8.2 Web-based data generated by the public
The way in which we interact with the web is changing very rapidly. Rather than passively reading content that has been posted up by others, many people now actively contribute content themselves. For example, if you buy a second-hand item from eBay, the chances are that the seller you buy from has been rated by other people who have previously bought goods from that seller. If you buy a book from Amazon, you may well find some reviews of the book written by other people who have read it. If you want to investigate a particular topic, such as visiting a city, you are likely to consult Wikipedia, an encyclopaedia which has been collectively written and edited by people from all over the world. This phenomenon – the idea that people can collaborate online to generate information and be content editors and writers, not just readers – has been termed ‘Web 2.0‘ (see below).
If Web 2.0 can generate book reviews, seller profiles and online encyclopaedias, can the technology also be used to generate spatial data that we might be able to use in a GIS? The answer appears to be ‘yes’. Probably the best known example of a spatial data set generated by the public is OpenStreetMap (OSM), a street map created by members of the public uploading GPS tracks from cycles and vehicles and editing attributes such as road names. OSM does contain some data of interest to those studying the environment: there is for example a land use layer within OSM, although this remains less complete than the street data that forms its core thematic interest. More recently, particularly in response to humanitarian crises, concentrated OSM update efforts have occurred to increase mapping accuracy for geographical locations with limited spatial information. For example, satellite imagery was made publically available for volunteers to digitise infrastructure in response to the devastating 2015 Nepal earthquakes (see references). Information generated collectively by the public in this way has been described as Volunteered Geographic Information (VGI), a term that was coined by Prof. Mike Goodchild in 2007.
There are several developments that have helped VGI data sets to develop. One development is a straightforward, flexible set of licencing arrangements known as creative commons that makes it easier for people to decide how they wish to share their information with others. On the technology side, VGI can be generated in many different ways. Many VGI sites draw on technologies such as Google Maps, but often also on other resources. The online photo repository site Flickr, for example, enables photographs to be geotagged with co-ordinates and uploaded onto the web. Online public domain photographs can then be queried using calls to an Application Programming Interface (API) for Flickr. Programmers with skills in languages such as Java can then identify photographs uploaded into Flickr taken in a particular area or time and tagged with a particular keyword. Similar analyses can be undertaken for a multitude of social media platforms where spatial data is recorded, consequently allowing spatial patterns to be analysed. For example, Twitter can provide geo-located Tweets which document activity around a particular topical issue. By searching for a ‘hashtag’ term, if spatial information has been stored then spatiotemporal tracking of data can occur e.g. searching “#Haiyan” to map progress of tropical cyclone Haiyan which ripped across the South Pacific and Southeast Asia in 2013. The latter example questions the term ‘volunteered’ and can raise a multitude of ethical questions regarding privacy of information.
More recently, following up on the development of environments like OSM that allow for collective creation and editing of geospatial data, sites have developed that enable amateur map makers – without access to specialist GIS software – to undertake map design and spatial analysis operations. These can range from basic web mapping interfaces that display and overlay data (e.g. ArcGIS Online https://www.arcgis.com/home/webmap/viewer.html) to more complex functionally in Open Source GIS software (e.g. QGIS http://www.qgis.org/en/site/). Similarly, sites have evolved which actively seek to engage the public in what has been termed ‘citizen science’, the contribution of spatial environmental information to develop collective datasets.
Historically, it has been argued by some that the highly technical skills and specialist software required to undertake GIS analyses prevent proper public scrutiny of the spatial analyses that underpin many environmental management decisions. Do developments like these mean that there is now potential for greater transparency and public scrutiny of the data processing that lies behind many environmental management decisions? This is questionable as much crowdsourced output data still remain inaccessible to the public, despite their contribution to the data collection processes.
Examples of environmental VGI
Although some of the best known examples of VGI do not involve environmental data sets, there are many examples that are environmental. These initiatives fall into different categories and often require varying levels of expertise to contribute data:
Started by professionals: Some sites are designed by professionals to enable the public to share ideas on a particular topic.
