NCAR’s Societal Impacts Program:

Research and OutreachJeffrey K. Lazo, Director

Julie Demuth, Associate ScientistNCAR Societal Impacts Program

NOAA/NWS SeminarNovember 1, 2007

Outline• Overview of NCAR’s Societal Impacts

Program (SIP)• Results from recent U.S. household

survey– Assess people’s views and attitudes about

weather forecast information, including uncertainty information

Overview of the SIP

Origins of SIP• Very little is currently known about:

– economic value of current and improved weather information

– users’ sources, perceptions, uses, and value of weather information

– users’ understanding of, use of, and preferences for weather forecast uncertainty information

– best practices for integrating new weather information into users’ environment

• Relevance to NOAA– Crucial to fulfilling NOAA mission– NOAA has minimal capacity to address these

issues – Greater demands for NOAA to justify research

and activities for budget decisions

Origins of SIP• Collaborative Program on Societal and

Economic Benefits of Weather Information– NOAA–NCAR collaboration– Peer reviewed– Unanimously accepted USWRP Interagency

Working Group, July 2003– Program commenced April 1, 2004, with hiring of

a Director• SIP as part of NCAR

– Independent non-federal– History of interdisciplinary research in the

interactions between weather and society– Unique within university community, tap

extensive expertise in social sciences

Scope of SIP• Mission

– Improve the societal gains from weather forecasting by infusing social science and economic research, methods, and capabilities into the planning, execution, and analysis of weather information, applications, and research directions.

• Implementation– Research– Outreach and education– WAS*IS program– Community support

Current SIP research• Communicating uncertainty in weather

forecasts– Will discuss methods and results in more detail

• Overall U.S. sector sensitivity assessment• Individual sector sensitivity assessment–

transportation• Warning decisions in extreme weather

events• Hurricane household valuation study• User needs assessment• Hydrometeorological testbed

Overall U.S. Sector Sensitivity Assessment (OUSSSA)

• “...one-third of the private industry activities, representing annual revenues of some $3 trillion, have some degree of weather and climate risk.” (Dutton, 2002)

• Evaluate sensitivity of 11 U.S. economic “super” sectors to weather variability

• Writing results for BAMS manuscript

Individual Sector Sensitivity Assessment–Transportation (ISSA-

T)• Assess transportation sector’s use and

value of weather forecast information– To understand how the transportation sector

is affected by weather and what weather information it deems useful

– To develop valid and reliable socio-economic methods for assessing use and value

• Methodology– 5 subsectors: air, rail, water, truck, pipeline– Eliciting judgments from experts in academia,

private sector, and government

Warning Decisions in Extreme Weather Events

• How extreme weather warnings are communicated, interpreted, and used by different participants

– Explore interactions among 4 groups: forecasters, public officials, media, public

– Focus on hurricanes and flash floods; challenges for decision making in the face of risk and uncertainty• Collaborators (NSF funded)

– Kathleen Tierney and Jeannette Sutton -- U. of Colorado, Natural Hazards Center

– Ann Bostrom -- U. of Washington, School of Public Affairs

Other SIP research• User needs assessment

– Summarize best practices for introducing weather decision-support technologies into new environments

• Hurricane household valuation study– Assess use and value of improved hurricane

forecasts to households• Hydrometeorological testbed

– Assess societal impacts and economic value for American River area in California

Weather and Society * Integrated Studies (WAS*IS)

• Vision: To change the weather enterprise by comprehensively and sustainably integrating social science into meteorological research and practice

• Mission– Build an interdisciplinary, grassroots community

of practitioners, researchers, & stakeholders who are dedicated to integrating meteorology and social science

– Provide opportunities to learn and examine ideas, methods, & examples related to integrated weather-society work

WAS*IS• Workshops

– Original Boulder WAS*IS (Nov 2005, Mar 2006)

– WAS*IS Norman (Apr 2006)– 2006 Summer WAS*IS (Jul 2006)– WAS*IS Australia (Jan-Feb 2007)– 2007 Summer WAS*IS (Jul 2007)• 145 WAS*ISers (so far)!

