Presentation by Robert Soden of University of Colorado, Boulder on joint Crisis Informatics workshop organized by Kathmandu Living Labs, Harvard Humanitarian Initiative and University of Colorado Boulder to draw lessons from Nepal Earthquake 2015.
Repurposing LNG terminals for Hydrogen Ammonia: Feasibility and Cost Saving
Collective Information Infrastructures During Disaster Response
1. Collective Information Infrastructures
During Disaster Response:
Initial Findings from the
2015 Nepal Earthquake
Robert Soden
PhD Student, Computer Science
University of Colorado, Boulder
2. Research Question
What are the characteristics of the
information infrastructure that supported
the work of GIS and mapping teams during
the Nepal earthquake response?
In particular, what has been the impact of
community mapping and open data
projects?
3. Information Infrastructure
the people, processes, and tools that support
the creation, maintenance, and use of
information
networks, protocols, standards, formal and
informal social arrangements
when they function they are invisible
4. Methodology
• 40 interviews with GIS and information
management officers from government,
international organizations, and Nepali
technical organizations
• Trying to develop a detailed understanding of
how these people did their work, what maps
did they produce and why, what data did they
use and why, how did they distribute their
products and how were they used
Information infrastructure is a funny word but I think it evokes a few things that are useful for our talk today. First, when we are talking about infomration and mapping in disaster response we aren’t talking about isolated datasets just sitting on one person’s laptop. We’re instead talking about networks of people, social and technical processes, hardware and software that allow for collaborative efforts to create, share, update, and communicate data. Second infrastructure is something that we need to invest in the creation and maintenance of and that we need to think very carefully about the design.
Ok so I just have a few minutes to present so I want to focus my remarks
OpenStreetMap was the default basemap for the majority of the GIS teams
This is true of both international organizations, many of whom have come to expect it, and government and local organization
What’s more interesting though is the incredible diversity of organizations who used it, the things they used it for, and the ways in which it was presented.
Detailed mapping of an area of 10,000km2 in 4 days, including coverage of road networks, hiking trails, built-up areas, building footprints, river crossings and temporary relief camps
Quadrupled road mileage and added 30% more buildings in 48 hours
Identified 15 priority areas, 8 of which have been completed and validated
Attracted over 2,000 volunteer contributors from around the world, 1/3 of whom are new mappers
Made maps available on the web as half hourly data exports, print maps, and offline maps for Android
USAID team with a paper basemap of gorkha
Search and rescue map by mapaction. Here OSM is used as a basemap and they add operational data on top of it.
PDF Map of IDP camps by the Red Cross. They have a complicated workflow by which they extract raw data from OSM and bring into their own open source GIS tools where they have a number of map templates set up
PDF Map published on the webiste of the national geographic information infrastructure project (ngiip) of road networks in lalitpur using OSM
Interactive ebmap produced by ICIMOD in support of the Ministry of Home Affairs which uses OSM as a basemap
Logistics map made by WFP
Mobile App developed by Kathmandu University and ICIMOD
Sometimes GIS people obsess over cartographic scale, precision, and positional accuracy when talking about GIS data. This is especially the case when it comes to data from crowdsourcing or volunteers. When talking about OpenStreetMap the first question that people always used to ask me was about data quality, how can we ensure it we have accurate data? A lot of the initial academic research into OSM, carried out by Muki Haklay at University College London going back as far as 2007 was about OSM data quality
Interestingly this has not been the case here in Nepal. When interviewing GIS teams about which data source they used and why, data quality was far from the most important consideration. If people did have concerns about it, they would usually say that it was sufficient or their purposes or that OSM was the best available. Far more important for these individiauls when choosing which data the would use for their maps were considerations such as access – OSM data is open whereas many other datasets were not, previous experience using OSM, or personal recommendations
Sometimes GIS people obsess over cartographic scale, precision, and positional accuracy when talking about GIS data. This is especially the case when it comes to data from crowdsourcing or volunteers. When talking about OpenStreetMap the first question that people always used to ask me was about data quality, how can we ensure it we have accurate data? A lot of the initial academic research into OSM, carried out by Muki Haklay at University College London going back as far as 2007 was about OSM data quality
Interestingly this has not been the case here in Nepal. When interviewing GIS teams about which data source they used and why, data quality was far from the most important consideration. If people did have concerns about it, they would usually say that it was sufficient or their purposes or that OSM was the best available. Far more important for these individiauls when choosing which data the would use for their maps were considerations such as access – OSM data is open whereas many other datasets were not, previous experience using OSM, or personal recommendations from trusted individuals
Sometimes GIS people obsess over cartographic scale, precision, and positional accuracy when talking about GIS data. This is especially the case when it comes to data from crowdsourcing or volunteers. When talking about OpenStreetMap the first question that people always used to ask me was about data quality, how can we ensure it we have accurate data? A lot of the initial academic research into OSM, carried out by Muki Haklay at University College London going back as far as 2007 was about OSM data quality
Interestingly this has not been the case here in Nepal. When interviewing GIS teams about which data source they used and why, data quality was far from the most important consideration. If people did have concerns about it, they would usually say that it was sufficient or their purposes or that OSM was the best available. Far more important for these individiauls when choosing which data the would use for their maps were considerations such as access – OSM data is open whereas many other datasets were not, previous experience using OSM, or personal recommendations from trusted individuals
Sometimes GIS people obsess over cartographic scale, precision, and positional accuracy when talking about GIS data. This is especially the case when it comes to data from crowdsourcing or volunteers. When talking about OpenStreetMap the first question that people always used to ask me was about data quality, how can we ensure it we have accurate data? A lot of the initial academic research into OSM, carried out by Muki Haklay at University College London going back as far as 2007 was about OSM data quality
Interestingly this has not been the case here in Nepal. When interviewing GIS teams about which data source they used and why, data quality was far from the most important consideration. If people did have concerns about it, they would usually say that it was sufficient or their purposes or that OSM was the best available. Far more important for these individiauls when choosing which data the would use for their maps were considerations such as access – OSM data is open whereas many other datasets were not, previous experience using OSM, or personal recommendations from trusted individuals
KLL played an important role both before and during the quake
Creation of data
Outreach and Community Building
Interface between local responders and the international OSM community
On demand map production
KLL not the only one
Information systems related to disasters are necessarily collective resources - there are many different users and uses involved No one actor has all of the information, everyone needs information that other organizations create or manage.
Yet our information infrastructures don't always accommodate this
The design of information infrastructure can help communities manage collective information resource successfully, or contribute to inefficiences and information failures, and in the case of disaster response, the failure to deliver lifesaving humanitarian aid in effective ways.
We saw during the Nepal Earthquake response and the example of OpenStreetMap that investment in information infrastructures that have openness, accessibility, and collaboration as components is a successful strategy. What we need to now as a community is to understand in more detail how to design them and continue to grow this effort in Nepal.
For that reason I am very happy to be here and participate in this workshop with you all today