are about product formats and include calls for more graphics and simpler templates.

“Reformatting the information for Twitter (and Facebook). It would be so much easier if the information came out from the local NWS office pre-formatted so that all emergency management offices (and the media) had to do was Retweet or Share it. When we have to translate and edit the information ourselves, we may not be conveying exactly what the NWS wants, and it takes valuable time.” “All capitals in the text documents makes it difficult to read.” “In the past messages were almost code-like.” “Highlight critical, or changed, information in larger font; make font choices more readable…” “Providing better maps and visual products to be used by emergency managers in making decisions.” “Maps with Google Earth overlay.”

NCAR Societal Impacts Program 29 ETTC Emergency Manager Survey

“I think more detailed, zoomed-in maps are easier to interpret. Maps that overlay orthophotography are especially meaningful to residents whose homes are in potential impact area.” “By implementing some of the graphics that you have shown in this survey.”

Many mention the need for more and better surge forecast products.

“Understanding how storm surge will effect separate areas of our County.” “It would be more productive if we could see the overall picture at once (e.g., both wind and surge forecasts).” “Enhanced storm surge height information.” “The water above land maps would be most useful to me.”

Several mention needing the products earlier in the storm cycle.

“Just need a forecast of the major strike area available earlier in the storm cycle.” “Give as much advance notice of the probability as possible.” “Longer lead times. Our major protective actions decision time is 72 hours prior to landfall. The expression ‘in the next 48 hours’ doesn’t face the reality that 48 hours is not time to plan anything.”

Well-known landmarks should be mentioned when describing areas under threat.

“The use of ‘little known/arbitrary’ locations to set the watches and warnings. I don’t know the locations used and have to regularly go look up where those points are—adding a county name to those points would be very helpful.”

More information is needed related to uncertainty.

“Add the additional 30% margin of error cone to indicate that they should be prepared for the possibility of a change in the storm track.”

The common call for special products for EMs is proposed a little differently here.

“I would like to see the NHC place the more complicated graphics intended for EM decision making somewhere not as easily accessible to the general public. Not saying a non-public portal, just something off the beaten path so that they do not confuse the general public. Seeing the model runs are also helpful. I think it’s silly that I have to go to other sources to get this information.”

NCAR Societal Impacts Program 30 ETTC Emergency Manager Survey

Other mentioned topics include more information on tsunami events, tide variations, and expected rain volume; fewer products overall; use of Twitter feeds; and more outreach to the public on where to locate the NHC website and its products.

NCAR Societal Impacts Program 31 ETTC Emergency Manager Survey

4. Summary

A key question was whether these EMs think the NWS should issue storm surge watches and warnings. This survey reflects strong support for both (83% and 87%, respectively). A proposed graphic for showing the area under a surge warning receives high marks on effectiveness, but with calls for higher resolution and more localized information. When shown three possible maps for illustrating local inundation, EMs prefer one that uses different colors to indicate levels of potential inundation. Most prefer that this be labeled Height of Water Above Land.

Moving beyond storm surge, those from TC regions were shown three versions of the Tropical Cyclone Forecast Cone, typically referred to as the Cone of Uncertainty. Although only 11% rate the current cone as excellent, 86% give it a positive rating (Excellent, Very Good, or Good) on ease of understanding and 84% on usefulness. It is preferred over a map that has a transparent gray cone with dashed lines and one showing a transparent cone with no lines. A problem with this exercise, however, was that as presented the current cone map did, in fact, provide more information (such as being in color and having more labeling and text). Therefore, this was not a valid test of cone preference. The cone issue requires further study.

The map communicating Potential for Damaging Winds using Low, Moderate, High and Extreme labels (each defined in terms of wind speeds) was well received. The different levels are depicted in different colors against a light gray U.S. map. This map receives very positive reviews. A forecast track cone was added to the wind map in the next two graphics. The first one shows the entire wind field and the second one masks the portion over water with the intent of better emphasizing land areas under threat. About two-thirds prefer the one showing the entire wind field.

The Arrival of Tropical Storm Force Winds map was created in response to EMs’ request for this information as this drives their preparation and evacuation timelines. The map developed for this purpose uses several distinct gradations of one color (red) to illustrate the approximate times (early Friday morning, Saturday, etc.). Support for this graphic is strong in terms of both ease of understanding and usefulness, with some reservations expressed about how the general public might use this information incorrectly.

NCAR Societal Impacts Program 32 ETTC Emergency Manager Survey

When asked about their relationships with local WFOs, all give positive responses. They were then afforded an opportunity to discuss any special issues or challenges they have in receiving and using NWS ET and TC products. Changes in product formats are called for most often, including more graphics and simpler templates, more storm surge products, earlier dissemination of forecast products in the storm cycle, reference to well-known landmarks, more information about forecast certainty, and greater access to more complicated products such as model runs.

About a third of these jurisdictions use commercial weather vendors. Most frequent reasons cited are more timely information, information tailored to the local areas, additional model information, and greater data reliability. When asked how often they use public media sources for weather information, local TV is at the top of the list. Most also use local WFO, NHC, other NOAA websites, followed by state emergency management websites.

In closing, many of these emergency managers express appreciation for the opportunity to provide input into NWS severe storm products and services and ask for continued collaboration.

NCAR Societal Impacts Program 33

References

Demuth, J., R. E. Morss, B.H. Morrow, and J. K. Lazo. 2012. “Creation and Communication of Hurricane Risk Information.’ Bulletin of the American Meteorological Society. August:1133– 1145.

Lazo, J.K. and D.M. Waldman. 2011. “Valuing Improved Hurricane Forecasts.” Economics Letters. 111(1): 43–46.

