For the past few weeks, I’ve been highlighting scientific findings made possible by citizen science that appeared in the literature each week. Methods of public engagement in sharing observations are not only useful to science, but also in a wide range of areas that need reliable information, such as urban planning, public health, environmental justice, and disaster relief. Consequently, the practice of citizen science is itself an area of innovation and active inquiry.
As I’ve pointed out before, the inquiry often includes comparisons of experts and amateurs. This weeks was no exception with the following two examples:
(1) Mining urban deprivation from foursquare: implicit crowdsourcing of city land use. by Quercia and Saez.
Even when people share information for non-scientific reason, that shared information can be useful for investigations. In this example, the authors created maps of land-uses based on data from the social-media users of Foursquare in London. These maps were comparable to proprietary commercial maps. Plus, the maps revealed that well-off neighborhoods were more likely to have amenities that promote health and wellbeing, such as dance studios, hobby shops, pools, tea rooms, movie theaters, and kid stores. The poor areas were more likely to have health threats, such as factories, light rail, airports, strip clubs, and whisky bars. (This paper was part of a special feature on citizen science in the IEEE journal Pervasive Computing. Other articles from the special feature are summarized in the footnote at the end of this post).
(2) Conducting disaster assessment with Spatial video, experts, and citizens. by Lue et al in Applied Geography 52:46-54.
In this case, Lue et al. compared the effectiveness of laypeople at CrisisMappers and experts at the American Red Cross using video to carrying out damage assessment after a natural disaster. Assessments from experienced and inexperienced people were similar, though assessing damage from video turned out to be a difficult task, irrespective of on-the-ground experience.
The potential for engaging crowds in solving problems related to disasters is widely recognized but there are many obstacles to overcome since the inquiry inherently needs to be quick. Individuals can be quick, but rarely can large crowds be coordinated to accomplish anything quickly. Let’s look at examples where citizen science happens quickly after earthquakes.
Earthquake studies and responses
There are an estimated 500,000 earthquakes annually. About 100,000 release enough energy close to the surface (that is, high magnitude) that the seismic waves produce shaking that people can feel (that is, high intensity). But we cannot anticipate when they will occur.
One great example to illustrate speedy citizen science is a project called Did You Feel It? In 1997 the USGS moved their post-earthquake survey from the mail to online. At first it was called Community Internet Intensity Maps, and more recently adopted the name Did You Feel It? With the (more than) 2,790,000 responses so far, Did You Feel It? uses an algorithm to quantify earthquake intensity by processing data on where, what was observed and what was experienced by people. Post-quake, people report remarkably similar experiences which leads to quick consensus on intensity, calculated to one decimal point, and detection of quakes even under 2.0 magnitude.
Some projects recruit people to host earthquake sensors. With these fairly effortless contributions, automated systems using low-cost micro-electronic accelerometers provide essential data to scientists. Participatory sensor networks include the Quake-Catcher Network, the Community Seismic Network, SeisMac, and iShake Cal (under development, this project uses iPhones because they already contain motions sensors).
Other projects involve incidental contributions, such as through the Twitter Earthquake Dispatch (@USGSted) algorithm developed by a team at the National Earthquake Information Center. There are 500 million people using Twitter. Cumulatively, these users publically document events. People use twitter to learn about events before those events appear in the news. Since people re-tweet, when a earthquake is mentioned, the word “earthquake” is likely to be amplified quickly – which is an easy signal to detect automatically. The TED algorithm (which looks for increases in use of the word “earthquake” in several languages) can detect a quake within 2 minutes, though it tends to miss small ones.
Using citizen science to advance our understanding of earthquakes is only half of the story. Citizen science can aid response and rescue efforts, even from a distance. Through The Global Earth Observation Catastrophe Assessment Network (GEO-CAN), people can use the Virtual Disaster Viewer (VDV) to view satellite imagery of an area before and after quake, mark differences, add notations of damage grades, and share the information with emergency responders. The Internet and smart phones have been leveraged for citizen science in ways that support the work of first responders to disasters, such as earthquakes in Wenchuan, China in 2008, Haiti in 2010, and Christchurch, NZ in 2011. Non-governmental efforts use citizen science too. The Humanitarian Open Street Map is an online platform for data sharing for humanitarian responses.
