Aerial surveys are routinely used to study marine mammal populations. The resolution of imagery from commercial satellites has improved to the point where individual marine mammals can be detected. It may therefore be possible to perform remote sensing of marine mammal populations from space. This paper presents an initial assessment of detectability. A simple signal and noise model is developed to predict the signal-to-noise ratio (SNR) for a whale-like target. SNR is calculated for three cases: detection limited by sensor quantization noise ("best case"); detection in real noise, using one spectral band image; and detection in real noise, using a two-band noise reduction technique. An Ikonos satellite image was used for a realistic noise spectrum. The calculations show that a canonical target of length =14 in and average spectral reflectivity can be detected up to a depth of 24 m. Thus a case can be made for using satellites to monitor marine mammal abundance and geographical distributions, and to observe migratory patterns in remote areas.
[1]
T. Lee,et al.
Shallow sea-floor reflectance and water depth derived by unmixing multispectral imagery
,
1992
.
[2]
L. D. Stuffle.
Bathymetry from hyperspectral imagery.
,
1996
.
[3]
David L.B. Jupp,et al.
Background and Extensions to Depth of Penetration (DOP) Mapping in Shallow Coastal Waters
,
1988
.
[4]
H. Urkowitz.
Signal Theory and Random Processes
,
1983
.
[5]
E. Gumbel,et al.
Statistics of extremes
,
1960
.
[6]
James V. Carretta,et al.
Distribution and abundance of marine mammals at San Clemente Island and surrounding offshore waters : results from aerial and ground surveys in 1998 and 1999
,
2000
.