Lazo, J.K., D.M. Waldman, B.H. Morrow, and J.A. Thacher. 2010. “Assessment of Household Evacuation Decision Making and the Benefits of Improved Hurricane Forecasting.” Weather and Forecasting. 25(1):207–219.

Lazrus, H., B.H. Morrow, R.E. Morss, and J.K. Lazo. March 2012. “Vulnerability Beyond Stereotypes: Context and Agency in Hurricane Risk Communication.” Submitted to Weather Climate and Society.

Margusity, H. 2012. “No Hurricane Warning for What Could be the Most Expensive Storm in History.” Available at: http://www.accuweather.com/en/weather-news/national-hurricane-center- no-a/839301. Accessed November 14, 2012.

Morrow, B.H. and J.K. Lazo. Under Review. “Coastal Emergency Managers’ Preferences for Storm Surge Forecast Communication.” Journal of Emergency Management.

Morrow, B.H. and J.K. Lazo. 2012. Coastal Emergency Managers On-Line Survey for Hurricane Forecast Improvement Program/Surge Roadmap Project. Final Report (under review).

NHC Public Affairs. 2012. “National Hurricane Center’s Views on the Use of Scales to Communicate the Storm Surge Hazard.” September 10. Available at: http://www.nhc.noaa.gov/news/20120910_pa_surgeScale.pdf. Accessed September 15, 2012.

Norcross, B. 2012. “Isaac – the Mess and the Message.” Bryan Norcross’ Official Blog. September 04. Available at: http://www.wunderground.com/blog/bnorcross/show.html. Accessed September 25, 2012.

ETTC Emergency Manager Survey

Schleifstein, M. 2012. “Surge Warnings Went Out Before Hurricane Isaac Hit.” The Times- Picayune. September 04. Available at: http://www.nola.com/hurricane/index.ssf/2012/09/surge_warnings_went_out_before.html. Accessed September 25, 2012.

Stemple, Kirk and Joseph Goldstein. 2012. “How a Beach Community Became a Deathtrap.” New York Times, November 10. Accessed at: http://www.nytimes.com/2012/11/11/nyregion/how-a-staten-island-community-became-a- deathtrap.html?pagewanted=all&_r=0. Accessed November 15, 2012.

NCAR Societal Impacts Program 35

Appendix A. Advance Letter and Invitation Email Sent to Broadcast Meteorologists

ETTC Emergency Manager Survey

Advance letter

NCAR Societal Impacts Program App. A-2 ETTC Emergency Manager Survey

Invitation email

June 26, 2011

From: National Center for Atmospheric Research Subject: National Weather Service Storm Surge Roadmap

Dear [insert name]

You were recently contacted about participating in a survey being conducted to provide guidance on the National Weather Service (NWS) Surge Roadmap Project. Information in this email will explain how to participate.

The survey will ask for your opinion about various forecast practices and products being considered for use on NWS websites when coastal areas are threatened by severe storms. Select broadcast meteorologists from major television stations in Atlantic, Gulf and Pacific coastal areas, as well as Alaska, are being asked to participate.

As an expert in the presentation of weather information, we are soliciting your opinion.

Click on the following link to take a survey: [link]

Please be sure that the link has not wrapped. If you are prompted for a key, enter the key as follows: [key]

Attention AOL Users: If the survey does not load when you click on the above link, please click here.

Dr. Jeff Lazo (National Corporation for Atmospheric Research) and Dr. Betty Morrow (SocResearch Miami) are directing the project. Research Exec, a professional Internet survey company is administering the survey. Do not hesitate to email us questions or concerns.

YOUR PARTICIPATION IS VERY IMPORTANT.

Thanking you in advance,

Jeff Lazo ([email protected]) and Betty Morrow ([email protected])

NCAR Societal Impacts Program App. A-3

Appendix B. Survey Questions and Summary of Responses

App. B-1

NWS Storm Surge – Emergency Manager Survey August 2012 OE Responses not included to maintain respondent confidentiality

INTRODUCTION

Important information about this survey. Please read!

You can influence how the National Weather Service (NWS) communicates tropical and extratropical system forecasts on its websites. We are collecting opinions from a select group of experienced broadcast meteorologists from coastal areas of the Atlantic, Gulf and Pacific, as well as Alaska. As part of the selected sample, your opinions are very important to the success of this project. The Societal Impacts Program at the National Center for Atmospheric Research (NCAR) is conducting this study to inform the development and testing of tropical and extratropical severe storm products. This is not a commercial survey. This survey builds on the results from earlier surveys, webinars and discussions in which you may have been involved. As a professional emergency manager we are very interested in your insights. The survey should take you about 20 minutes to complete. We will analyze your responses together with all others, thus preserving confidentiality. Neither your name nor that of your agency or jurisdiction will be reported in the results. Completing this survey is voluntary. By clicking on the “AGREE AND CONTINUE” button below you are indicating that you have read this and agree to participate in this survey. Otherwise, please click on the “Exit” button.

[“Agree and continue” button – required to click on this to continue – if not, they will exit the survey. “Exit” button will take them to a Thank You screen and end data collection]

App. B-1 ETTC Emergency Manager Survey

HAZARD VULNERABILITY OF YOUR JURISDICTION

1.) How often, if at all is your jurisdiction affected by each of these weather-related hazards?

