okanagan mountain park fire kelowna bc 2003
play

Okanagan Mountain Park Fire (Kelowna, BC, 2003) 64000 acres, $33.8 - PowerPoint PPT Presentation

V ISUALIZING W ILDFIRE N ARRATIVES LAS 2015 D ECEMBER 2, 2015 C HRISTOPHER G. H EALEY B RANDA N OWELL AJ F AAS N ORTH C AROLINA S TATE U NIVERSITY S AN J OSE S TATE U NIVERSITY SIC PARVIS MAGNA NC STATE UNIVERSITY


  1. V ISUALIZING W ILDFIRE N ARRATIVES LAS 2015 D ECEMBER 2, 2015 C HRISTOPHER G. H EALEY B RANDA N OWELL AJ F AAS N ORTH C AROLINA S TATE U NIVERSITY S AN J OSE S TATE U NIVERSITY SIC PARVIS MAGNA NC STATE UNIVERSITY

  2. http://www.icarus.ca/icarus/?p=1024 Okanagan Mountain Park Fire (Kelowna, BC, 2003) 64000 acres, $33.8 million,239 homes destroyed

  3. N ATIONAL C OHESIVE S TRATEGY To safely and effective extinguish fire, when needed; use fire where allowable; manage our natural resources; and as a Nation, live with wildland fire. National Cohesive Wildland Fire Management Strategy April, 2014 http://www.forestsandrangelands.gov/strategy

  4. Number of Fires and Acres Burned 300 12,000 Thousands Thousands 250 10,000 200 8,000 Acres Fires 150 6,000 100 4,000 50 2,000 0 0 1960 1962 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 Year Fires Acres National Interagency Fire Center http://www.nifc.gov/fireInfo/fireInfo_stats_totalFires.html

  5. Acres Thousands 1,000 2,000 2,500 3,000 1,500 3,500 500 0 1825 Miramachi, 3M 1845 Great Fire, 1.5M 1853 Yaquina, 450K 1868 Coos, 300K http://www.nifc.gov/fireInfo/fireInfo_stats_histSigFires.html 1881 Lower Michigan, 2.5M National Interagency Fire Center 1898 South Carolina, 3M Twenty-Five Largest Fires By Acres 1903 Adirondack, 637K 1910 Great Idaho, 3M 1918 Cloquet-Moose Lake, 1.2M 1933 Tillamook, 311K 1987 Seige of '87, 640K 1988 Yellowstone, 1.6M Year 1997 Inowak, 610K 1999 Dunn-Glen, 288K 2002 Rodeo-Chediski, 462K 2003 Cedar, 275K 2004 Taylor, 1.3M 2006 East Amarillo, 907K 2007 Big Turnaround, 388K 2007 Murphy, 652K 2010 Long Butte, 300K 2011 Wallow, 538K 2012 Whitewater-Baldy, 298K 2012 Long Draw, 558K 2013 Rim, 257K

  6. USFS / DOI Wildfire Costs 2,000 Millions 1,800 1,600 1,400 1,200 Dollars 1,000 800 600 400 200 0 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 Year US Forest Service Department of Interior National Interagency Fire Center http://www.nifc.gov/fireInfo/fireInfo_documents/SuppCosts.pdf

  7. P ROJECT O BJECTIVES • What are dominant wildfire and risk narratives communicated through social media? • How are narratives shaped by ecological, social, and political characteristics? • How do narratives evolve over time and in response to actual wildfire events? • How can social media be used to engage the public and wildfire risk and risk-mitigation?

