IEEE Geoscience and Remote Sensing Magazine - September 2016 - 39

SCIENCE OF THE PEOPLE
Citizens have a long history of being involved in scientific
research or what has more recently been referred to as CS
[1]. One of the main drivers behind the recent proliferation
of CS projects has been technological, i.e., interactivity made
possible via Web 2.0, resulting in a thriving culture of social
media, movement toward the Internet of Things through
smart sensors, and global positioning system (GPS)-enabled
mobile devices. These components have made it possible
for citizens to become environmental sensors, collecting
and analyzing information on a massive scale that also has
real scientific value. There are many successful examples of
CS that have led to new scientific discoveries, such as new
knowledge about protein structures [2], discovering new galaxies [3], and websites for public reporting of illegal logging/
deforestation [4] and illegal waste dumping [5]. These have
demonstrated how citizens can have a visible impact on the
environment and local governance.
Another significant development has been the opening up of satellite imagery for viewing purposes through
providers, such as Google Earth and Bing, that has given citizens access to vast volumes of spatial data about
the entire world. This trend continues as we move into
the Sentinel Era of big data, where access to the data is
truly open. Complementing the vast amounts of information already being collected via Sentinel 1, which is
part of the European Space Agency's (ESA) Copernicus
program, there are several planned Sentinel missions in
the near future. Other initiatives include the Biomass
mission planned for 2021 and the 131 satellites that were
launched in 2015 by Planet Labs [6], where there will be
an open-access data model. All of these missions will require greater volumes of calibration and validation data.
The collection of ground truth for remotely sensed products has traditionally been undertaken by experts. However, with tightening budgets and an explosion in the volume
and frequency of data acquisition, new sources need to be
considered, particularly those from citizens. CS and crowdsourcing represent considerable opportunities to support
data collection for EO. At the same time, citizen involvement can promote EO more widely through raising awareness and education, which is often a secondary but fundamental goal of many CS projects.
In addition to vast quantities of data from space, CS and
crowdsourcing, i.e., the involvement of citizens in tasks such
as data collection, are also generating big data (especially
from social media, e.g., geotagged photographs), particularly in terms of the frequency and variety of data sources.
For this reason, crowdsourced data present numerous challenges, such as managing large volumes of data from diverse
inputs, how to ensure data quality, how to build up communities and motivate participants, and how to ensure the
sustainability of crowdsourcing activities. There is already a
growing body of literature on many of these issues from both
ecology/conservation [7]-[9] and the geographic literature
[10]-[13]. Overcoming these challenges will be critical if the
september 2016

ieee Geoscience and remote sensinG maGazine

data collected by citizens are to become a serious and rigorous input to support EO in the future. The aim of this article
is to highlight the potential of CS and crowdsourcing for the
calibration and validation of EO in the context of current and
future big data streams from space.
CROWDSOURCING, CITIZEN SCIENCE,
AND EARTH OBSERVATION
DEFINITIONS
CS can be defined as the involvement of the wider public in
scientific research from data collection to research design
[14], [15]. The term CS first appeared in a book of the same
name by Irwin in 1995, where Irwin expressed CS as the
idea of local knowledge to complement knowledge from
more scientific sources [54]. Around the same time, Rick
Bonney of the Cornell Laboratory of Ornithology used the
term as a synonym for public participation in scientific research (PPSR) [16].
PPSR and CS are only two terms of many that have appeared in the literature to describe the same basic phenomenon in which citizens have been involved in carrying out
some type of task. Another commonly used term is crowdsourcing, which was coined by Howe [17]. Combining the
words crowd and outsourcing, it literally means to outsource
tasks to the crowd. Also within Howe's definition is the idea
that this model has value for businesses and is therefore
often used in a more commercial sense. Crowdsourcing
platforms such as Amazon Mechanical Turk represent one
mechanism in which businesses can find low-cost labor to
carry out a range of microtasks, but it may also be an inexpensive source of data [18].
Volunteered geographic information (VGI) has emerged from the geographical literature [19] with a main focus
on citizens as sensors, gathering spatially referenced data
and providing it voluntarily. In VGI, end users contribute
geographic information to augment and replace existing
sources of information, such as printed maps, remotely
sensed images, and other web content. We also make the
distinction between active data collection, where participants go out and take measurements, and more passive
data collection from social media, where sensors are connected to the Internet and data are automatically collected,
e.g., data from amateur weather stations. A primary benefit
of VGI is that users are often more familiar with local geographic conditions and might contribute local geographic
information more often and faster than governmental
mapping organizations. End users may therefore be better
at detecting changes in their local environments.
One of the most famous examples of VGI is OpenStreetMap (OSM), a community mapping initiative to provide
open and free access to basic mapped features. Goodchild
[20] notes that georegistration errors between authoritative sources and nonauthoritative sources are often similar,
while a number of articles have shown reasonable positional accuracy between OSM and authoritative data [21],
39

Table of Contents for the Digital Edition of IEEE Geoscience and Remote Sensing Magazine - September 2016