Sub-question Variable Name Never Rarely Occasionally Often Frequently Mean Median SD n # missing Q1 1 2 3 4 5 75 31 11 0 0 1.45 1 0.66 117 0 Tsunami 64.1% 26.5% 9.4% 0.0% 0.0% 39 40 22 7 9 2.21 2 1.19 117 0 Snowstorms 33.3% 34.2% 18.8% 6.0% 7.7% 5 42 51 16 3 2.74 3 0.84 117 0 Hail 4.3% 35.9% 43.6% 13.7% 2.6% 1 1 30 46 39 4.03 4 0.84 117 0 Heavy rain 0.9% 0.9% 25.6% 39.3% 33.3% 10 29 40 29 9 2.98 3 1.07 117 0 Wildfires 8.5% 24.8% 34.2% 24.8% 7.7% 8 14 36 41 18 3.40 4 1.10 117 0 High waves 6.8% 12.0% 30.8% 35.0% 15.4% 2 1 24 55 35 4.03 4 0.84 117 0 Coastal storms 1.7% 0.9% 20.5% 47.0% 29.9% 74 19 16 6 2 1.66 1 1.01 117 0 Mud slides 63.2% 16.2% 13.7% 5.1% 1.7% 14 24 50 17 12 2.91 3 1.11 117 0 Flash floods 12.0% 20.5% 42.7% 14.5% 10.3% 39 44 25 8 1 2.04 2 0.95 117 0 Ice storms 33.3% 37.6% 21.4% 6.8% 0.9% 18 40 45 12 2 2.49 3 0.93 117 0 Tornadoes 15.4% 34.2% 38.5% 10.3% 1.7% 6 16 45 37 13 3.30 3 1.01 117 0 Heat waves 5.1% 13.7% 38.5% 31.6% 11.1%

2.) What other weather-related hazards might significantly affect your area? OPEN-ENDED RESPONSE

NCAR Societal Impacts Program App. A-2 ETTC Emergency Manager Survey

3.) In general, how prepared do you think people in your jurisdiction are for each of these hazards?

Not at all Somewhat Very Extremely Does not Sub-question Variable Name Prepared Mean Median SD n # missing prepared prepared prepared prepared apply Q3 1 2 3 4 5 6 48 28 9 4 2 24 1.73 1 0.97 115 2 Tsunami 41.7% 24.3% 7.8% 3.5% 1.7% 20.9% 34 29 17 11 1 23 2.09 2 1.07 115 2 Snowstorms 29.6% 25.2% 14.8% 9.6% 0.9% 20.0% 21 45 35 9 2 3 2.34 2 0.94 115 2 Hail 18.3% 39.1% 30.4% 7.8% 1.7% 2.6% 0 17 55 31 12 0 3.33 3 0.86 115 2 Heavy rain 0.0% 14.8% 47.8% 27.0% 10.4% 0.0% 11 51 30 14 3 6 2.51 2 0.94 115 2 Wildfires 9.6% 44.3% 26.1% 12.2% 2.6% 5.2% 6 39 40 21 4 5 2.80 3 0.94 115 2 High waves 5.2% 33.9% 34.8% 18.3% 3.5% 4.3% 0 24 38 39 12 2 3.35 3 0.93 115 2 Coastal storms 0.0% 20.9% 33.0% 33.9% 10.4% 1.7% 39 23 9 3 0 41 1.68 1 0.85 115 2 Mud slides 33.9% 20.0% 7.8% 2.6% 0.0% 35.7% 11 52 29 12 3 8 2.48 2 0.92 115 2 Flash floods 9.6% 45.2% 25.2% 10.4% 2.6% 7.0% 34 45 10 4 1 21 1.86 2 0.85 115 2 Ice storms 29.6% 39.1% 8.7% 3.5% 0.9% 18.3% 26 52 23 7 3 4 2.18 2 0.96 115 2 Tornadoes 22.6% 45.2% 20.0% 6.1% 2.6% 3.5% 4 39 44 20 5 3 2.85 3 0.91 115 2 Heat waves 3.5% 33.9% 38.3% 17.4% 4.3% 2.6%

NCAR Societal Impacts Program App. A-3 ETTC Emergency Manager Survey

4.) Tropical storms include tropical cyclones and hurricanes. An extratropical storm is a large, strong coastal storm that causes large waves and high water levels along the coast, causing flooding and severe erosion. Extratropical storms include extra-tropical cyclones, sometimes called mid-latitude cyclones or wave cyclones, in the middle or high latitudes. For instance, a nor’easter is a common extratropical storm that moves along the east coast of North America with winds blowing from a northeasterly direction. How would you rate the vulnerability of your jurisdiction to tropical storms and extratropical storms that produce at least gale-force (39-54 mph) winds?

Extremely Very Somewhat Not at all Sub-question Variable Name Vulnerable Mean Median SD n # missing vulnerable vulnerable vulnerable vulnerable Q4 1 2 3 4 5 48 24 12 18 11 2.29 2 1.41 113 4 Tropical Storms 42.5% 21.2% 10.6% 15.9% 9.7% 41 21 16 25 10 2.49 2 1.40 113 4 Extratropical Storms 36.8% 18.4% 14.0% 21.9% 8.8%

5.) All tropical and extratropical storm systems are different, but in general, if a severe storm were predicted for your area, to what extent would you be concerned about each of the following hazards?

Extremely Very A little Not concerned Does not # Sub-question Variable Name Concerned Mean Median SD n concerned concerned concerned at all apply missing Q5 1 2 3 4 5 6 49 42 16 6 0 0 1.81 2 0.87 113 4 Storm Surge 43.4% 37.2% 14.2% 5.3% 0.0% 0.0% 39 54 18 2 0 0 1.85 2 0.75 113 4 Wind 34.5% 47.8% 15.9% 1.8% 0.0% 0.0% 36 34 19 12 5 7 2.21 2 1.17 113 4 Tornadoes 31.9% 30.1% 16.8% 10.6% 4.4% 6.2% 37 39 25 11 0 1 2.09 2 0.97 113 4 Inland Flooding 32.7% 34.5% 22.1% 9.7% 0.0% 0.9% 10 16 21 16 10 40 3.00 3 1.25 113 4 Heavy Snow 8.8% 14.2% 18.6% 14.2% 8.8% 35.4%

NCAR Societal Impacts Program App. A-4 ETTC Emergency Manager Survey

6.) What other hazards associated with tropical or extratropical storms are of special concern in your area? OPEN-ENDED RESPONSE

7.) What portion of your jurisdiction’s population do you think live in areas vulnerable to storm surge or coastal flooding?

