Abstract: Congwen Wang
Mentor: Micheal W. Sears
Species distribution models are often based on data with a spatial resolution that is much larger than the actual environments that individual organisms experience. Further, these models assume that all habitats within a spatial unit are homogeneous. Both of these assumptions are potentially problematic, yet these models are common tools for ecologists when making predictions of the impacts of climate change on the fate of populations and distributions of species. We have begun to construct more realistic models that incorporate landscape features into classical biophysical models of organisms. To continue this work, we will both refine the computational techniques as well as ground truth the model with empirical data. Computationally, we will acquire digital elevation models at the highest resolution possible. We will then sample these data at lower resolutions. All resulting data sets will then be analyzed through our biophysical model and statistical comparisons will be made to see how different important biological metrics differ across the different spatial resolutions. Further, we will compare the results to the operative temperatures of lizard models that we place in the same areas for which our digital elevation models correspond. We will both validate that our equations for estimating operative temperatures in the biophysical model make appropriate calculations and we will compare temperatures measured from the environment to those predicted at different spatial resolutions. Our hope is to identify the appropriate spatial scale that captures relevant thermal variability in real world environments to be used in computational models of species ranges.
Supported by the Bryn Mawr Summer Science Program