Society For Risk Analysis Annual Meeting 2012

Advancing Analysis

Session Schedule & Abstracts


* Disclaimer: All presentations represent the views of the authors, and not the organizations that support their research. Please apply the standard disclaimer that any opinions, findings, and conclusions or recommendations in abstracts, posters, and presentations at the meeting are those of the authors and do not necessarily reflect the views of any other organization or agency. Meeting attendees and authors should be aware that this disclaimer is intended to apply to all abstracts contained in this document. Authors who wish to emphasize this disclaimer should do so in their presentation or poster. In an effort to make the abstracts as concise as possible and easy for meeting participants to read, the abstracts have been formatted such that they exclude references to papers, affiliations, and/or funding sources. Authors who wish to provide attendees with this information should do so in their presentation or poster.

Common abbreviations

M3-D
Symposium: Using Maps to Communicate Geospatial Risk and Uncertainty: Weather Forecasts and Environmental Hazards

Room: Pacific Concourse F   1:30 - 3 PM

Chair(s): Lori Severtson

Sponsored by RCSG

Maps are now commonly used to communicate geospatial risk and hazard information to a variety of audiences for the purpose of informing decisions. Despite this, few studies have explored if and how the features of a map influence decisions and related beliefs. This symposium brings together speakers from various disciplines and institutions who will share study findings on this topic. Topics covered in these studies include weather forecast warnings for tornadoes and hurricanes, and environmental hazards of climate change and contaminated groundwater. These researchers explored the following influences on decisions and related beliefs: (1) symbols and text to convey the geospatial probability of forecasted tornado warnings, (2) the effect of alternate hurricane forecast maps on evacuation intent, (3) three different visual representations to convey the uncertainty of modeled climate change outcomes, and (4) map symbols and a supplementary alphanumeric table to convey the geospatial distribution of arsenic test results in private residential water wells. The primary disciplines represented by the speakers in this symposium include risk communication, geographic information science, and cartography.



M3-D.1  13:30  The influence of map design and hazard proximity on risk beliefs and mitigation intentions for maps of arsenic in private residential water wells. Severtson DJ*; University of Wisconsin-Madison   djsevert@wisc.edu

Abstract: The purpose of this repeated measures field experiment was to assess how four map-related variables influenced outcomes of beliefs about the mapped hazard, ambiguity, emotion, and behavioral intentions. Outcomes were assessed at the time of viewing and 3 weeks later. Maps depicted water test results for arsenic in private residential wells. Map variables were: (1) map color (symbolic or non-symbolic color), (2) a symbol to denote areas with no data (with or without symbol), (3) an alphanumeric table that supplemented map information (with or without table), and (4) an estimate of participants’ proximity-based hazard at their perceived map location. A full factorial design for the 3 dichotomous map variables generated 8 map interventions. A table-only control was also included. The sample was homeowners with private wells from a county with elevated arsenic levels. Spatial stratification and randomization were used to generate the sample and assign households to 1 of the 9 conditions. Of 1224 mailed Survey 1s, 809 (66%) were completed. Of those participants, 733 (91%) completed Survey 2. Structural equation modeling was used to examine the influence of map variables on intentions to test water for arsenic; how beliefs, ambiguity, and emotion mediated these relationships; how these relationships changed over time; and the influences of interaction effects among map variables. These relationships were examined within a context of participant’s characteristics that included prior risk beliefs, perceived numeracy, gender, and years in home. Preliminary results suggest that some map variables moderated the influences of others and that some map variables and interactions influenced how modeled relationships changed over time.