Variable # None 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Other Mean Median SD n Name missing

Q7 1 2 3 4 5 6 7 8 9 10 11 12 0 18 20 17 10 7 9 12 6 4 4 4 5.41 5 2.90 111 6

0.0% 16.2% 18.0% 15.3% 9.0% 6.3% 8.1% 10.8% 5.4% 3.6% 3.6% 3.6%

8.) Of those who are vulnerable, what portion would do you think adequately understand their storm surge or coastal flooding vulnerability?

Variable # None 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Other Mean Median SD n Name missing

Q8 1 2 3 4 5 6 7 8 9 10 11 12 1 17 10 22 11 20 5 4 11 8 1 1 5.38 5 2.61 111 6

0.9% 15.3% 9.0% 19.8% 9.9% 18.0% 4.5% 3.6% 9.9% 7.2% 0.9% 0.9%

9.) How many times would you estimate your agency has been activated during the last 10 years for the following?

More # Sub-question Variable Name 0 1-2 3-4 5-6 7-8 9-10 Mean Median SD n than 10 missing Q9 1 2 3 4 5 6 7 16 11 17 26 15 5 20 3.98 4 1.96 110 7 Tropical Storms 14.5% 10.0% 15.5% 23.6% 13.6% 4.5% 18.2% 22 12 29 22 7 5 13 3.43 3 1.87 110 7 Hurricanes 20.0% 10.9% 26.4% 20.0% 6.4% 4.5% 11.8% 28 31 17 14 7 6 7 2.88 2 1.79 110 7 Extratropical Storms 25.5% 28.2% 15.5% 12.7% 6.4% 5.5% 6.4%

NCAR Societal Impacts Program App. A-5 ETTC Emergency Manager Survey

STORM SURGE FORECAST PRODUCT

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.

12.) To what extent do you agree or disagree with the following options:

Strongly Strongly # Sub-question Variable Name Agree Undecided Disagree Mean Median SD n Disagree Disagree missing Q12 1 2 3 4 5 The NWS should 9 2 3 41 55 4.19 5 1.15 110 7 issue Storm Surge 8.2% 1.8% 2.7% 37.3% 50.0% Warnings The NWS should 9 3 7 42 49 4.08 4 1.17 110 7 issue Storm Surge 8.2% 2.7% 6.4% 38.2% 44.5% Watches

13.) Why or why not? OPEN-ENDED RESPONSE

14.) If a separate warning product is issued for the surge or coastal flooding associated with tropical and extratropical storms, how would you evaluate each of these possible names?

Very # Sub-question Variable Name Poor Fair Good Excellent Mean Median SD n Good missing Q14 1 2 3 4 5 Extreme Coastal 15 16 26 28 22 3.24 3 1.32 107 10

Flood Warning 14.0% 15.0% 24.3% 26.2% 20.6% 3 8 24 41 31 3.83 4 1.02 107 10 Storm Surge Warning 2.8% 7.5% 22.4% 38.3% 29.0% 22 21 32 21 11 2.79 3 1.26 107 10 Storm Surge Risk 20.6% 19.6% 29.9% 19.6% 10.3% 22 25 31 18 11 2.73 3 1.26 107 10 Storm Surge Danger 20.6% 23.4% 29.0% 16.8% 10.3%

NCAR Societal Impacts Program App. A-6 ETTC Emergency Manager Survey

15.) What other name would you suggest or what other comments would you like to make about what it should be called? OPEN-ENDED RESPONSE

16.) Do you think a separate NWS warning for surge would result in the public paying greater attention to a coastal flooding or storm surge threat?

Definitely Probably # Variable Name Possibly Probably Definitely Mean Median SD n not not missing Q16 1 2 3 4 5 1 9 32 38 27 3.76 4 0.96 107 10

0.9% 8.4% 29.9% 35.5% 25.2%

17.) Do you think a separate NWS warning for storm surge would result in a better informed response by your jurisdiction?

Definitely Probably # Variable Name Possibly Probably Definitely Mean Median SD n not not missing Q17 1 2 3 4 5 2 6 23 43 33 3.93 4 0.96 107 10

1.9% 5.6% 21.5% 40.2% 30.8%

18.) Would a separate NWS warning for storm surge result in greater emphasis of coastal flooding or storm surge threat in your emergency management decision making?

Definitely Probably # Variable Name Possibly Probably Definitely Mean Median SD n not not missing Q18 1 2 3 4 5 4 6 22 34 41 3.95 4 1.08 107 10

3.7% 5.6% 20.6% 31.8% 38.3%

NCAR Societal Impacts Program App. A-7 ETTC Emergency Manager Survey

If a separate warning for storm surge is issued, a map will show the area included in the warning, as currently done for other warnings. (Local inundation maps will also be released.) The color purple is used in this example in order to avoid colors used for other NWS watches and warnings.

19.) How effective do you think a map like this would be in communicating to the public the area under storm surge or coastal flooding warning?

