Nicholas Gray Heavens 

Global climate models typically simulate climate at much larger spatial scales than tropical mountain glaciers and many of the world’s lakes. Yet some of the proxy data that can be used to validate models of past climate on land come from and are related to these geographic features. Validating global climate models using these proxies requires some method of downscaling global climate model output to the resolution of the relevant geographic feature. One possible downscaling technique is to use the global climate model to generate a time-evolving atmospheric state that can be used to force the model’s own land component at much higher resolution than the global model to resolve the relevant geographic features and/or local topography as well as the processes of interest. This dataset contains the output from 50 experiments that use the Community Earth System Model version 2 (CESM2) and its land component, the Community Land Model (version 5) to downscale climate simulations of pre-industrial, Last Glacial Maximum, and an extreme glacial climate with gross similarities to hypothetical late Paleozoic conditions (but within a Last Glacial Maximum paleogeography) to better resolve tropical mountain glaciers in various parts of the world and in one case, a lake. These experiments largely interpolate the land properties assumed by the relevant global climate model simulation to a 100 point per degree grid in a domain that is 1 in longitude and 2 in latitude. Global, high-resolution topography, and sometimes lake distribution datasets are then used to provide information about topography and other properties on the 100 ppd grid. The experiments are then run for 30 or 60 years. The dataset also contains global climate model simulation output of the model experiments used to generate the atmospheric forcing, simple analyses of tropical mountain glacier stability and extent in the experiments, and the code necessary to reproduce these experiments or apply them to other spatial domains.