On sampling the ocean using underwater gliders

The sampling characteristics of an underwater glider are addressed through comparison with contemporaneous measurements from a ship survey using a towed vehicle. The comparison uses the underwater glider Spray and the towed vehicle SeaSoar north of Hawaii along 158 degrees W between 22.75 degrees N and 34.5 degrees N. A Spray dive from the surface to 1000 m and back took 5.6 h and covered 5.3 km, resulting in a horizontal speed of 0.26 m s(-1). SeaSoar undulated between the surface and 400 m, completing a cycle in 11 min while covering 2.6 km, for a speed of 3.9 m s(-1). Adjacent profiles of temperature and salinity are compared between the two platforms to prove that each is accurate. Spray and SeaSoar data are compared through sections, isopycnal spatial series, and wave number spectra. The relative slowness of the glider results in the projection of high-frequency oceanic variability, such as internal waves, onto spatial structure. The projection is caused by Doppler smearing because of finite speed and aliasing due to discrete sampling. The projected variability is apparent in properties measured on depth surfaces or in isopycnal depth. No projected variability is seen in observations of properties on constant density surfaces because internal waves are intrinsically filtered. Wave number spectra suggest that projected variability affects properties at constant depth at wavelengths shorter than 30 km. These results imply that isobaric quantities, like geostrophic shear, are valid at wavelengths longer than 30 km, while isopycnal quantities, like spice, may be analyzed for scales as small as a glider measures.