Not Not very Very Extremely # Variable Name effective at Effective Mean Median SD n effective effective effective missing all Q19 1 2 3 4 5 0 8 29 44 26 3.82 4 0.89 107 10

0.0% 7.5% 27.1% 41.1% 24.3%

20.) Please provide any additional comments or thoughts about this map. OPEN-ENDED RESPONSE

NCAR Societal Impacts Program App. A-8 ETTC Emergency Manager Survey

COASTAL FLOODING MAPS

To improve storm surge communication in tropical events the National Hurricane Center (NHC) is developing storm-specific inundation maps for tropical cyclones for its websites. Similar maps could be developed for extratropical storm surge events in the future. Therefore, we welcome the opinions of those of you from regions not subject to tropical cyclones. These maps show estimates of the amount of ocean water (including tides) expected on top of the land. The elevation above sea level of that general area will already have been subtracted.

21.) We are searching for the best way to label this so people understand. Please assess how clear or easy to understand you think each of these labels would be for the public.

Not at all Not very Extremely # Sub-question Variable Name Clear Very clear Mean Median SD n clear clear clear missing Q21 1 2 3 4 5 Above Ground 10 25 38 22 10 2.97 3 1.10 105 12

Level 9.5% 23.8% 36.2% 21.0% 9.5% 11 48 29 13 4 2.53 2 0.97 105 12 Above Ground 10.5% 45.7% 27.6% 12.4% 3.8% 33 52 14 3 3 1.96 2 0.91 105 12 Above Elevation 31.4% 49.5% 13.3% 2.9% 2.9% Depth of Water 5 18 41 29 12 3.24 3 1.02 105 12

Above Land 4.8% 17.1% 39.0% 27.6% 11.4% Height of Water 2 16 37 30 20 3.48 3 1.03 105 12

Above Land 1.9% 15.2% 35.2% 28.6% 19.0%

22.) Please suggest another label you think would be clear to the public. OPEN-ENDED RESPONSE

NCAR Societal Impacts Program App. A-9 ETTC Emergency Manager Survey

Next you will see 3 different ways coastal flooding might be mapped. A past hurricane forecast for Lee County, Florida is used for the example. The NHC would put maps like this on its website for any area under a surge warning. Similar maps might be developed in the future for extratropical storms. You will be asked to comment on each map individually, and then to choose the one you think would be most effective. Please click CONTINUE to proceed.

23.) Taking the perspective of the general public, how would you evaluate this map in terms of ease of understanding?

Variable Name Poor Fair Good Very Good Excellent Mean Median SD n # missing Q23 1 2 3 4 5 3 16 39 34 13 3.36 3 0.98 105 12

2.9% 15.2% 37.1% 32.4% 12.4%

24.) How would you rate the usefulness of this map for communicating to the public what they need to know about potential storm surge?

Variable Name Poor Fair Good Very Good Excellent Mean Median SD n # missing Q24 1 2 3 4 5 6 20 39 27 13 3.20 3 1.07 105 12

5.7% 19.0% 37.1% 25.7% 12.4%

25.) Please provide any additional comments or thoughts about this map. OPEN-ENDED RESPONSE

NCAR Societal Impacts Program App. A-10 ETTC Emergency Manager Survey

26.) Taking the perspective of the general public, how would you evaluate this map in terms of ease of understanding?

Very # Variable Name Poor Fair Good Excellent Mean Median SD n Good missing Q26 1 2 3 4 5 6 20 29 36 12 3.27 3 1.09 103 14

5.8% 19.4% 28.2% 35.0% 11.7%

27.) How would you rate the usefulness of this map for communicating to the public what they need to know about potential storm surge?

Very # Variable Name Poor Fair Good Excellent Mean Median SD n Good missing Q27 1 2 3 4 5 10 18 34 29 12 3.15 3 1.14 103 14

9.7% 17.5% 33.0% 28.2% 11.7%

28.) Please provide any additional comments or thoughts about this map. OPEN-ENDED RESPONSE

NCAR Societal Impacts Program App. A-11 ETTC Emergency Manager Survey

29.) Taking the perspective of the general public, how would you evaluate this map in terms of ease of understanding?

Very Variable Name Poor Fair Good Excellent Mean Median SD n # missing Good Q29 1 2 3 4 5 4 10 12 37 39 3.95 4 1.12 102 15

3.9% 9.8% 11.8% 36.3% 38.2%

30.) How would you rate the usefulness of this map for communicating to the public what they need to know about potential storm surge?

Very # Variable Name Poor Fair Good Excellent Mean Median SD n Good missing Q30 1 2 3 4 5 5 11 12 37 37 3.88 4 1.16 102 15

4.9% 10.8% 11.8% 36.3% 36.3%

31.) Please provide any additional comments or thoughts about this map. OPEN-ENDED RESPONSE

NCAR Societal Impacts Program App. A-12 ETTC Emergency Manager Survey

All Blue

Blue with Shading

Multi-Colored

NCAR Societal Impacts Program App. A-13 ETTC Emergency Manager Survey

32.) If the NHC uses only one of these inundation maps on its website, which do you think is BEST?

Blue with Shading – All Blue – one shade of Multi-Colored – colors shades of blue show range # Variable Name blue shows entire area show 4 levels with potential Mean Median SD n of potential storm surge missing under threat range of surge depth for each depth Q32 1 2 3 14 7 81 2.66 3 0.71 102 15

13.7% 6.9% 79.4%

33.) Why do you think it is best? OPEN-ENDED RESPONSE

NCAR Societal Impacts Program App. A-14 ETTC Emergency Manager Survey

34.) Thinking about the one you choose, how would you rate it on these qualities?

Sub-question Variable Name Poor Fair Good Very Good Excellent Mean Median SD n # missing Q34 1 2 3 4 5 Easy to 0 3 15 50 34 4.13 4 0.77 102 15

Understand 0.0% 2.9% 14.7% 49.0% 33.3% Provides Useful 2 0 16 51 33 4.11 4 0.81 102 15

Information 2.0% 0.0% 15.7% 50.0% 32.4% Communicates 2 3 12 48 37 4.13 4 0.88 102 15 the Risk 2.0% 2.9% 11.8% 47.1% 36.3% Promotes 5 9 19 45 24 3.73 4 1.07 102 15

Protective Action 4.9% 8.8% 18.6% 44.1% 23.5%

35.) Please provide any additional comments or thoughts regarding surge inundation maps. OPEN-ENDED RESPONSE

36.) We need to ask again in order to direct you to the next section of the survey - Do you work in an area that can be impacted by tropical cyclones?

