In a previous editorial I encouraged more work in physical geography related to zoonoses such as COVID-19 (Malanson, 2020). Here, I want to raise a caution. As physical geographers and other environmental scientists, we learn early that correlation is not causation. In addition to simply spurious correlations, the more pernicious problems of multicollinearity, serial and spatial autocorrelation, spatial and temporal nonstationarity, and mismatched scales are well-known problems that can also arise in studies of the physical geography of COVID-19. While many studies are sound, enough are questionable to prompt this editorial. Potential problems could include the following:

中文翻译：

---

COVID-19 和自然地理 – 黄牌

在之前的一篇社论中，我鼓励在与 COVID-19 等人畜共患病相关的自然地理领域开展更多工作（Malanson，2020 年）。在这里，我想提个醒。作为自然地理学家和其他环境科学家，我们很早就了解到相关性不是因果关系。除了简单的虚假相关之外，多重共线性、序列和空间自相关、空间和时间的非平稳性以及尺度不匹配等更有害的问题也是众所周知的问题，这些问题也可能出现在 COVID-19 的自然地理研究中。虽然许多研究都是合理的，但足以引发这篇社论的质疑。潜在的问题可能包括以下内容：