Understanding the impact of along-transect resolution on acoustic surveys

Abstract

Resolution can be defined as the minimum distance between two consecutive sampling points taken by an instrument. In acoustic surveys, the main parameter determining the resolution of sampling along a transect is the distance between successive echosounder transmissions or “pings”. An increase in either the time interval between pings or the speed of the vessel increases the inter-ping distance, hence decreasing the effective sampling resolution. This study investigated whether a loss in along-transect resolution affects the mean backscattered acoustic energy, leading to uncertainty and/or bias in abundance estimates. To this end, a real acoustic survey was echo-integrated, followed by the application of a systematic resampling scheme to simulate a decrease in pinging resolution. For each transect, the mean backscattered acoustic energy calculated at each resolution was compared with that at the original resolution. Transects were characterised according to their heterogeneity and spatial autocorrelation to investigate their effect on the relationship between abundance error and sampling resolution. Uncertainty was seen to increase with decreasing resolution, with higher heterogeneity and lower spatial autocorrelation accelerating the rise in imprecision. Although the mean bias across replicates was zero, the asymmetry of the bias distributions increased with decreasing resolution, leading to an increasing probability and magnitude of underestimation (https://aztigps.shinyapps.io/PingRateStudio/).