Variable Name Yes No Q36 1 2 81 21

79.4% 20.6%  If “No”, go to Q63

NCAR Societal Impacts Program App. A-15 ETTC Emergency Manager Survey

TROPICAL CYCLONE FORECAST CONE

Below is the Forecast Cone as it appeared in the Hurricane Irene forecast. As you know, the Cone represents the probable track of the center of a tropical cyclone. It uses official forecast errors over a 5-year sample to estimate that the center will remain within the cone approximately two-thirds of the time.

37.) Taking the perspective of the general public, how would you evaluate this map in terms of ease of understanding?

Very # Variable Name Poor Fair Good Excellent Mean Median SD n* Good missing Q37 1 2 3 4 5 1 10 30 31 9 3.46 4 0.90 81 0

1.2% 12.3% 37.0% 38.3% 11.1%  n* = 81 because it does not include the 21 who answered “no” to Q36

NCAR Societal Impacts Program App. A-16 ETTC Emergency Manager Survey

38.) How would you rate the usefulness of this map for communicating to the public what they need to know about the track of a threatening tropical cyclone?

Very # Variable Name Poor Fair Good Excellent Mean Median SD n* Good missing Q38 1 2 3 4 5 1 12 25 36 7 3.44 4 0.89 81 0

1.2% 14.8% 30.9% 44.4% 8.6%

39.) Please provide any additional comments or thoughts about this graphic. OPEN-ENDED RESPONSE

NCAR Societal Impacts Program App. A-17 ETTC Emergency Manager Survey

40.) The Forecast Cone in the next graphic is transparent with DASHED LINES to indicate that the storm can extend beyond the cone. Taking the perspective of the general public, how would you evaluate this map in terms of ease of understanding?

Very # Variable Name Poor Fair Good Excellent Mean Median SD n* Good missing Q40 1 2 3 4 5 2 26 30 19 4 2.96 3 0.93 81 0

2.5% 32.1% 37.0% 23.5% 4.9%

41.) How would you rate the usefulness of this map for communicating to the public what they need to know about their chance of being affected by this storm?

Very # Variable Name Poor Fair Good Excellent Mean Median SD n* Good missing Q41 1 2 3 4 5 5 23 31 19 3 2.90 3 0.96 81 0

6.2% 28.4% 38.3% 23.5% 3.7%

42.) Please provide any additional comments or thoughts about this graphic. OPEN-ENDED RESPONSE

NCAR Societal Impacts Program App. A-18 ETTC Emergency Manager Survey

43.) The Forecast Cone in the next graphic is transparent with NO LINES to indicate that the storm can extend beyond the cone. Taking the perspective of the general public, how would you evaluate this map in terms of ease of understanding?

Very # Variable Name Poor Fair Good Excellent Mean Median SD n* Good missing Q43 1 2 3 4 5 8 16 37 18 2 2.88 3 0.95 81 0

9.9% 19.8% 45.7% 22.2% 2.5%

44.) How would you rate the usefulness of this map for communicating to the public what they need to know about their chance of being affected by this storm?

Very # Variable Name Poor Fair Good Excellent Mean Median SD n* Good missing Q44 1 2 3 4 5 8 15 38 18 2 2.89 3 0.95 81 0

9.9% 18.5% 46.9% 22.2% 2.5%

45.) Please provide any additional comments or thoughts about this graphic. OPEN-ENDED RESPONSE

NCAR Societal Impacts Program App. A-19 ETTC Emergency Manager Survey

Current Map

Map with Dashed Lines

Map with no Dashed Lines

NCAR Societal Impacts Program App. A-20 ETTC Emergency Manager Survey

46.) Which do you prefer?

Current forecast Transparent cone Transparent cone # Variable Name Mean Median SD n* cone with dashed lines with no lines missing Q46 1 2 3 52 12 17 1.57 1 0.82 81 0

64.2% 14.8% 21.0%

47.) Please provide your suggestions or comments regarding the Track Forecast Cone. OPEN-ENDED RESPONSE

NCAR Societal Impacts Program App. A-21 ETTC Emergency Manager Survey

POTENTIAL FOR DAMAGING WINDS The next map shows areas with varying potential for experiencing damaging winds from this storm.

48.) Taking the perspective of the general public, how would you evaluate this map in terms of ease of understanding?

Variable Name Poor Fair Good Very Good Excellent Mean Median SD n* # missing

Q48 1 2 3 4 5 1 6 28 29 16 3.66 4 0.93 80 1

1.3% 7.5% 35.0% 36.3% 20.0%

49.) How would you rate the usefulness of this map for communicating to the public what they need to know about their chances of experiencing damaging winds?

Variable Name Poor Fair Good Very Good Excellent Mean Median SD n* # missing

Q49 1 2 3 4 5 1 4 30 29 16 3.69 4 0.89 80 1

1.3% 5.0% 37.5% 36.3% 20.0%

50.) Please provide any additional comments or thoughts about this graphic. OPEN-ENDED RESPONSE

NCAR Societal Impacts Program App. A-22 ETTC Emergency Manager Survey

WIND PLUS CONE The next two maps combine the track cone and damaging wind information.

51.) Taking the perspective of the general public, how would you evaluate this map in terms of ease of understanding?