M3-D.2  13:50  Representations of risk space and decision-making in tornado warnings. Klockow KE*, McPherson RA, Thomas RP; University of Oklahoma   kklockow@ou.edu

Abstract: This presentation will highlight preliminary results from a set of cognitive decision experiments, testing the impacts of warning design choices on protective responses in a geospatial context. The presentation will specifically highlight the potential for communicating warning information probabilistically, e.g., per results of the National Severe Storms Laboratory Warn-on-Forecast program. Applying theories of risky choice and cartography, deterministic and probabilistic representations of tornado risk will be generated for use in the decision experiments. Evaluation will be based on the cumulative outcome of warning responses after numerous trials and subjective ratings of trust placed in the forecast system. Verbally delivered warning tiers will also be tested in the design. Results will highlight effects on warning response given distance-to-storm, boundary inclusion/exclusion, the interaction of graphical and verbal information, and symbolic elements introduced with probabilistic designs. A test for spatial cognition and risk aversion, in addition to select socio-demographic information, will be used to examine differences in performance and preference among participants.

M3-D.3  14:10  Mapping climate change uncertainty: A pilot study of effects on risk perceptions and decision making. Retchless DP*; Penn State University   dpr173@psu.edu

Abstract: While some research has shown that providing uncertainty information about a hazard may influence risk perceptions and improve decision making, the relationship between cartographic depictions of uncertain climate change, risk perceptions, and decision making remains largely unexplored. To address this gap, I conduct a pilot study to assess how three different ways of representing uncertainty on maps of climate change affect risk perceptions and decision making. Each of the maps shows a projection of global temperature change from 1981-2010 to 2071-2100 developed using the ensemble mean of 26 different CMIP5 computer models. Uncertainty representation methods tested include hatching or stippling (popular in IPCC and U.S. climate change reports); saturation ramps; and a “squares” method where the size of the colored portion of the gridded climate output will increase with certainty. A fourth map that does not show uncertainty is included as a control. Each of at least 24 participants is shown one of these four maps, and ranks regions indicated on the map in order of: magnitude of temperature increase; level of uncertainty (except in control); likelihood of harm to regional ecosystems; and suitability for an expensive infrastructure project with specific temperature requirements. Rankings and response times are compared across maps. A follow-up question asks participants how helpful they found the uncertainty information when making the infrastructure decision. I hypothesize that most map users will find the uncertainty information helpful, and that the choice of cartographic technique for representing uncertainty will have a significant effect on risk perceptions and decision making. It is hoped that the results will inform decisions about how – and whether – to incorporate uncertainty information into maps of climate change.

M3-D.4  14:30  Beyond the cone of uncertainty: Effect of alternate hurricane forecast maps on evacuation intent. Trumbo C*, Peek L, Lueck M, Marlatt H, McNolldy B; 1-4 COLORADO STATE UNIVERSITY; 5 UNIVERSITY OF MIAMI   ctrumbo@mac.com

Abstract: The National Hurricane Center (NHC) issues its hurricane forecasts to include a mapped representation in the form of a cone. According to the NHC: "The cone represents the probable position of a tropical cyclone's circulation center, and is made by drawing a set of circles centered at each forecast point—12, 24, 36, 48, and 72 hours for a three-day forecast, as well as 96 and 120 hours for a five-day forecast. The radius of each circle is equal to encompass two-thirds of the historical official forecast errors for the preceding five year period. The cone is then constructed by drawing a tangent line that connects the outside boundary of all the circles. The National Hurricane Center states that the entire track of the tropical cyclone 'can be expected to remain within the cone roughly 60-70% of the time.'" Public understanding of the forecast cone graphic has not been without controversy. Research evidence exists indicating that individuals don't fully understand the probabilistic nature of the area under the cone. Further, many individuals place excessive confidence on the centerline of the projection. As a consequence, various efforts are underway to examine alternate approaches to providing visual representation of hurricane track forecasts. This study presents the results of a small field experiment that tests the effect of including omnibus coastal impact information in addition to the cone. The inclusion of coastal impact depiction is tested for two storm scenarios, a projected direct hit and a side-glance. Data for the 2X2 experiment were collected in June 2012 from study participants sampled from within a 10-mile coastal buffer running from Wilmington, NC to Brownsville, TX. Participants are part of a broader panel study in which the first wave of data collection was conducted in July 2010 (n = 653, 56% adjusted rate) and the second panel wave conducted in June 2011 (n = 450, 72% adjusted rate for panelist retention). Results pending.



[back to schedule]