Temperature changes at the Earth surface propagate into the subsurface and leave a thermal signature in the underlying soil and rock. Inversions of subsurface temperature measurements yield reconstructions of ground surface temperature (GST) histories that provide estimates of climatic changes. A question remaining in the interpretation of reconstructed GST histories is the extent to which GST changes reflect changes principally in surface air temperature (SAT), or whether other factors may be significant. Here we use a Land Surface Processes (LSP) model to examine the influence of precipitation changes on GST and subsurface temperature and moisture fields on annual to decadal timescales. We model soil and vegetation conditions representative of a prairie region in the southern Great Plains of North America and force the model with meteorological data synthesized from a typical year in the region. Model responses are observed after changes in the amount of daily precipitation, the intensity and frequency of daily precipitation, and the diurnal and seasonal timing of precipitation. We show that: (1) increasing daily precipitation cools mean annual GST, (2) increasing the intensity and reducing the frequency of daily precipitation, while holding the annual amount of precipitation constant, cools mean annual GST, and (3) shifting maximum precipitation to occur in the warmest months cools mean annual GST. We compare modeled results to observed precipitation changes during the 20th century and conclude that the observed precipitation changes would cause only small changes to GST within the modeled region, on the order of 0.1 K or less.