FeaturesDialogues: climate care

XXIX.1 January - February 2022
Page: 44
Digital Citation

Reimagining environmental data

Robert Soden

back to top 

When we think about climate change, so often we think with and through data. Statistics such as parts per million of atmospheric CO2, thresholds of global average temperature, and annual tons of greenhouse gas emissions form our understanding of potential futures. Numbers ranging from meters of sea-level rise, to counts of so-called climate migrants forced to leave their home cities or countries thanks to heat and drought, to the year-over-year records being set by tropical storm damages shape our understanding of climate change's impacts. Even efforts to communicate about climate science through narrative or art often rely heavily on models and forecasts of possible futures to give the work a realistic underpinning. These models, and the understandings they yield, are powered by global information infrastructures cobbled together out of satellites, tide gauges, data standards, epistemic cultures, and legal mandates that specify how climate data should be collected across jurisdictions.

back to top  Insights

Environmental data and models shape our understanding of the world around us, our place in it, and our relations with one another.
HCI has a wealth of methods and knowledge that can inform how we collect, analyze, and use data in ways that support justice and sustainability.

Environmental data tells us so much about the world around us and our place within it. The maps, models, and dashboards that we use to understand the environment influence how environmental problems, such as climate change and its many impacts, are understood and delimit the realm of possible solutions. They have a powerful impact on how we come into relationship with nature, and how these relationships evolve and are sustained over time. In doing so, they shape both who we are and what we mean to one another. In science and technology studies, scholars refer to these relationships as "imaginaries," or collective, place-based, and culturally and institutionally reinforced practices and beliefs that help structure our connections to the world. Such imaginaries are intimately connected with the information infrastructures we use to apprehend this world [1]. The everyday politics of care will thus be shaped by both our imaginaries and the infrastructures we act through and with.

Despite the promise of these tools, all data is partial and limited. This argument echoes commonly heard phrases such as "the map is not the territory" or "there is no such thing as raw data." All models of the world are, by necessity, incomplete. It is their very incompleteness that allows them to scale and gives them their power. What is too often forgotten is that when abstracting certain aspects of the environment through data, we are always making choices about what and what not to include. In making these choices, we have a tendency to reproduce, sometimes unknowingly, our own priorities, prejudices, and perspectives as well as those of the wider society in which we live. When we attend to the choices we make when deciding how to gather, analyze, and use environmental data, we create the opportunity for things to be otherwise.

If we want to attend to our relations with the world and with one another with care, we need to be much more careful about the environmental data we collect, how we collect it, and how we use it. With that in mind, I want to suggest six ways that we can begin to reimagine environmental data.

  • We should lend attention and support to efforts that aim to challenge or unsettle predominant understandings of environmental issues. As has been widely documented in the field of critical cartography, the history of environmental information, from the origins of mapping to contemporary satellites, GPS, and geographic information systems, is bound up with the worldviews and priorities of military and empire. Despite recent efforts to democratize or decolonize such tools, traces of these priorities persist, such as commonly used map projections that exaggerate the size of North America and Europe as compared with other parts of the world. Efforts to undermine these legacies, such as David Garcia's work (https://www.patreon.com/mapmakerdavid), which combines cutting-edge digital cartography with subversive representational choices—using Indigenous place names, centering maps on oceanic rather than land features, removing national boundaries—can help reveal the subtle and not-so-subtle ways that environmental data reflects the interests of the powerful and suggest alternatives.

In a similar vein, environmental justice groups have been deploying participatory sensing methods for decades as a means to challenge industry accounts of air or water quality in racialized or low-income communities. In addition to creating an evidence base to hold polluters accountable, participatory data-collection practices also can serve movement building by drawing members of affected frontline communities into activist networks. For example, the West Oakland Environmental Indicators Project (WOEIP; https://woeip.org/) takes Oakland, California, residents on "toxic tours" through polluted areas, pointing out offending sites and recording air-quality measurements using backpack-size sensors along the way. According to WOEIP organizers, the data itself is often the least important part of the tours, as compared with the role that the tours play in volunteer recruitment and in raising awareness of air quality issues in the region. HCI scholars Amanda Meng and Carl DiSalvo argue that "counter-data action" such as WOEIP's toxic tours are important community empowerment tools for their contributions to "resource mobilization, relationship building, and development of critical conscious[ness]" [2].