Variable Name Poor Fair Good Very Good Excellent Mean Median SD n* # missing Q51 1 2 3 4 5 7 13 19 32 9 3.29 4 1.14 80 1

8.8% 16.3% 23.8% 40.0% 11.3%

52.) How would you rate the usefulness of this map for communicating to the public what they need to know about their potential for being affected by this storm?

Variable Name Poor Fair Good Very Good Excellent Mean Median SD n* # missing Q52 1 2 3 4 5 7 13 18 34 8 3.29 4 1.13 80 1

8.8% 16.3% 22.5% 42.5% 10.0%

53.) Please provide any additional comments or thoughts about this graphic. OPEN-ENDED RESPONSE

NCAR Societal Impacts Program App. A-23 ETTC Emergency Manager Survey

54.) This graphic also combines the track and wind information, but this time it’s only illustrated over the land area. Taking the perspective of the general public, how would you evaluate this map in terms of ease of understanding?

Variable Name Poor Fair Good Very Good Excellent Mean Median SD n* # missing Q54 1 2 3 4 5 8 19 17 28 8 3.11 3 1.18 80 1

10.0% 23.8% 21.3% 35.0% 10.0%

55.) How would you rate the usefulness of this map for communicating to the public what they need to know about their potential for being affected by this storm?

Variable Name Poor Fair Good Very Good Excellent Mean Median SD n* # missing

Q55 1 2 3 4 5 8 21 14 29 8 3.10 3 1.20 80 1

10.0% 26.3% 17.5% 36.3% 10.0%

56.) Please provide any additional comments or thoughts about this graphic. OPEN-ENDED RESPONSE

NCAR Societal Impacts Program App. A-24 ETTC Emergency Manager Survey

On Land & Water

On Land Only

NCAR Societal Impacts Program App. A-25 ETTC Emergency Manager Survey

57.) The areas most likely to be affected by this storm are coastal North Carolina and Virginia. Which map do you think best communicates the urgency to those areas?

Variable Name On land & water On land only Not sure/don’t know Mean Median SD n* # missing Q57 1 2 3 51 26 3 1.34 1 0.48 80 1

63.8% 32.5% 3.8%

58.) Why do you think this is the best one? OPEN-ENDED RESPONSE

59.) Do you have any comments you’d like to make about these graphics? OPEN-ENDED RESPONSE

NCAR Societal Impacts Program App. A-26 ETTC Emergency Manager Survey

ARRIVAL OF TROPICAL STORM FORCE WINDS Emergency managers have been asking NWS to provide information about when to expect the arrival of tropical storm force winds (39 mph or higher that last at least one minute) as all preparations should be completed by then. The following map depicts the approximate time when areas on the East Coast should expect Tropical Storm Force Winds from this storm.

60.) Taking the perspective of the general public, how would you evaluate this map in terms of ease of understanding?

Variable Name Poor Fair Good Very Good Excellent Mean Median SD n* # missing Q60 1 2 3 4 5 4 10 28 29 9 3.36 3 1.01 80 1

5.0% 12.5% 35.0% 36.3% 11.3%

61.) How would you rate the usefulness of this map for communicating to the public when they should be ready?

Variable Name Poor Fair Good Very Good Excellent Mean Median SD n* # missing Q61 1 2 3 4 5 5 11 24 31 9 3.35 4 1.06 80 1

6.3% 13.8% 30.0% 38.8% 11.3%

62.) Please provide any additional comments or thoughts about this graphic. OPEN-ENDED RESPONSE

NCAR Societal Impacts Program App. A-27 ETTC Emergency Manager Survey

TROPICAL AND EXTRATROPICAL WEATHER INFORMATION SOURCES

63.) What, if any, special issues or challenges do you have in receiving and using NWS tropical and/or severe costal storm forecast and warning data in your work? OPEN-ENDED RESPONSE

64.) What, if anything, is the single most important change the NWS could make to improve its severe coastal storm forecast and warning products and services? OPEN-ENDED RESPONSE

65.) In what town or city is the NWS office for your jurisdiction located? If more than one serves your area, please list them all. If you are not sure or you do not know, simply click Continue to skip this question. OPEN-ENDED RESPONSE

66.) How would you rate your relationship with your local National Weather Service Forecast Office(s)?

Very # Variable Name Poor Fair Good Excellent Mean Median SD n Good missing Q66 1 2 3 4 5 0 0 5 19 77 4.71 5 0.55 101 16

0.0% 0.0% 5.0% 18.8% 76.2%

67.) Please provide any additional comments or thoughts about that relationship. OPEN-ENDED RESPONSE

68.) Do you use a paid commercial vendor for weather forecast information?

Variable Name Yes No Q68 1 2 29 72

28.7% 71.3%  if “yes”, go to Q68a and Q68b

NCAR Societal Impacts Program App. A-28 ETTC Emergency Manager Survey

68a.) Please list your paid commercial vendor for weather forecast information? OPEN-ENDED RESPONSE

68b.) To what extent is each of these an important reason for using a commercial vendor in addition to the products and services NWS provides?

Not Not very Very Extremely Sub-question Variable Name important Important Mean Median SD n* # missing important important important at all Q68b 1 2 3 4 5 2 3 19 2 3 3.03 3 0.94 29 0 Data reliability 6.9% 10.3% 65.5% 6.9% 10.3% 5 6 11 4 3 2.79 3 1.21 29 0 Better graphics 17.2% 20.7% 37.9% 13.8% 10.3% 2 5 11 8 3 3.17 3 1.07 29 0 Tailored to my market 6.9% 17.2% 37.9% 27.6% 10.3% 3 2 11 7 6 3.38 3 1.21 29 0 Timeliness 10.3% 6.9% 37.9% 24.1% 20.7% 2 5 12 8 2 3.10 3 1.01 29 0 More model information 6.9% 17.2% 41.4% 27.6% 6.9% 5 4 11 5 4 2.97 3 1.27 29 0 Interpretation of NWS data 17.2% 13.8% 37.9% 17.2% 13.8%  n* = 29 because it does not include the 72 who answered “no” to Q68

NCAR Societal Impacts Program App. A-29 ETTC Emergency Manager Survey

69.) If a hurricane or severe coastal storm threatens your area, to what extent do you go to each of the following sources for your forecast information?

