A Vast Machine

A Vast Machine: Computer Models, Climate Data, and the Politics of Global Warming

Paul N. Edwards
Series: Infrastructures
Copyright Date: 2010
Published by: MIT Press
Pages: 552
https://www.jstor.org/stable/j.ctt5hhds1
  • Cite this Item
  • Book Info
    A Vast Machine
    Book Description:

    Global warming skeptics often fall back on the argument that the scientific case for global warming is all model predictions, nothing but simulation; they warn us that we need to wait for real data, "sound science." In A Vast Machine Paul Edwards has news for these skeptics: without models, there are no data. Today, no collection of signals or observations -- even from satellites, which can "see" the whole planet with a single instrument -- becomes global in time and space without passing through a series of data models. Everything we know about the world's climate we know through models. Edwards offers an engaging and innovative history of how scientists learned to understand the atmosphere -- to measure it, trace its past, and model its future.

    eISBN: 978-0-262-29410-2
    Subjects: General Science, Technology

Table of Contents

  1. Front Matter
    (pp. i-vi)
  2. Table of Contents
    (pp. vii-viii)
  3. Acknowledgments
    (pp. ix-xii)
  4. Introduction
    (pp. xiii-xxviii)

    Unless you have been in a coma since 1988, you have certainly heard or read a story that goes something like this: Global warming is a myth. It’s all model predictions, nothing but simulations. Before you believe it, wait for real data. “The climate-studies people always tend to overestimate their models,” the physicist Freeman Dyson told an interviewer in April 2009. “They forget they are only models.”¹ In the countless political controversies over climate change, the debate often shakes out into a contest: models versus data.

    This supposed contest is at best an illusion, at worst a deliberate deception—becausewithout...

  5. 1 Thinking Globally
    (pp. 1-26)

    In 1968, three American astronauts became the first human beings ever to see Earth’s full disk from space. President Lyndon B. Johnson mailed framed copies of the Apollo mission’s photographs to the leaders of every nation as an allegory of the inevitable unity that encompasses all human division and diversity and binds us to the natural world.

    By then, of course, representations of Earth as a globe were already centuries old. Nevertheless, many saw a transfiguring power in the awesome beauty of those famous photographs. That small blue ball, spinning alone in darkness: it hit you like a thunderclap, a...

  6. 2 Global Space, Universal Time: Seeing the Planetary Atmosphere
    (pp. 27-48)

    Today we see world maps almost everywhere we go. Backdrops to the nightly news, they appear transparent, obvious, unmediated. We seem to grasp their God’s-eye view intuitively, without thought. GPS receivers in our phones and our cars pinpoint us precisely on the global grid. In all their incarnations, from Mercator projections to parlor globes to interactive GPS, maps are information technologies of the first order. They are “objects to think with,” in Sherry Turkle’s felicitous phrase.¹

    Behind the seeming immediacy of global maps and images lie vast bodies of complex and expensive collective and collaborative work and social learning accomplished...

  7. 3 Standards and Networks: International Meteorology and the Réseau Mondial
    (pp. 49-60)

    Universal time was only one among many standards sought and achieved by scientists, engineers, commercial enterprises, and governments during the latter half of the nineteenth century. Indeed, standardization itself should be seen as a major characteristic of this historical period. For one thing, the period witnessed the first widespread manufacture of interchangeable parts, the most important innovation of the industrial age. Yet a great deal of standardization occurred in the realms of organization, technique, and practice, rather than in the realm of technologyper se. Industrial mass production depended not only on machine tools capable of the precision necessary to...

  8. 4 Climatology and Climate Change before World War II
    (pp. 61-82)

    As we saw in chapter 2, the basic structure of the global circulation was well established by the middle of the nineteenth century. So were the fundamental forces driving that motion. Yet the causal relationship between the circulation and the climate remained poorly understood. Even as late as World War II, meteorologists could still say little about this relationship with any certainty.

    There were two main reasons for this. First, until the 1930s virtually all weather and climate data had been collected at the surface. Few direct measurements existed that might be used to chart the details of circulatory structures...

  9. 5 Friction
    (pp. 83-110)

    You toss your laptop into a backpack, tuck your Blackberry into a shirt pocket, email your colleague a spreadsheet from the back seat of a taxicab. Once-inconceivable computer power is now ubiquitous. That power drives a colossal, networked information infrastructure through which pass terabytes of data and communication (not to mention vast amounts of “spam”) each day. When one is caught up in the cascade of words, images, and numbers, in the frenetic traffic from screen to screen, it is easy to lose sight of the infrastructure—to forget that, underneath that glistening surface of free-flowing information, computing remains a material...

  10. 6 Numerical Weather Prediction
    (pp. 111-138)

    Between 1945 and 1965, digital computers revolutionized weather forecasting, transforming an intuitive art into the first computational science. Unlike many scientific revolutions, this one was planned. Numerical weather prediction became the civilian showcase for a machine invented in wartime to support specifically military needs. Scientists conceived and carried out the first experiments with numerical forecasting in the earliest days of electronic computing, years before commercial computers became widely available, as a joint project of American military research agencies and the US Weather Bureau. A principal architect of that project was John von Neumann, who saw parallels between the science of...