  • We should expand who we consider to be users of environmental data. Too often, the only audiences considered are experts or policymakers. As recent crisis informatics research on disaster risk communication has demonstrated, the public has distinctly different and varied needs [3]; information products created for one purpose may lead to faulty decision making when used for another. Max Liboiron's study of disaster data created following Hurricane Sandy in 2012 showed that official statistics tended to reflect the interests and worldviews of the government, neglecting the priorities expressed by communities affected by the storm [4]. Such neglect has material consequences, shaping what forms of disaster assistance are provided and to whom. Other research has shown that even official disaster information, such as situation reports used by humanitarian agencies, are typically not designed with the specific needs of their users in mind [5]. HCI research, by recognizing the variety of users that exist for environmental data, can better target their needs.

  • Work on environmental data should account for the labor, energy, and materials that go into its creation and maintenance. Fantastical stories about how AI will revolutionize environmental monitoring are channeling billions of dollars in funding to Silicon Valley tech companies, masking the environmental impacts and the often difficult and poorly compensated work necessary for their proprietary algorithms to function. This is occurring while in other parts of the U.S., many of the flood maps that inform the National Flood Insurance Program as well as planning decisions and public understanding of risk are decades out of date, in part due to lack of funding. Cindy Lin's ethnographic research with Indonesia's national mapping agency documents not only the deeply situated nature of the labor required to produce forest data but also the ways in which emerging data science methods are changing both how this labor is performed and what counts as a forest [6]. We cannot come to grips with, or meaningfully intervene in, the information infrastructures that shape our understanding of the environment without attending to the work and work settings that operate them.

  • We need to develop new models of data ownership, governance, and sharing. In this vein, Maggie Jack and Seyram Avle argue for a feminist geopolitics of technology focused on "place, everyday surviving and thriving, and community" [7]. These values, when applied to the social, legal, and economic frameworks that currently govern environmental data, would suggest that quite significant changes are necessary. The work of Digital Democracy provides ideas on what this could look like. Their collaboration with the Waorani people in the Ecuadorian Amazon, for example, leverages digital mapping tools to develop community-created and-owned maps of millions of acres in territory in support of Indigenous land claims against fossil fuel companies (https://www.digital-democracy.org/ourwork/waorani/; https://waoresist.amazonfrontlines.org/). Their open-source tool kit and the community of earth defenders that uses it go beyond prevailing models of open data and participatory mapping to develop locally specific and cocreated solutions that give control over environmental data to the communities who live in the areas the data describes.

  • Building on the broader turn to the arts in HCI, we should be looking to collaborations with the arts as sources of inspiration and critique. The San Francisco—based Climate Music Project (https://climatemusic.org/) uses climate data—atmospheric CO2 concentrations, average temperatures, etc.—to adjust the volume, tempo, and pitch of original music scores over the course of a performance. Their live shows include string instruments, drums, and keyboards and original video projected on a screen behind the musicians. The performances take audiences through a timeline of human history that begins prior to agriculture and traverses several potential climate futures, based on Intergovernmental Panel on Climate Change projections. As temperatures rise and climate impacts increase, melodies that began as soft and harmonious grow discordant and chaotic, stressing the urgency of the problem and the necessity to respond. Other work seeks to help climate data achieve greater resonance by placing it in the local context it describes, powerfully indicating which communities will be affected and what is at risk. For example, Eve Mosher's High Water Line (https://highwaterline.org/) takes sea-level-rise projections for several coastal areas around the U.S. and, using chalk lines on sidewalks and pavements in each city, delineates areas predicted to be underwater.

In both the Climate Change Music Project and High Water Line, art is used as a means of communicating science in more compelling ways. In contrast, other projects seek to make more of an upstream contribution to how science is conducted through developing critical perspectives on environmental data. The Disaster and Climate Change Artathon, for example, assembled 30 scientists and artists for a two-day workshop on developing speculative art projects that upend traditional approaches to environmental data. The themes, including affect, irony, and political economy, suggest novel and compelling directions for future research in environmental informatics [8]. As currently practiced, however, speculative practice in this space is not without its limits. A recent review of speculative design research within the sustainable HCI research area found that such projects on the whole are largely focused on the Global North and ignore distributional equity [9]. These findings support other calls for reconsidering the role of speculative design practice in questions of sustainability, climate, and HCI more broadly [10].

