Collaborative proposal: The political and Security Impacts of Abrupt Climate Change on Modern Populations: An Integrated Computational Model

自20世纪70年代以来，气候科学家开发了越来越复杂的世界气候系统模型。这些模型正在对未来几十年可能发生的气候变化进行预测。这种变化肯定会引起人类的反应，许多评论家已经提出了世界某些地区可能的反应。然而，人口是一个复杂的系统，其中系统一部分的变化可能对系统的其他部分产生重大影响，这意味着人类对气候变化的反应也必须作为一个复杂系统中的过程来研究，该系统可能至少涉及所有人类社会。然而，到目前为止，社会科学还没有建立大规模人口系统的模型，这些模型整合了气候模型的预测。造成这种忽视的一个原因可能是，这类工作不仅需要了解人类行为的基本原则及其建模技术，而且还需要了解（1）当前的气候模型;（2）地理信息系统，它提供了一种组织全球地理数据，包括人口数据的手段;（3）用于分析从人口模拟产生的大量数据的统计技术;以及（4）人类对最近的气候变化的实际反应（例如，达尔富尔）和更遥远的过去（例如，古玛雅）。因此，本项目包括两名具有行为建模背景的政治科学家;一名在气候建模方面有广泛工作的气候学家;一名地理信息系统专家;一名应用数学和统计学专家;以及一名考古学家。行为模型的初步步骤将根据最近（例如达尔富尔）和更遥远的过去（例如，玛雅崩溃）。在这种情况下，如果开发中的模型的预测与已知的人类反应不匹配，则将调整其参数，并随后针对进一步的此类情况进行测试。考古证据表明，气候变化在过去破坏了许多文明，但气候变化本身的破坏性往往小于人类对这些变化的反应，包括大规模移民和武装冲突。在相互依存的现代世界中，人类对气候变化的反应有可能涟漪到直接受影响的地区之外，而在不同的、也许是相当偏远的地区对这种变化的反应可能会以消极的方式相互作用。因此，关于人类对预期气候变化-包括突然变化-可能作出的反应的完善模型将使决策者不仅能够预测当地气候变化的当地后果，而且能够预测地理上可能相当遥远的后果。该项目朝着人类对气候变化反应的更全面发展的模型迈出了第一步，就像早期的气候模型现在可以被视为朝着目前存在的更全面发展的气候模型迈出的第一步一样。

Since the 1970s, climate scientists have developed increasingly sophisticated models of the world climate system. These models are converging on predictions about climate changes that are likely to occur in coming decades. Such changes are certain to provoke human responses and many commentators have suggested what those responses are likely to be in particular parts of the world. Nevertheless, human populations are complex systems in which changes in one part of the system can have significant consequences for other, perhaps quite remote, parts of the system - meaning that human responses to climate change must also be studied as processes in a complex system involving, potentially at least, all human societies. To date, however, the social sciences have not produced models of large scale human population systems that integrate predictions from climate models. One reason for this neglect may be that such work requires knowledge not only of the basic principles of human behavior and the techniques for modeling it, but also of (1) current climate models; (2) geographic information systems (GIS) which provide a means for organizing geographic data worldwide, including human population data; (3) statistical techniques for analyzing the large amounts of data produced from population simulations; and (4) actual human responses to climate changes both in the recent past (e.g., Darfur) and in the more remote past (e.g., ancient Maya). Accordingly, the present project includes two political scientists with background in behavioral modeling; a climatologist who has worked extensively on climate modeling; an expert in geographic information systems; an expert in applied mathematics and statistics; and an archaeologist. Initial steps toward the behavioral model will be tested against known cases of climate change in the recent past (e.g. Darfur) and the more remote past (e.g., the Mayan collapse). To the extent predictions from the developing model do not match known human responses in such cases, its parameters will be adjusted - and, subsequently, tested against further such cases. Archaeological evidence shows that climate change has undermined many civilizations in the past but also that the climate changes per se were often less damaging than the human responses to those changes - including mass migration and armed conflict. In the interdependent modern world, human responses to climate changes have the potential to ripple well beyond an immediately affected area, and responses to such changes in different and perhaps quite remote areas might interact in negative ways. Thus, well developed models of likely human responses to expected climate changes - including abrupt changes - will allow policy makers to anticipate not only local consequences of local climate changes, but also geographically perhaps quite remote consequences. The project takes initial steps toward more fully developed models of human responses to climate change, just as early climate models can now be seen as initial steps toward the more fully developed climate models that presently exist.