Delineating Sea-Level Rise Inundation: An Exploration of Data Structures and Performance Optimization

Abstract

Based on a conservative projection by the IPCC (IPCC 2007), inundation caused by sea level rise will likely disrupt the physical, economic, and social systems in coastal regions around the world. This research proposed an innovative method to calculate the minimum sea level rise required to inundate a cell in a Digital Elevation Model (DEM). The method, which accounts for water connectivity when determining inundation height for each cell, performs better than the simple “bathtub” approach, especially with sea level rises below 1 m. Several implementation data structures are proposed and compared. The combination of a binary heap and hash table data structure gives the most efficient implementation. The implementation is further parallelized using a master / worker paradigm. The parallel approach significantly outperforms serial implementations with respect to running time and memory footprint. Performance can be further improved with additional processing cores and using the supercomputing resources in the XSEDE (Towns, et al., 2014) program.