A number of authors, including me, have argued that the output of our most complex climate models, that is, of global climate models and Earth system models, should be assessed possibilistically. Worries about the viability of doing so have also been expressed. I examine the assessment of the output of relatively simple climate models in the context of discovery and point out that this assessment is of epistemic possibilities. At the same time, I show that the concept of epistemic possibility used in the relevant studies does not fit available analyses of this concept. Moreover, I provide an alternative analysis that does fit the studies and broad climate modelling practices as well as meshes with my existing view that climate model assessment should typically be of real possibilities. On my analysis, to assert that a proposition is epistemically possible is to assert that it is not known to be false and is consistent with at least approximate knowledge of the basic way things are. I, finally, consider some of the implications of my discussion for available possibilistic views of climate model assessment and for worries about such views. I conclude that my view helps to address worries about such assessment and permits using the full range of climate models in it.

包括我在内的许多作者认为，我们最复杂的气候模型（即全球气候模型和地球系统模型）的输出应该进行可能性评估。人们也表达了对这样做可行性的担忧。我在发现的背景下研究了对相对简单的气候模型的输出的评估，并指出这种评估具有认知可能性。同时，我表明相关研究中使用的认知可能性的概念并不适合对该概念的现有分析。此外，我提供了一种替代分析，它确实适合研究和广泛的气候建模实践，也符合我现有的观点，即气候模型评估通常应该具有真正的可能性。根据我的分析，断言一个命题在认识上是可能的，就是断言它不知道是错误的，并且至少与事物基本方式的近似知识一致。最后，我考虑了我的讨论对气候模型评估的现有可能性观点以及对此类观点的担忧的一些影响。我的结论是，我的观点有助于解决对此类评估的担忧，并允许在其中使用全方位的气候模型。