Sub-question Variable Name Never Rarely Occasionally Often Frequently Mean Median SD n # missing

Q69 1 2 3 4 5 4 14 23 37 23 3.60 4 1.11 101 16 Local TV news 4.0% 13.9% 22.8% 36.6% 22.8% National TV (e.g., 8 21 33 23 16 3.18 3 1.17 101 16 ABC, CBS, NBC, CNN, or FOX) 7.9% 20.8% 32.7% 22.8% 15.8% The Weather Channel 6 16 34 22 23 3.40 3 1.18 101 16 on TV 5.9% 15.8% 33.7% 21.8% 22.8% 8 26 40 17 10 2.95 3 1.07 101 16 Local Radio 7.9% 25.7% 39.6% 16.8% 9.9% 17 18 33 19 14 2.95 3 1.27 101 16 Weather Underground 16.8% 17.8% 32.7% 18.8% 13.9% 8 20 22 21 30 3.45 4 1.32 101 16 NOAA Weather Radio 7.9% 19.8% 21.8% 20.8% 29.7% 2 1 7 29 62 4.47 5 0.83 101 16 Internet 2.0% 1.0% 6.9% 28.7% 61.4%

70.) What other public sources do you use? OPEN-ENDED RESPONSE

NCAR Societal Impacts Program App. A-30 ETTC Emergency Manager Survey

71.) How much will you probably use each of these websites for storm information?

Sub-question Variable Name Never Rarely Occasionally Often Frequently Mean Median SD n # missing

Q71 1 2 3 4 5 8 24 37 16 15 3.06 3 1.15 100 17 Local TV news websites 8.0% 24.0% 37.0% 16.0% 15.0% National TV news 15 29 35 12 9 2.71 3 1.14 100 17 websites 15.0% 29.0% 35.0% 12.0% 9.0% National Hurricane 14 2 5 18 61 4.10 5 1.42 100 17

Center website 14.0% 2.0% 5.0% 18.0% 61.0% Local office of weather 1 0 8 23 68 4.57 5 0.73 100 17 service website 1.0% 0.0% 8.0% 23.0% 68.0% 2 8 19 33 38 3.97 4 1.04 100 17 Other NOAA website 2.0% 8.0% 19.0% 33.0% 38.0% 11 22 40 15 12 2.95 3 1.14 100 17 Weather Channel website 11.0% 22.0% 40.0% 15.0% 12.0%

Weather Underground 19 23 25 17 16 2.88 3 1.34 100 17 website 19.0% 23.0% 25.0% 17.0% 16.0% 10 10 48 18 14 3.16 3 1.11 100 17 Other weather website 10.0% 10.0% 48.0% 18.0% 14.0% State emergency 3 13 29 28 27 3.63 4 1.11 100 17 management website 3.0% 13.0% 29.0% 28.0% 27.0%

72.) Are there any other websites you’d like to mention? OPEN-ENDED RESPONSE

NCAR Societal Impacts Program App. A-31 ETTC Emergency Manager Survey

SOCIO-DEMOGRAPHICS

D1. In what state or territory are you located? OPEN-ENDED RESPONSE

D2. What is the ZIPCODE of your office? OPEN-ENDED RESPONSE

D3. What is the official name of your agency? OPEN-ENDED RESPONSE

D4. Which best describes your jurisdiction?

Variable Name City or town County, parish, or borough Regional State Other Mean Median SD n # missing QD4 1 2 3 4 5 8 91 0 1 0 1.94 2 0.34 100 17

8.0% 91.0% 0.0% 1.0% 0.0%

D5. Please list the major counties, parishes or boroughs located within your jurisdiction. OPEN-ENDED RESPONSE

D6. Please list any major cities located within your jurisdiction. OPEN-ENDED RESPONSE

NCAR Societal Impacts Program App. A-32 ETTC Emergency Manager Survey

D7. What is the approximate population of your jurisdiction?

Between Between Between Under Over Variable Name 50,000 and 100,000 and 500,000 and Mean Median SD n # missing 50,000 1,000,000 100,000 500,000 1,000,000 QD7 1 2 3 4 5 33 20 31 8 8 2.38 2 1.25 100 17

33.0% 20.0% 31.0% 8.0% 8.0%

D8. What is your official title?

Director or Chief Director of Emergency Variable Name of Emergency Public Safety Other Mean Median SD n # missing Manager Management (or equivalent) QD8 1 2 3 4 58 28 0 14 1.70 1 1.03 100 17

58.0% 28.0% 0.0% 14.0%

D9. How many years have you been in emergency management? OPEN-ENDED RESPONSE

D10. How long have you been in your current position? OPEN-ENDED RESPONSE

D12. Finally – If you know of another emergency manager who would provide useful information for this survey please provide us with his or her name, email address, and telephone number and we will invite them to complete the survey as well. Note- they cannot simply complete it using the same link you were provided but must be "invited." OPEN-ENDED RESPONSE

D13. What else would you like to say regarding the NWS tropical or extratropical cyclone products or about this survey? OPEN-ENDED RESPONSE

COMPLETION: We greatly appreciate the time you took to complete this survey. Thank you! Please click the FINISH button below to record your responses.

NCAR Societal Impacts Program App. A-33