2023_programme: Detection performance uncertainty due to imprecise environmental knowledge

Back


  • Session: 24. Sonar Technology
    Organiser(s): N/A
  • Lecture: Detection performance uncertainty due to imprecise environmental knowledge [invited]
    Paper ID: 1928
    Author(s): Brooker Daniel, Fialkowski Laurie
    Presenter: Brooker Daniel
    Abstract: This work explores the uncertainty in detection metric outputs that are due to imprecise environmental knowledge. Detection performance is modeled for a single hydrophone using a detector based on the signal-to-noise ratio at a single acoustic receiver in a range-independent environment. The receiver is fixed in shallow water near a shipping lane. The signal of interest is a narrow-band submerged target at a constant depth, with a variable point of closest approach to the detector. All of the acoustic propagation is modeled using a parabolic equation model; the noise due to surface waves is simulated as a sheet of surface sources, with the source level based on wind speed; noise due to surface ships is modeled by assuming each ship is a point source at a depth of 6m, with a source level prescribed by the ship type, length, and speed. The wind speed data and shipping data from the Ambient Sound Modeling Workshop in July 2022 are used as inputs. Uncertainty in the cumulative probability of detection is generated via a distribution of sound speed profiles, emulating the uncertainty present when forecasted or historical conditions are used to simulate real world environments. The sound speed profile is typical of shallow water in summer conditions for temperate climates, where warm surface ducts are present above the thermocline. The perturbations to the sound speed profile consist of changes in the depth and width of the surface duct, which are difficult to predict precisely, and vary temporally. The uncertainty in detection performance as a function of acoustic receiver depth is explored. Additionally, other sources of uncertainty are considered, including wind speed and ship source levels. This works demonstrates the importance of including uncertainty metrics when using model outputs to make policy decisions. [Work Sponsored by the Office of Naval Research]
  • Corresponding author: Dr Daniel Brooker
    Affiliation: US Navy Research Lab
    Country: United States
    e-mail:

Back