Most citizen science does not hinge on speedy reporting. For example, about 15 percent of all bird observations in Project FeederWatch this past winter arrived on paper sheets marked with number 2 pencils, delivered to the Lab of Ornithology by the USPS. It is convenient that people can enter their bird observations online, but it is far from crucial, and even by mail the projects are likely faster than global collaborations in the early, quill-pen-wielding days of citizen science. Earthquakes carry urgency. At such times, information needs to be centralized rapidly. And as the old drug-testing joke goes, the one drug that won’t be found among postal workers is speed. Most extend that joke to other government agencies, all of which can get mired in bureaucracy. Then how is it that, for example, the USGS has been able to carry out speedy citizen science with Did You Feel It?
Last fall, the USGS and the Woodrow Wilson Center released a report on the tools of citizen seismology. The report is intended to provide lessons for other government entities wanting to develop citizen-science projects.
The trick to crafting projects with fast response times is to get all ducks in a row in anticipation of the inevitable events – there are 350 tiny quakes daily for study and invariably the big disasters that require emergency aid. Any entanglement in red tape must be prevented by moving through the obstacle course beforehand. It means navigating practical, legal, technical, policy, and ethical considerations of the whole system well ahead of time. Risk management associated with disasters is serious work: recall that in wake of deaths from after-shocks of the earthquake in L’Aquila in 2009, Italian courts convicted six scientists for manslaughter, sentencing them to 6 years in prison and $10.2 million in fines. Speedy citizen science means creating a mutually clear environment conducive to government-citizen collaborations.
The report covers two frequent stumbling blocks: the Privacy Act and the Paperwork Reduction Act. Most seismologists don’t have training in working with human subjects or experience with policies on how government entities can interact with citizens. The laws can be particularly cumbersome because they were not enacted with citizen science in mind.
The Federal Privacy Act of 1974 covers policies and procedures for how the government must handle personally identifiable information (name, social security number, finger prints, voice, photographs, and more). Until an agency begins assembling public data submissions, typically from computers with unique IP addresses, they may not have fully explored the scope of this Act.
The intentions of the Paperwork Reduction Act of 1980 are to reduce the burden of paperwork that the government imposes on citizens. Ironically, the Act puts an overburden of paperwork on agency staff and requests to gain approval for public data collection must be submitted to the Office of Management and Budget. The approval process lasts a minimum of 90 days, which includes 60-day and 30-day public comment periods. In 2009, a Presidential Directive about Open Government resulted in inclusion of social media in the mandates of this Act.
Taken together, the tools of citizen seismology provide rapid detection, information for emergency response, and information dissemination. We need agencies to get prepared. When crises arises, we need government systems in place to quickly draw on crowds to collectively build global maps of rapidly changing conditions, and assist in emergency response.
Yes, despite red tape, agencies can provide tools for crowds to work fast. Do you feel the ground shake? The citizen science possibilities register at 9 on my Richter scale.
Photo credit: Haiti earthquake damage in UN Photo/Logan Abassi, UN Development Programme, & screen capture of Virtual Disaster Viewer.
More Roundup: other articles in the special feature in Pervasive Computing included:
(1) Stevens et al. Taking participatory citizen science to extremes.
These authors help marginalized communities have a voice by supporting the communities to share their indigenous knowledge. They illustrate ways to structure citizen science to stimulate inclusion, and thereby empower communities. Visit the Extreme Citizen Science group at University College London to learn more.
(2) Bahanamonde et al. Mining private information from public data: the Transantigao case.
They use publicly available information on the smartcards cards that passengers use for daily travel on the public transportation system in Transantiago, Chile. Even though the data were anonymized, the researchers demonstrated how they can hone in on where people live and thus revealed and explored the privacy implications.
(3) Angus et al. Public goods: using pervasive computing to inspire grassroots activism.
The author illustrate how artists and engineers can work together. The artists bring the cultural interventions and the engineers bring the technical solutions. Together they offer creative, low-cost tech air pollution devices to create experiences around local concerns that mobilize communities.