1st Place, Poster Presentation
Potential for Increased Detection Time with Multi-Static Active Sonar
By Nathan Faulkner, McCallum High School
Supervisor: Jerry Mitchell, Signal and Information Sciences Laboratory, ARL:UT

Abstract
There are two types of receiver processing for active sonar, monostatic receiver processing, where the vessel that transmits the signal is the only one that receives that signal, and bistatic receiver processing where one vessel transmits a signal and a different vessel receives the signal. Traditionally monostatic processing is used almost exclusively; however the goal of this project is to see if bistatic receiver processing can offer an increase in the detection opportunity of a target. When a signal is transmitted the target may be turned so that the angle where you would receive the highest target strength, like a sweet spot, is not where you transmitted. With monostatic processing you are only receiving the energy from one angle, so for one vessel there is only one chance of coming close to this sweet spot. In bistatic processing when a vessel is receiving the transmissions from every other vessel in the fleet, it is more likely that for at least one of these transmitters this vessel is near a sweet spot. A program was created to roughly simulate a few common Naval Scenarios in which vessels maneuver through the water while a target attempts to intercept them. The geometry from these Scenarios was then used to calculate all the monostatic and bistatic echo levels for each case at each point in time. With certain setups results showed that using bistatic processing did increase the detection opportunity of the target for the particular formations.
Return to top

2nd Place, Poster Presentation
Bubble Curtain Noise Abatement Experiment
By Kyle Ford, Anderson High School
Supervisor: Preston Wilson and Kevin Lee, Environmental Sciences Laboratory, ARL:UT

Abstract
Recently among industrial, governmental, and environmentalist groups there has been a growing perception of the impact that anthropogenic noise could possibly have on marine ecosystems. The purpose of this research is to design, build, and test a system that uses encapsulated bubbles to abate the noise produced by industrial underwater equipment (ex. offshore drilling or pile driving) in order to minimize potential stress on the environment. A prototype system consisting of a curtain of tethered, rubber-encapsulated bubbles with approximate resonance frequencies of 50 Hz was tested at the Lake Travis Test Station. A source barge equipped with an industrial shaker was placed inside the curtain to simulate the noise from the hull of a ship generated by rotating machinery. A five-element vertical hydrophone array was deployed at various depths and ranges from the source barge to record changes in radiated noise due to the curtain. The results were promising, showing that near 50 Hz, the noise was reduced by as much as 38 dB. Over a broader range of frequencies, a sound pressure level reduction of 10 dB was observed in the 30 Hz – 300 Hz frequency band. Removing the top 20% of the curtain had little effect on the noise reduction, and furthermore when the number of bubbles in the curtain was reduced by a factor of two, the amount of noise was decreased, demonstrating control over attenuation. Results indicate a system utilizing a curtain of encapsulated bubbles would be effective in reducing underwater manmade noise.
Return to top

3rd Place, Poster Presentation
Automated Detection of Outliers in Bathymetric Data
By Ben Laird, Pflugerville High School
Supervisor: David Munton and Lisa McFarland, Space and Geophysics Laboratory, ARL:UT

Abstract
Current measurement systems for ocean floor mapping, specifically swath sonar systems, are capable of producing millions of geospatial measurements that consist of a latitude, longitude, and depth value. These measurements inevitably contain outliers: points that do not fit the local and/or global trend of the data. Presently, time-consuming manual detection of these outliers is required and is accomplished by manual inspection of interactive maps. Implementing an automated system that performs a statistical analysis on the data to identify possible outliers would greatly increase the speed of processing this data. Due to the complexity of the topography of the ocean floor we cannot justify the use of traditional statistical methods for outlier detection that require assuming a normal distribution. In this project I developed a method of detection that successfully identifies outliers within test data sets. Due to the geospatial variation of the data and to improve computational efficiency the field of data is divided into geographically smaller regions and a median absolute deviation score, a robust statistical measurement which quantifies the likelihood that each point is an outlier, is calculated for each point. Each point with a score greater than a preset value is flagged as a potential outlier for further review. This method was tested on simulated data that allows us to measure the effectiveness of the procedure in different scenarios. Initial testing has provided promising results of outlier identification making this system a viable option to significantly aid in the development of ocean floor topography maps and databases.
Return to top