  8. Colorado Springs Fire Department Wildfire Mitigation http://www.springsgov.com/Page.aspx?NavID=101

  9. P ROJECT P LAN 1. Capture, index, store wildfire incident Twitter communication 2. Perform thematic and sentiment analysis of tweets 3. Analyze and visualize information flow within socialmedia networks • Community engagement: for risk mitigation prior to a wildfire • Communication: between emergency management and communities during wildfire events

  10. D ATA C APTURE • Capturing tweets with keywords “wildfire” and “forest service” since May 14, 2013 – 3 million tweets stored in MySQL database • Extracting relevant tweet properties – date and time – author – body – geolocation • DenverCP | -104.994593,39.746012 | Wildfire burning SW of Beulah closes Hwy. 165: BEULAH, Colo. A small wildfire burned in Pueblo County just... http://t.co/FfPeLrpIS6 | Sun Jun 01 02:21:12 +0000 2014

  11. T EXT V ISUALIZATION • Visualize properties of “documents” in a “collection” • We analyze individual tweets as short text documents – sentiment, topics, term frequency, geolocation, affinity

  12. S INGLE T WEET V ISUALIZATION unpleasant pleasant pleasure hue sedate active arousal luminance low confidence high confidence rating variation opacity low confidence high confidence rating frequency size

  13. S ENTIMENT S CATTERPLOT King Fire (El Dorado County, CA, 2014) 97717 acres, $91 million, 80 residences destroyed

  14. S ENTIMENT V ALUES King Fire (El Dorado County, CA, 2014) 97717 acres, $91 million, 80 residences destroyed

  15. T OPIC C LUSTERS King Fire (El Dorado County, CA, 2014) 97717 acres, $91 million, 80 residences destroyed

  16. T OPIC C LUSTERS King Fire (El Dorado County, CA, 2014) 97717 acres, $91 million, 80 residences destroyed

  17. S ENTIMENT H EATMAP King Fire (El Dorado County, CA, 2014) 97717 acres, $91 million, 80 residences destroyed

  18. S ENTIMENT T AG C LOUD King Fire (El Dorado County, CA, 2014) 97717 acres, $91 million, 80 residences destroyed

  19. S ENTIMENT T IMELINE King Fire (El Dorado County, CA, 2014) 97717 acres, $91 million, 80 residences destroyed

  20. G EOLOCATION King Fire (El Dorado County, CA, 2014) 97717 acres, $91 million, 80 residences destroyed

  21. A FFINITY G RAPH King Fire (El Dorado County, CA, 2014) 97717 acres, $91 million, 80 residences destroyed

  22. T WEET T EXT King Fire (El Dorado County, CA, 2014) 97717 acres, $91 million, 80 residences destroyed

  23. C URRENT R ESEARCH C HALLENGES • Contextual tweet filtering: via natural language processing • Narrative threads: construction and visualization • Real-world validation: with JFS public information officers Visualizing character threads from “Star Wars IV: A New Hope”

  24. TWITTER NARRATIVE • Identify narrative threads passing through “anchor” tweet Sep 23, 10:15am psapconnectnews: Californi a Wildfire Crews Brace For Weather Shift : The King Fire thread 1 region is expected to experience erractic win… http://t.co/Qg9RKNC10B thread 2 anchor tweet thread 3

  25. THREAD CONSTRUCTION 1. Find all tweets above similarity ε to anchor tweet 2. Bundle similar tweets plus anchor into “similarity sets” 3. Convert similarity sets into DAGs where: – tweets connected if pairwise similarity > τ, τ < ε – edge direction based on time, older to newer 4. Identify source nodes as nodes with no incoming edges 5. Apply Bellman-Ford on DAGs to find longest path from every source node 6. Keep any longest paths that contain anchor, these form narrative threads

  26. C ONTACT I NFORMATION C HRISTOPHER G. H EALEY HEALEY @ NCSU . EDU WWW . CSC . NCSU . EDU / FACULTY / HEALEY F IRE CHASERS P ROJECT RESEARCH . CNR . NCSU . EDU / BLOGS / FIRECHASERS T WEET V ISUALIZER WWW . CSC . NCSU . EDU / FACULTY / HEALEY / TWEET _ VIZ / TWEET _ APP NC STATE UNIVERSITY

Download Presentation
Download Policy: The content available on the website is offered to you 'AS IS' for your personal information and use only. It cannot be commercialized, licensed, or distributed on other websites without prior consent from the author. To download a presentation, simply click this link. If you encounter any difficulties during the download process, it's possible that the publisher has removed the file from their server.

Recommend


More recommend