- For example, several environmental projects in Australia, reviewed in the reports here (http://researchonline.jcu.edu.au/14915/), have attempted to use web-based GIS as a means of public participation.
- eBird (http://ebird.org/content/ebird/about/) is a site that enables amateur ornithologists to upload and share bird sightings (follow the ‘user sightings’ link, then choose a US state and a period of several months to see some typical data).
- Another example was crowdsourcing data for rapid damage assessment following the 2010 Haiti earthquake to increase spatial information to aid response and recovery operations (Ghosh et al. 2011). Some basic understanding of satellite imagery was required by volunteers to visually process remote sensing imagery to digitise areas detected as damaged.
- Earthwatch was the founding organisation that supported the development of Global Freshwater Watch, a platform which engages citizen scientists in collecting and uploading water quality samples to contribute to an ongoing water quality monitoring database. https://freshwaterwatch.thewaterhub.org/
- A similar initiative is the World Water Monitoring Challenge http://www.monitorwater.org/
- Zooniverse is a website which crowdsources multiple projects, some of which have spatial mapping elements to them, such as the ‘Floating Forests’ project which is mapping global distributions of kelp (seaweed). https://www.zooniverse.org/projects/
- This community mapping platform enables community groups to monitor and map environmental quality in their neighbourhoods, with air quality a particular focus. https://communitymaps.org.uk/welcome
Environmental VGI as a by-product of other activity: Sometimes, environmental information can be generated as a sub-conscious by-product of other activity on the web. For example, flood events are often so striking that people are prepared to upload imagery of flooding. Spontaneously, many people uploaded photos of the 2007 summer floods in England onto sites such as Flickr and the BBC web site (e.g. http://www.bbc.co.uk/berkshire/content/articles/2007/07/23/flood_map_feature.shtml). These photos provide additional information about the depth and timing of flood events. Similarly, the much-publicised use of Twitter for mapping weather (e.g. https://uksnowmap.com/) finds very recent ‘tweets’ from people and extracts locations where people had mentioned particular weather conditions (e.g. ‘sunny’; ‘rain’; or used weather-related hashtagsc). The result was a ‘live’ map that depicted weather conditions, as described by people at different locations across the globe. ESRI has developed the capacity to add twitter feeds to ArcGIS Online web apps, drawing particularly on their disaster response experience (https://www.esri.com/arcgis-blog/products/sharing-collaboration/sharing-collaboration/add-tweets-using-the-public-information-template/).
Whether the VGI phenomenon will continue to grow is a matter of debate, as is the credibility of data generated in this way.
To illustrate how GIS data capture functionality has started to migrate into the ‘cloud’ and become available to those without access to GIS software, try undertaking the exercise in this pdf file, which uses geospatial data from the GeoWiki platform.
References (Essential reading for this learning object indicated by *)
The National Centre for Geographic Information and Analysis runs an ongoing project on Volunteered Geographic Information. More details are available here: http://www.ncgia.ucsb.edu/projects/vgi/
The Web 2.0 concept is normally associated with Tim O’Reilly. You can find out more about this idea here: http://oreilly.com/web2/archive/what-is-web-20.html
Possibly the most widely known example of Volunteered Geographic Information is OpenStreetMap, which can be found here: http://www.openstreetmap.org/
The Humanitarian OpenStreetmap Team (HOT) web site gives further examples of crowdsourcing around natural disasters, such as the Haiti earthquake: http://www.hotosm.org/
For more information on the Haiti crowdsourcing initiative see this paper:
Ghosh, S., Huyck, C. K., Greene, M., Gill, S. P., Bevington, J., Svekla, W., DesRoches, R., and Eguchi, R. T. (2011) Crowdsourcing for Rapid Damage Assessment: The Global Earth Observation Catastrophe Assessment Network (GEO-CAN). Earthquake Spectra October 2011, Vol. 27, No. S1, S179-S198.
Additional resources used in the practical exercise include this link:
…and Google Earth: http://www.google.com/earth/download/ge/agree.html