Finding cause–effect relationships behind observed phenomena remains a challenge in spatial analysis. In recent years, much progress in causal inference has been made in statistics, economics, epidemiology and computer sciences, but limited progress has been made in spatial statistics due to the nonrandom, nonrepeatability and synchronism of spatial data. In this paper, we investigate the problem. We first refine the issues of causal inference, then discuss the causal inference issue in spatial statistics, next review the causal inference methods in other disciplines and analyze their potential to be used with cross-sectional data, and finally we look forward prospect of causal inference in spatial statistics.

寻找观察到的现象背后的因果关系仍然是空间分析的挑战。近年来，因果推断在统计学、经济学、流行病学和计算机科学方面取得了很大进展，但由于空间数据的非随机性、不可重复性和同步性，空间统计方面的进展有限。在本文中，我们调查了这个问题。我们首先细化因果推理的问题，然后讨论空间统计中的因果推理问题，然后回顾其他学科的因果推理方法并分析它们与横截面数据一起使用的潜力，最后展望因果推理的前景在空间统计中。