Socio-Economic Research on Hurricane Forecasts and Warnings

Jeffrey K. LazoSocietal Impacts Program

National Center for Atmospheric Research

Interdepartmental Hurricane ConferenceMarch 2, 2010

Picture “borrowed” from http://www.atmos.washington.edu/~houze/

Overview

“Understanding Hurricane Response for Improved Stakeholder / User Reaction.”

My Presentation

Part 1: Hurricane Forecast Socio-Economic Working Group

Part 2: Benefits of Improved Hurricane Forecasting

Part 3: Current Research– Hurricane Forecast Improvement Project– Communicating Hurricane Information – Warning Decisions: Extreme Weather Events

Part 1: Hurricane Forecast Socio-Economic Working Group

“The Group will recommend research initiatives and projects that can be supported through interagency cooperation, funding for public and private sector academic and commercial research enterprises, and partnerships with private sector information consumers.”

Hurricane Isabelle - September 18 2003

Hurricane Forecast Socio-Economic Working GroupFocal Areas

1. Warning Process

2. Decision Making

3. Evacuation Response Behavior

4. Societal Impacts and Valuation

Plan to develop:

• applied research agenda to generate short-term immediate benefits

• basic research agenda addressing fundamental theoretical and exploratory research designed to generate long-term improvements

• methods to enable the social science research community to gather and further develop research priorities and future agendas

• concept for a long-term, multidisciplinary, institutional approach to undertaking identified research priorities.

Hurricane Forecast Socio-Economic Working Group

Part 2: Benefits of Improved Hurricane Forecasting

Purpose: exploring methods for deriving household values for improved hurricane forecasts– Non-market valuation approach– Small sample implementation– Evacuation decision making– Benefit estimation

Forthcoming in Weather and Forecasting

Benefits of Improved Hurricane ForecastingEvacuation decision making

HurricaneHurricane

3.79(1.40)

3.40(1.47)

2.58(1.42)

1.96(1.20)

1.47(0.82)

1.00

2.00

3.00

4.00

5.00

Category 1 Category 2 Category 3 Category 4 Category 5

Hurricane Category on Saffir-Simpson Scale

Mea

n R

espo

nse

Mean Likelihood of Evacuation by Hurricane CategoryStandard deviation reported in parenthesis.

1 = “Not at all likely” to 5 = “Extremely likely” n = 80

WTP Calculation: Improve Baseline to Intermediate on All Attributes

AttributeBaseline

(all 48 hours in advance)

Intermediate Improvement

Diff.Marg. WTP

WTP

Time of expected landfall

± 8 hours ± 6 hours 2 $2.18 $4.36

Maximum wind speed

± 20 mph ± 15 mph 5 $0.26 $1.30

Projected location of landfall

± 100 miles ± 80 miles 20 $0.23 $4.60

Expected storm surge

±8’ of height above sea level

± 6’ of height above sea level

2 $2.04 $4.08

Total WTP $14.34

Part 3: Current Research

Societal Impacts Program / NCAR

– with multiple co-PIs

1. Hurricane Forecast Improvement Project

2. Communicating Hurricane Information

3. Hurricane and Flood Warning Decisions

4. Storm Surge

Hurricane Forecast Improvement Project

• Post-Katrina Assessments • HFIP

Hurricane Forecast Improvement Project

• HFIP Metrics– Reduce average track error by 50% for Days 1 through 5.

– Reduce average intensity error by 50% for Days 1 through 5.

– Increase the probability of detection (POD) for rapid intensity change to 90% at Day 1 decreasing linearly to 60% at Day 5, and decrease the false alarm ratio (FAR) for rapid intensity change to 10% for Day 1 increasing linearly to 30% at Day 5.

– Extend the lead time for hurricane forecasts out to Day 7

• 10-year program – multiple team research areas– About $20m/yr for 10 years

– Current social science – about $150k

Hurricane Forecast Improvement Project

• Socio-Economic Impacts Assessment– Assessment of Emergency Managers - Betty Morrow

in-depth focused interviews emergency managers stakeholder communities (hospitals / transportation / etc)

– Household valuation – Jeff Lazo non-market stated choice assessment adapted Benefits of Improved Hurricane Forecasting attribute set from HFIP 400 sample across the vulnerable region

Communicating Hurricane Information

• Examining the Hurricane Warning System: Content, Channels, and Comprehension – NSF-NOAA joint announcement of opportunity

• Research foci– How is the content of hurricane forecast and warning

messages generated, and what products result?– What are the channels through which hurricane forecast

and warning information is communicated, and what drives channel selection and use?

– How does the public (including vulnerable populations) comprehend and react to specific components of the forecast and warning messages?

Communicating Hurricane Information

• All text and graphical products at Days 4, 3, 2, and 1 prior to landfall

• Parallel studies in Miami and Houston areas• Mock hurricane hitting these areas

Communicating Hurricane Information

Communicating Hurricane Information

• Methods:– interviews and observations of message development with

forecasters, broadcast media, emergency managers– survey examining how members of the public access

information and their comprehension of and reactions to different messages

– focus groups with vulnerable populations (Miami only)– laboratory tests of sample messages with members of the

public– multi-method synthesis of public component and feedback to

forecast and emergency management communities through Expert Advisory Board

1) Message content – decisions, factors that influence the content of messages provided to others

2) Forecast/job mechanics – steps, actions, factors affecting mechanics of forecasters, EMs, and broadcasters

3) Interactions – content & channels of information (a) among partners, and (b) provided to recipients

4) Message development – data & information influencing process of what information is used in creating warnings

5) Interpretations – challenges of and factors influencing how to analyze and consider the data and/or message content

6) Uncertainty – data or messages that include or reference some ambiguity about the current state or future situation

Communicating Hurricane Information

Hurricane and Flood Warning Decisions

• Warning Decisions in Extreme Weather Events: An Integrated Multi-Method Approach – Funding from NSF Human and Social Dynamics program– 3 year project

• Research foci– How are hurricane / flash flood warnings

communicated, obtained, interpreted, and used in decision making by participants in the warning process?

– Challenges for decision making in the face of risk and uncertainty

Hurricane and Flood Warning Decisions

Hurricane and Flood Warning Decisions

• Parallel studies – Flash floods in Boulder, Colorado

– Hurricanes in Miami, Florida

• Methods– interviews, focus group discussions with forecasters, media,

public officials

– mental models with forecasters, media, public officials, members of public

– stated-preference survey with members of public (Miami only)

– multi-method synthesis

– stakeholder workshop

Likelihood of flash flooding occurring in BOU during warning or watch? Likelihood of flash flooding occurring in BOU during warning or watch?

Researchers Advisors

Ann Bostrom – Risk Communication

Julie Demuth – Meteorology / Communication

Gina Eosco – Communication

Somer Erickson – Emergency Management

Brandi Gilbert – Sociology

Hugh Gladwin – Sociology

Jennifer Hudson – Public Administration

Matthew Jensen – Mgmt. Information Systems

Jeff Lazo – Economics

Claude Miller – Communication

Betty Morrow – Sociology

Rebecca Morss – Meteorology

Dan O’Hair – Communication

Kathleen Tierney – Sociology

Jennifer Thacher – Economics

Don Waldman – Economics

David BernardFrank Billingsley

Luis CarreraChristopher Davis

Mark DeMaria Kelvin Droegemeier

Gene HafeleTim Heller

Greg HollandChuck LanzaMax Mayfield

Bryan NorcrossFrank Redding

Jamie Rhome