  11. 7 The Infinite Forecast
    (pp. 139-186)

    The concentration of computing resources at a few institutions probably affected no field more than it affected climatology. In the 1960s and the 1970s, this data-driven, regionally oriented, descriptive field would be transformed into a theory-driven, globally oriented discipline increasingly focused on forecasting the future. The change would be wrought not by traditional climatologists, but by scientists based in theoretical meteorology and computer programming, working at a handful of institutions endowed with what were then enormous computing resources. Unlike traditional climatologists, who searched for regularities in data from the past, this new generation of scientists sought to simulate the climate,...

  12. 8 Making Global Data
    (pp. 187-228)

    As the era of numerical forecasting dawned, atmospheric scientists began to realize that the structure of their discipline had been turned on its head by the computer. In the 1940s the stream of data had already become a flood; both forecasters and climatologists were collecting far more information than either could ever hope to use with the technologies then at their disposal. Yet by the late 1950s, as NWP models reached hemispheric scales and general circulation modeling began, forecasters and climatologists saw that they would soon not have nearlyenoughdata—at least not in the right formats (standardized and...

  13. 9 The First WWW
    (pp. 229-250)

    Ruskin dreamed of a “perfect system of methodical and simultaneous observations.” Richardson imagined a “forecast-factory.” Those were visions of a centrally planned, perfectly uniform observing and forecasting system whose every part and every person works according to a single set of standards and techniques. For the myriad of reasons discussed in earlier chapters, to build such a system in the real world you would need godlike authority and unlimited resources, and even then you would probably fail. Instead, infrastructural globalism succeeded in meteorology because it adopted a particular development strategy—a strategy partly inherited, partly chosen, and partly forced upon...

  14. 10 Making Data Global
    (pp. 251-286)

    Preceding chapters examined the rise of a global weather and climate data internetwork, from the kluged pre–World War II system to the systematic planning of the World Weather Watch for coordinated global observing, communication, and data processing. This is what I have been callingmaking global data: collecting planetary data in standard forms, through interconnected networks, to build data images of global weather and general circulation. Yet difficult as it was, collecting planetary data was only one dimension of the overall effort. This chapter explores the complementary project ofmaking data global: building complete, coherent, and consistent global data...

  15. 11 Data Wars
    (pp. 287-322)

    Climate is essentially the history of weather, averaged over time. So one might be forgiven for assuming that climatedataare simply weather data, averaged over time. After all, we are talking about the same primary variables (temperature, pressure, wind, etc.), measured by the same instruments at the same places, usually by the same people, transmitted through the same communication systems, under standards set by agencies of the same organizations.

    Yet in the course of writing this book I was regularly met with blank stares, puzzlement, and hostile comments. “You’re talking aboutweatherdata, notclimatedata,” some interlocutors said,...

  16. 12 Reanalysis: The Do-Over
    (pp. 323-336)

    From the earliest national and global networks through the 1980s, every empirical study of global climate derived from the separate stream of “climate data.” Climatological stations calculated their own averages, maxima, minima, and other figures. Central collectors later inverted the climate data infrastructure, scanning for both isolated and systematic errors and working out ways to adjust for them, seeking to “homogenize” the record. All of these efforts presumed—for the very good reasons discussed above—that only traditional “climate data” could form the basis of that record.

    But as numerical weather prediction skill advanced and computer power grew, a new...

  17. 13 Parametrics and the Limits of Knowledge
    (pp. 337-356)

    In chapters 8–12, I explored how scientists use models to process, interpret, and reconstruct observations, tomake data global. Models are how we know what we know about the history of climate. They are also why we can keep learning more about the history of climate, without ever securing a single definitive version of that history once and for all. The past shimmers. What about the future?

    If we need models to make sense of data about the past, we need them all the more to predict whether, and how, the climate will change. Chapter 7 described how climate...

  18. 14 Simulation Models and Atmospheric Politics, 1960–1992
    (pp. 357-396)

    Between 1960 and 1980, as World Weather Watch systems came online and the Global Atmospheric Research Program built toward the crescendo of the Global Weather Experiment, concerns about anthropogenic greenhouse warming slowly began to travel outward from a small scientific elite into policy circles.

    Policymakers care much less about the past than about the future. The overriding questions for them take the form “What will happen if . . . ?” For most policymakers or policy institutions even to notice any given issue among the thousands that confront them daily generally requires at least three things: a crisis, a constituency...

  19. 15 Signal and Noise: Consensus, Controversy, and Climate Change
    (pp. 397-430)

    If you make a widget and you want people all over the world to use it, you need to do three things. First, you need to make the widget work—not just where you are, under your local conditions, but everywhere else too. It should be bulletproof, stable, and reliable. It should also be simple, accessible, and cheap, so that most people can afford it and will be capable of using it. Second, you need to distribute it around the world. Making it isn’t enough; you have to get it out there, everywhere, to everyone. Finally, your widget needs to...

  20. Conclusion
    (pp. 431-440)

    If engineers are sociologists, as Michel Callon and Bruno Latour have taught us, then climate scientists are historians.¹ Their work is never done. Their discipline compels every generation of climate scientists to revisit the same data, the same events—digging through the archives to ferret out new evidence, correct some previous interpretation, or find some new way to deduce the story behind the numbers. Just as with human history, we will never get a single, unshakeable narrative of the global climate’s past. Instead we get versions of the atmosphere, a shimmering mass of proliferating data images, convergent yet never identical.²...

  21. Notes
    (pp. 441-508)
  22. Index
    (pp. 509-518)