Dissertation Research: Dividing the Waters: Data, Models, and Environmental Policy in the American Southwest and U.S.-Mexico Border Region

这篇科学和技术研究论文改进补助金探索了大规模计算机模拟在美国西南部和美墨边境地区环境治理制度中的变化。自20世纪90年代初以来，模型设计(转向基于软件和面向对象的战略)和规模(长期增加资金并扩大到新的地理位置)的变化，将计算机模拟推向了更接近环境管理、政策和争议的核心。这产生了相互矛盾的影响：一方面，模型在区域水资源管理的实践和政治中享有新的日益突出的地位；另一方面，它们的成功在规划者、政策制定者和各种环境公众中引发了关于其基本性质和认识地位的广泛辩论：模型产生了什么样的知识，我们如何相信这种知识？如何在模型设计和使用的层面上表达、管理和判定不确定性、可信度和信任的关系？在调解公众环境争议方面，模特可以做什么，也应该做什么？这项研究通过其更广泛的制度和社会背景追踪一种特殊的环境事实制作艺术，探索在频繁强烈的地区水政治聚光灯下运作的模型建造者、用户和公众的基础设施工作。研究方法是比较的和人种学的，利用了三个主要案例的原始实地调查：加州水资源部和美国垦荒局开展的CALSIM II过程，最初旨在模拟州水资源和中央山谷项目的相互作用，但越来越多地被用作全州范围内系统的一般预测模型；RiverWare，由美国垦荒局和田纳西州山谷管理局开发的建模环境，用于科罗拉多州、田纳西州和其他管理严重的流域的决策支持；和边界+20，这是一个由国会资助的西南环境研究和政策中心赞助的基于团体的建模练习，试图建立可信的20年期货，作为美墨边境两国环境政策和合作的基础。资金将主要用于支持在加利福尼亚州、科罗拉多州和墨西哥北部对模型开发人员和用户进行采访、人种学观察和档案研究的旅行。在科学、技术和科学研究领域内，这项研究将有助于对分布式和计算机数据系统在地球科学和公共政策的共同建设中的作用进行新的研究。除了与STS学者相关外，这项研究还将谈到地球科学家、政策制定者和环境公众面临的紧迫关切，因为新的数据和信息技术以及新的实践模式被部署到整个环境领域。最具体地说，这项研究将提出关于在生态、制度和政治背景下适当设计和使用大规模计算机模拟的经验教训。

This Science and Technology Studies Dissertation Improvement Grant explores the changing place of large-scale computer simulations within regimes of environmental governance in the American Southwest and U.S.-Mexico border region. Since the early 1990s, changes in model design (a shift towards software-based and object-oriented strategies) and scale (a secular increase in funding and expansion to new geographic locales) have pushed computer simulations closer to the heart of environmental management, policy and controversy. This has produced contradictory effects: on one hand, models enjoy a new and growing prominence in the practice and politics of regional water management; on the other, their very success has provoked widespread debate among planners, policymakers and various environmental publics over their fundamental nature and epistemic status: What sort of knowledge do models produce, and how are we to credit this knowledge? How are relations of uncertainty, credibility and trust expressed, managed and adjudicated at the level of model design and use? What work can and should models do in mediating public environmental controversies? This study traces a particular art of environmental fact-making through its wider institutional and social context, exploring the infrastructural work of model builders, users and publics operating under the frequently intense spotlight of regional water politics. The study methodology is comparative and ethnographic, drawing on original fieldwork into three primary cases: the CALSIM II process undertaken by the California Department of Water Resources and the U.S. Bureau of Reclamation, originally designed to model the interaction of the State Water and Central Valley Projects, but increasingly being deployed as a general predictive model for the state-wide system as a whole; RiverWare, a modeling environment developed by the U.S. Bureau of Reclamation and the Tennessee Valley Authority for decision support on the Colorado, Tennessee and other heavily-managed basins; and Border+20, a group-based modeling exercise sponsored by the Congressionally-funded Southwest Center for Environmental Research and Policy that attempts to establish credible 20-year futures as a basis for binational environmental policy and cooperation along the U.S.-Mexico border. Funds will be used primarily to support travel to conduct interviews, ethnographic observation and archival research with model developers and users in California, Colorado, and Northern Mexico. Within the field of STS, the study will contribute to emerging research on the role of distributed and computer-based data systems in the mutual construction of the earth sciences and public policy. Beyond its relevance for STS scholars, the study will speak to pressing concerns faced by earth scientists, policy makers and environmental publics as new data and information technologies and new modes of practice are deployed throughout the environmental field. Most concretely, the study will suggest empirically-grounded lessons concerning the appropriate design and use of large-scale computer simulations in their ecological, institutional, and political contexts.