Topographic and vegetation structure controls on microclimate in complex landscapes: A case study from Mount Rainier National Park

Background/Question/Methods

Anticipating the impacts of climate change on species geographic distributions is limited by the mismatch in spatial scale between the types of climate data that are available (individual station data, coarse-scale gridded products) and the conditions that organisms actually experience. This scale mismatch can lead to a mischaracterization of the climate niches of species, and erroneous conclusions about population persistence under climate change. Moreover, characterizing climate at a spatial grain commensurate with other factors affecting the distributions of organisms, such as forest structure, may allow us to identify areas where climate suitability for species will be more or less sensitive to environmental change (e.g. climatic refugia). In this analysis, we used a hierarchical Bayesian state-space model to relate local climate from an extensive network of autonomous microclimate sensors located across Mount Rainier National Park (Washington, USA) to high resolution topographic and vegetation structure attributes derived from LiDAR.

Results/Conclusions

Our modeling framework allowed us to quantify seasonal variation in drivers of microclimate conditions while accounting for spatial and temporal autocorrelation of daily measurements. We found that incident solar radiation, topographic position, cold-air drainage, and canopy structure all jointly influence climatic conditions at small spatial scales, and those relationships vary predictably by season. Specifically, our results suggest that riparian areas in valley bottoms have lower (~1 °C) wintertime temperatures under stable synoptic conditions, but higher summertime high temperatures than ridges at the same elevation. Surprisingly, evidence for enhanced buffering of conditions in riparian areas relative to adjacent slopes is equivocal, although this part of the analysis suffers from low statistical power (n=14 sites). Higher-elevation sites were more strongly decoupled from regional climates than low-elevation sites, but were also more sensitive to regional temperature changes (e.g. a 1 °C increase in regional temperature corresponded with a >1 °C change at these sites). The diverse, stable and predictable relationships that we found between local and regional conditions suggest that different environments within this small (~5.5 km²) region may be subject to somewhat divergent climate trajectories, with consequences for climate-mediated range shifts of species and the disassembly and re-assembly of communities.