Some components of uncertainty in the CCAMLR 2000 acoustical survey of krill

Quickly studied and quantified were sources of uncertainty in the CCAMLR 2000 acoustical estimate of krill abundance in FAO statistical area 48. The variance in system calibration was evaluated in relation to the effects of variant water temperature and salinity on sound speed, sound absorption, and beam characteristics. Uncertainty in krill target strength (TS) was estimated using a distorted wave born approximation model of krill TS, fitted with measured distributions of animal lengths and orientations. The uncertainty in species delineation was also investigated through the same scattering model. The combined measurement and sampling variance was estimated from a simulation which assumed the three frequency measurements provided independent estimates of the krill biomass. The overall variance (1 1.33%) was little different from the sampling variance (1 1.38%). That is, the measurement variance may be negligible relative to the sampling variance due to the large number of measurements averaged to derive the ultimate biomass estimate. Some potential sources of bias were also explored. The latter (ie. stemming from uncertainties in target strength, species delineation, bubble attenuation, thresholding, area definition, etc.), may be more appreciable components of measurement uncertainty. Further studies are warranted.