The AllTaxa web interface allows exploration of species distribution modeling results from the AllTaxa Biodiversity Inventory (ATBI) of the Great Smoky Mountains National Park. It presents work done by the NSF Center for Remote Data Analysis and Visualization (RDAV) in collaboration with NIMBioS and Discover Life in America (DLIA).
Species distribution modeling involves making a computer model of where a species is likely to be present in a geographic area, given some observed locations and environmental information about the area. In our work in the Great Smoky Mountains National Park, the observed locations of species come from the ATBI, an extensive effort sponsored by DLIA to catalog every species in the Park. We also use ten different environmental layers (i.e. soil type, vegetation, elevation, etc.) provided by the Park at 30-meter resolution.
Usually, species modeling is performed at small scale, exploring a single species at a time. By using high-performance computing, we accelerate the process, producing models for over 500 species across the Park and allowing comparative analytics on the results. We used Nautilus, an SGI supercomputer, to perform thousands of runs of the MaxEnt Java code used for species modeling as well as pixel-by-pixel comparisons of the resulting models. Our AllTaxa web interface allows exploration of the modeling results and the relationships among species.
For more information, see:
Scott Simmerman, Jingyuan Wang, James Osborne, Kimberly Shook, Jian Huang, William Godsoe and Theodore Simons, "Exploring Similarities Among Many Species Distributions", Proc. of the Extreme Science and Enginnering Discovery Environment Conference (XSEDE12), pp. 38:1-38:9, Chicago, IL, 2012. pdf
Scott Simmerman, James Osborne and Jian Huang, "Eden: Simiplified Management of Atypical HPC Jobs", IEEE Computing in Science and Engineering (CiSE), ?(?), pp. ??-??, in press, 2012. url