WAS*IS• Current and future activities

– Program evaluation– Edited compendium (NCAR assessment initiative

funding)– Interaction with AMS Policy Program– 2008 Summer WAS*IS workshop (tentative)

• Output– WAS*IS overview paper by Demuth et al. forthcoming in

November issue of BAMS– Conferences and meetings– Networking and collaborations

• Outcome– NWS “Advanced” Workshop, October 24-25– New research and activities

SIP outreach and education• Website

– Extreme Weather Sourcebook– Societal Aspects Page– Digital Library

• Weather and Society Watch Newsletter• Economics Primer• WxSoc newsgroup

SIP community support• Hurricane Forecast Socio-Economic Working

Group– Special issue Natural Hazards Review (August 2007)– Joint NSF-NOAA funding opportunity (FY08)

• THORPEX– Societal and Economic Research and Applications (SERA) WG– NAT SERA (BAMS article forthcoming)

• Economics primer– NOAA funded

• WMO Public Weather Service Taskforce• WMO CAS Strategic Planning• AMS Annual Partnership Topic on Hurricane

Disasters

Nationwide Survey of U.S. Households

To assess people’s views and attitudes about weather forecasts and weather forecast uncertainty

Motivation: Community recognition

• Per NRC (2006) “Completing the Forecast” report recommendations:– Overarching Rec 2: collaborating with users and

partners in the Enterprise and engaging and using social and behavioral science expertise

– Rec 4.4: study uncertainty terms, icons, and other communications methods

• AMS (2002) statement on enhancing weather information with probability forecasts

• Other NRC reports, conference sessions, etc. • Recent formation of the AMS Ad Hoc

Committee on Uncertainty in Forecasts (ACUF)

Objectives• To support the meteorological

community in effectively providing weather forecast information, including uncertainty, by:– Assessing people’s sources, perceptions, uses

and value of weather forecast information Also assess how these factors change over time

– Exploring people’s perception of, interpretation of, and preferences for weather forecast uncertainty information

Survey design and implementation

• Survey questions developed:– based on questions asked in previous survey

research– to investigate fundamental research

questions and issues raised by previous related work

• Pre-tested survey during development and implementation

• Implemented survey in November 2006– Controlled-access, web-based with sample

population provided by survey sampling company• Analysis based on N=1465 respondents

Weather forecast research questions

• From where and how often do people get weather forecast information?

• For what locations or regions do people get forecasts?

• What times of day do people get forecasts?

• For what reasons do people get forecasts?• What weather forecast parameters are

important to people?• What is people’s willingness to pay for

forecasts?

How often do you get forecasts from the sources listed below?

0 5 10 15 20 25 30 35 40

Local TV

Cable TV

Commercial or public radio

Other webpages

Newspapers

NWS webpages

Friends, family, co-workers, etc.

NOAA Weather Radio

Telephone weather info source

Cell phone, PDA, pager, etc.

Average household

accesses weather forecasts 115

times per month!

With over 113 million U.S. households, this totals to over 150 Billion forecasts

accessed a year!

Mean Monthly Access

On average, how often do you use weather forecasts for the activities

listed below?

0% 20% 40% 60% 80% 100%

Simply knowing whatthe weather will be like

Planning how to dressself or children

Planning weekendactivities

Planning travel

Planning yard work oroutdoor house work

Planning socialactivities

Planning getting towork or school

Planning job activities

Rarely or never Less than half the time About half the timeMore than half the time Usually or always Not applicable

How important is it to you to have the information below as part of a weather

forecast?