  • Finally, HCI scholars are increasingly turning to history as a source of insight and inspiration. While design tends to orient toward the future, calls for greater attention to history go back decades and are gaining momentum in the field [11]. Paul Edwards's book on the development of the global climate-modeling system spans nearly two centuries and sheds light on how design choices and modeling approaches adopted far in the past help shape the contemporary politics of climate change [12]. Ongoing work by my colleague David Lallemant into the history of commonly used earthquake-modeling software reveals how surprising connections between data standards, international development priorities, and changing norms in the field of civil engineering shape the way much of the world perceives earthquake risk. Uncovering the lineage of contemporary tools and approaches helps us understand them further, revealing their contingencies as well as offering hints at alternatives.

The photograph Earthrise (Figure 1) was taken from lunar orbit by astronaut William Anders on December 24, 1968, during the Apollo 8 mission. Credited with giving inspiration to the U.S. environmental movement and called by some the most influential environmental photograph ever taken, Earthrise was shot just two weeks after Douglas Engelbart's presentation at the ACM/IEEE Computer Society Fall Joint Computer Conference in San Francisco. Engelbart's talk, which featured technologies including the mouse, hypertext, and collaborative real-time editing, presented a far-reaching vision of what personal computing could be. It was later called the "Mother of All Demos" and is one of the foundational moments of HCI. These moments, as powerful as they were, are part of a much longer history of the way in which technology intervenes in our understanding of ourselves and the world around us. We have yet to come to terms with this and are still developing the necessary cultural resources to do so, both from a theoretical perspective and in how we design and inhabit our everyday lives. A critical element in efforts to do so will be to reimagine the environmental data and information technologies that shape our relationships with nature, ourselves, and one another.

ins01.gif Figure 1. Earthrise.

back to top  References

1. Soden, R. and Kauffman, N. Infrastructuring the imaginary: How sea-level rise comes to matter in the San Francisco Bay Area. Proc. of the 2019 CHI Conference on Human Factors in Computing Systems. ACM, New York, 2019, 1–11.

2. Meng, A. and DiSalvo, C. Grassroots resource mobilization through counter-data action. Big Data & Society 5, 2 (2018).

3. Bica, M., Weinberg, J., and Palen, L. Achieving accuracy through ambiguity: The interactivity of risk communication in severe weather events. Computer Supported Cooperative Work (CSCW) 29, 5 (2020), 587–623.

4. Liboiron, M. Disaster data, data activism: Grassroots responses to representing Superstorm Sandy. In Extreme Weather and Global Media. J. Leyda and D. Negra, eds. Routledge, 2015, 144–162.

5. Finn, M. and Oreglia, E. A fundamentally confused document: Situation reports and the work of producing humanitarian information. Proc. of the 19th ACM Conference on Computer-Supported Cooperative Work and Social Computing. ACM, New York, 2016, 1349–1362.

6. Lin, C. How to make a forest. E-Flux. Apr. 2020; https://www.e-flux.com/architecture/at-the-border/325757/how-to-make-a-forest/

7. Jack, M. and Avle, S. A feminist geopolitics of technology. Global Perspectives 2, 1 (2021).

8. Soden, R., Hamel, P., Lallemant, D., and Pierce, J. The Disaster and Climate Change Artathon: Staging art/science collaborations in crisis informatics. Proc. of the 2020 ACM Designing Interactive Systems Conference. ACM, New York, 2020, 1273–1286.

9. Soden, R., Pathak, P., and Dogget, O. What we speculate about when we speculate about sustainable HCI. Proc. of the ACM Conference on Computing and Sustainable Societies. ACM, New York, 2021.

10. Kozubaev, S., Elsden, C., Howell, N., Søndergaard, M.L.J., Merrill, N., Schulte, B., and Wong, R.Y. Expanding modes of reflection in design futuring. Proc. of the 2020 CHI Conference on Human Factors in Computing Systems. ACM, New York, 2020, 1–15.

11. Soden, R., Ribes, D., Avle, S., and Sutherland, W. Time for historicism in CSCW: An invitation. Proc. of the ACM Conference on Computer Supported Cooperative Work & Social Computing. ACM, New York, 2021.

12. Edwards, P.N. A Vast Machine: Computer Models, Climate Data, and the Politics of Global Warming. MIT Press, 2010.

back to top  Author

Robert Soden is an assistant professor in computer science and in the School of the Environment at the University of Toronto. His research draws on design, social sciences, and the humanities to evaluate and improve the information systems used to respond to environmental challenges like disasters and climate change. soden@cs.toronto.edu

back to top 

©2022 ACM  1072-5520/22/01  $15.00

Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee.

The Digital Library is published by the Association for Computing Machinery. Copyright © 2022 ACM, Inc.

Post Comment

No Comments Found