Comparison of land-based vs. floating calibration targets used in aerial photogrammetric measurements of whale lengths

摘要

Measurements from vertical aerial photographs have been applied to a variety of cetacean species, e.g., bowhead (Balaena mysticetus), gray (Eschrichtius robustus), blue (Balaenoptera musculus), right whales (Eubalaena australis), and dolphins (Delphinidae spp.) to document many life-history parameters including: whale length, a proxy for age; growth rate; nutritive condition; and reproductive status (Best and Rüther 1992, Koski et al. 1992, Angliss et al. 1995, Gilpatrick 1996, Perryman and Lynn 2002, Cosens and Blouw 2003, Gilpatrick and Perryman 2008). The many variables that complicate measurements made from aerial photography demand fairly precise corrections for altitude, lens characteristics, aircraft motion, and potential problems between camera, altimeter, and data entry systems (Angliss et al. 1995). These variables can be documented and corrected through precise calibration tests conducted at least once each field season. Most aerial photogrammetric studies of marine mammals have used land-based calibration targets to scale measurements from photographs to actual whale measurements (e.g., Best and Rüther 1992, Cosens and Blouw 2003, Koski et al. 2006) because it has been logistically difficult to set up known-length calibration targets on water during brief periods when the sea is relatively calm. A vessel is required to tow targets far enough from shore to avoid having shore features confuse the radar altimeter signal. However, Angliss et al. (1995) described floating fish pens used by the National Marine Fisheries Service (NMFS) to calibrate aerial photogrammetric measurements, and others (e.g., Gilpatrick 1996, Perryman and Lynn 2002, Gilpatrick and Perryman 2008) also used floating calibration targets. Perryman and Lynn (2002) assumed that radar altimeters calibrated over floating targets would be more accurate to correct whale measurements than would land-based calibrations. However, they did not actually substantiate this assumption by conducting comparisons of altimeter performance over land relative to over the sea surface. Gilpatrick (1996) found a positive bias in radar altimeter values over water in that the radar altimeter overestimated distance, but this was not compared to land-based targets. Other studies have attempted to make these comparisons, but the data are sparse and results inconclusive (e.g., Nerini et al. 1987, Withrow and Angliss 1992), and no land vs. sea calibration results have been published. Accordingly, this note describes the first experiment to compare land-based vs. floating targets photographed from the same aircraft on the same day and provides a critical test in establishing the accuracy of whale measurements made via aerial photogrammetry.