�ݺ�ߣ

Editor's Notes

• #2: AlanGlennon, spatial@ucsb
• #3: This morning I'm going to discuss two use cases of non-authoritative crisis mapping: the Jesusita Fire and the Haitian Earthquake
• #4: In our area, we have numerous potential natural and manmade threats: from the land, sea, and air.
• #5: Generally, when responding to these crises, we think about response centered around government agencies and NGOS --
• #6: a room full of experts mapping and getting things right,accurate… and then vetted information is released via a press conference or bulletin.
• #7: …vetted information is released via a press conference or bulletin.
• #8: ...and from there, the press repackages it as stories for tv, radio, newspaper, and online.
• #9: In the progression, the public is primarily a consumer of data.
• #10: With the rise of the Internet, mobile phones, social networks, and online discussion, the public has begun inserting itself into the chain.
• #11: So, in this regard, let's look at the Jesusita Fire. The fire was started about 1:30pm on May 5, 2009. …the latest of several significant fires in our area. ...and for each of these fires, Internet discussion has been increasing and happening faster.
• #12: So, with the Jesusita Fire, when I learned about it, I started looking into ways to capture the information flows as the occurred during the fire.
• #13: I began writing a timeline of events.Noticing items like the first Internet references to the fire began less than thirty minutes after ignition.
• #14: Note, that the fire is at this point, a wildland fire -- away from any structures -- the smoke was visible though.In fact, after only a few minutes of origin postings, there's a complaint that there's no news anywhere.
• #15: Like anyone else, I had numerous data sources to pull from... and besides keeping notes on the simple accounting of events, I decided a map would work well to synthesize the data.The idea was not particularly original though -- I did a websearch and found several other online maps tagged with the name "Jesusita" -- particularly using Google's MyMaps feature.
• #16: MyMaps is a feature of Google Maps that allows users to annotate their existing base maps with points, polylines, and polygons, and share them, either as view only or as a collaborative. Among the JesusitaMyMaps, one -- by a user named "Ethan" already contained some county warning boundaries and was also fully editable by anyone. By the way, Ethan is actually the infant son of UCSB staff member Mark Grosch.
• #17: As I used the map, I began communicating with my fellow grad students and they started assisting too. We drew evacuation areas, the fire boundary, wind direction, school closings, and anything else we thought relevant. The crowd fixed erroneous data when it was contributed. The map was actively monitored until all evacuation were canceled on May 14th.
• #18: By the end of the Jesusita Fire, at least 34 related Google MyMaps had been created. The top seven garned over one million users, but the Ethan map led them all with nearly 600,000. The question was, "why?"
• #19: perhaps it involves community. Among all the maps, the Ethan map would allow users to change data and discuss when they were unsure. People could ask questions and get public responses. The map also persisted; this in contrast to some news media maps that were abandoned each day with data relevant to that day's story.
• #20: but it didn't all work well. all with a limited symbology set, all Google MyMap look similar. there were occasional errors and some were difficult to correct. we had exuberant users contribute large unrelated datasets, and we had no mechanism to elegantly correct them. -- we had to crowdsource the tedious corrective work.
• #21: So -- in all, there were many lessons learned from the Jesusitacrowdsourced mapping event -- particularly regarding the need for curatorial tools and the dynamics of real-time community discussion. By the way, since I am moving through some of these slides fast -- these presentation is available online at http://slideshare.net/glennon
• #22: The second case is the January 2010 Haitian earthquake. Unlike Jesusita, I did not conduct any direct mapping efforts. I have only been a keen observer of the situation -- as I am interested in leveraging the lessons and avoiding its failings. First, this was a massive natural and humanitarian disaster -- 10s of thousands of deaths and structures destroyed. The world attention was focused on this event.
• #23: I am sure many Internet-based efforts were attempted during the relief effort, but two in particular, appear to have been taken hold -- Ushahidi and OpenStreetMap.Ushahidi is an open source project that allows affected populations to send a message, via SMS, smartphone, email, or webpage, to a centralized repository.
• #24: For Haiti, a Ushahidi instance was up and running within 2 hours -- set up by a group of several students at tufts University. Through the course of the relief effort, the students would curate incoming messages, translate them, geocode them onto a map, and get the information to the Red Cross. Along the way, the engage several hundred local volunteers at the university and online -- and had thousands of Haitian volunteers help them with translation.
• #25: The second important crowdsourced mapping tool from the Haitian quake is OpenStreetMap. OpenStreetMap is an open editable global map project -- like a wikipedia for maps. Within 26 hours of the quake, Google and GeoEye provided post event imagery and a group of volunteers made the imagery available to the open source community. The community responded.
• #26: It created a services based on the imagery to be used in osm - marked damaged areas - used open source tools to create overlays, and provided data products (like shapefiles and GPS file exports updated every 5 minutes). Within days, OpenStreetMap became known on the ground as the best, most up-to-date map of Port-au-Prince. Here's an example of the work. before and after.
• #27: Pre-event
• #28: Post-event
• #29: Again not everything worked perfectly, large spatial datasets can be difficult to transfer and often had to be manually carried from one field group to the next, technology had to be assembled on the fly, and trust and partnerships had to be created out of necessity.
• #30: The bottom line though is that these grassroots, real-time community efforts can create useful data and, particularly in the case of Haiti, save lives.
• #31: To close, how can we on the south coast engage the successes of Jesusita and Haiti, and steer clear of the pitfalls? I think the first is to prepare for the community discussion, to know that it will happen, and have an idea of what to expect.have interested partners, like us, get together and figure out what's the best way to collaborateknow what's out there and how the community is likely to use it.equip ourselves with expertise to get things done.
• #32: This meeting today is the place to start such an effort.…Crisis Commons work session afterhours on the second Tuesday of the month.
• #33: Thank you.