Bathymetric mapping with QuickBird data

Abstract

Two algorithms are used to determine bathymetry in the littoral region using QuickBird multispectral satellite observations. The algorithms determine water-leaving radiance and convert this to water depth values. The first algorithm uses a ratio of two wavebands and the second uses the sum of several wavebands. Relative bathymetric errors are determined for the clear water of Looe Key (USA) and the turbid water of Plymouth Sound (UK). Bathymetric measurements from LIDAR and chart data are compared to derived depths to assess their accuracies. An amended version of the ratio method is proposed for use in turbid water to improve accuracy. The results show that the standard ratio and turbidity algorithms have a relative error of 11.7% and 16.5% respectively in clear water. In turbid water the average error of the turbidity algorithm is 11.6% and the amended ratio algorithm average error is 13%.