0% 20% 40% 60% 80% 100%

When precip will occurChance of precip

Where precip will occurType of precip

High tempAmount of precip

Chance of amount of precipLow temp

Wind speedHumidity

Time of day high temp will occurTime of day low temp will occur

CloudinessWind direction

Not at all important A little important Somewhat important Very important Extremely important

Do you feel that the NWS services you receive are worth more than, exactly, or less than $N a year to your household? • Given values for $N varied among $2, $5, $10,

$30, $60, $90, $120, $150, $180, $210, $240

$0

$50

$100

$150

$200

$250

0% 20% 40% 60% 80% 100%

Fitted Percent of Respondents

Will

ingn

ess

to P

ay

Uncertainty research questions• How much confidence do people have in

different types of weather forecasts?• Do people infer uncertainty into deterministic

forecasts and, if so, how much?• How do people interpret a type of uncertainty

forecast that is already commonly available and familiar: probability of precipitation forecasts?

• To what extent do people prefer to receive deterministic vs. uncertainty-explicit forecasts?

• In what formats do people prefer to receive forecast uncertainty information?

• Do people infer uncertainty into deterministic forecasts and, if so, how much?

(Perception)

Suppose the forecast high temperature for tomorrow for your

area is 75°F.What do you think the actual high temperature will be?

0%

10%

20%

30%

40%

50%

75°F 74-76°F(± 1°F)

73-77°F(± 2°F)

70-80°F(± 5°F)

65-85°F(± 10°F)

Other

Perc

ent o

f Res

pond

ents

• To what extent do people prefer to receive deterministic vs. uncertainty-explicit forecasts?

• In what formats do people prefer to receive forecast uncertainty information?(Preferences)

0% 10% 20% 30% 40% 50%

Prefer Channel A(deterministic)

Prefer Channel B(uncertainty)

Like both channels

Like neither channel

I don't know

Percent of Respondents

Suppose you are watching the local evening news …

• The Channel A weather forecaster says the high temperature will be 76°F tomorrow

• The Channel B weather forecaster says the high temperature will be between 74°F and 78°F tomorrow.

All the choices below are the same as a probability of

precipitation of 20%.Do you like the information given this way?

• Chance of precipitation is 20%• There is a 1 in 5 chance of

precipitation• The odds are 1 to 4 that it will rain• There is a slight chance of rain

tomorrow

Percent Frequency Odds Text

Asked this question 3 ways – using PoPs of 20%, 50%, and 80% with corresponding text descriptions from

NWS

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Percent Frequency Odds TextN = 489,489, 487

PoP of 20%

Percent of respondents who said “yes”

PoP of 50%PoP of 80%

To summarize• Weather forecast views and attitudes,

including on uncertainty– 150 billion served!– Majority of people like uncertainty info and many

prefer it– People have preferences for how uncertainty

information is conveyed

• Dissemination of results– Submitted manuscript with uncertainty results to

Weather and Forecasting this week– Will submit manuscript with sources, uses, perceptions

info to BAMS in early 2008Connecting physical and social science in these

ways will help more effectively communicate weather forecast information, including

uncertainty!

Future work• Future research questions to pursue

– People’s interpretations of and preferences for uncertainty information in other formats (including graphics), in different weather situations, across different media, etc.

– People’s use of different types of uncertainty information

• Grant funded this survey• Seeking additional funding for future

research in this area

Importance of social science• Empirical research employing both

quantitative and qualitative social science methods are needed– E.g., surveys, focus groups, interviews

• Need range of social science expertise– E.g., economics, communication, psychology,

decision science

Societal Impacts Program... aims to improve societal gains from weather forecasting by infusing social science and economic research, methods, and capabilities into the planning, execution, and analysis of weather information, applications, and research directions• research

• WAS*IS• outreach and education• community support

Thank you!• Contact

– Jeff Lazo, SIP Director (lazo@ucar.edu)– Julie Demuth (jdemuth@ucar.edu)– Rebecca Morss (morss@ucar.edu)

• www.sip.ucar.edu• www.sip.ucar.edu/wasis