Discipline: Computer Sciences and Information Management
Subcategory: Computer Science & Information Systems
Nubia E. Morton - North Carolina Central University
It is well established that the water pressure created under the impoundment of large dams can trigger earthquakes. This is known as reservoir-induced seismicity (RIS) . In this project, we employ data fusion to investigate how the different attributes of dams including height, surface area, and volume are likely to trigger extra earthquakes in areas that are already prone to earthquakes. We fuse data from three different datasets: earthquakes (13 features), dams (22 features), faults (21 features), with minimal overlap of features. Data fusion is the process of combining different datasets to produce more consistent and useful information than any of the original datasets provided individually. Each dataset was preprocessed by adopting standard data cleaning techniques; additionally, feature engineering techniques, such as binning (latitude, longitude), variable creation and transformation, were employed to provide consistent data over the contiguous US. We visualized the data using R. By plotting the points for the earthquakes and dams on the map of the US based on the given longitudes and latitudes, we observe that the pattern for dams exceeding a certain height was similar to that of earthquakes exceeding a certain depth and magnitude; this is not the case for the dams? surface area or volume. We mapped the dams with different variables exceeding a certain threshold and the earthquakes that occurred within a certain radius of those dams, and we observed that the most clusters of earthquakes formed around the specified dams when based on the height rather than surface area or volume. Analysis of the visuals found that the heights of the dams are strongly correlated with the frequency and intensity of earthquakes. Western US had the highest average height of dams but the lowest average surface area of dams. It is well documented that earthquakes are the strongest and most frequent in the western US largely due to tectonic stresses in that region. Further investigation will be carried out to determine quantitatively the impact of the presence of reservoirs on frequency and magnitude of earthquakes. In addition to the heights of the dams, other characteristics of the dams will be combined using Dempster’s rule of combination (DRC). References:  Gupta, H. K. 2002. A Review of Recent Studies of Triggered Earthquakes by Artificial Water Reservoirs with Special Emphasis on Earthquakes in Koyna, India. Earth-Science Reviews, Elsevier, 58.  Talwani, P. 1997. On the Nature of Reservoir-Induced Seismicity, Pure and Applied Geophysics, 150.
Funder Acknowledgement(s): This work is supported by the National Science Foundation under Grant No. HRD-1238547.
Faculty Advisor: Alade Tokuta, email@example.com
Role: Implemented code for the project, created visuals, did exploratory analysis of the data, and applied data fusion.