We consider here instruments that measure radiance only along a linear path, i.e. that do not simultaneously scan from side to side.
For completeness we shall deal first here with radiometric sensing of water bodies carried out from boats, even though this is, of course, strictly speaking not 'remote' sensing. As a quantitative measure of ocean colour, Jerlov (1974) defined the colour index to be the ratio of nadir radiance (i.e. vertically upwards flux, nadir angle 0„ = 0°) within the water in the blue at 450 nm to that in the green at 520 nm, and developed a colour meter consisting of a pair of downward pointing Gershun tubes with a blue and a green interference filter, respectively, to measure it. The meter was designed to be lowered manually from shipboard to any depth. Neuymin et al. (1982) chose a somewhat different pair of wavelengths in the blue and green for the colour index, namely 440 and 550 nm, which has the advantage that this ratio is particularly sensitive to algal pigments, and consequently this is the pair of wavelengths most favoured for remote sensing of oceanic phytoplankton (see later). As it happens, Neuymin et al. chose to define their colour index in terms of the green to blue ratio (in the upwelling flux), rather than vice versa, and so their index increases with phytoplankton content. Their instrument was located at 5 to 6 m depth in a shaft running through the vessel, and could provide a continuous record of the colour index of the sea along the ship's track.
Bukata, Jerome and Bruton (1988) developed a shipboard radiometer mounted on a 4.5 m boom from the bow, at 4.5 m above the water surface. Nadir radiance from the lake water surface was measured in four broad wavebands in the blue, green, red and near-infrared regions, together with downward solar irradiance in the same four wavebands. Measurements were taken continually from the moving vessel at half-second intervals.
Deschamps et al. (2004) have developed a hand-held radiometer (SIMBAD) for shipboard use, which measures the emergent flux from the ocean in five spectral bands - 443, 490, 560, 670 and 870 nm - with 10 nm bandwidths. To reduce sunglint and skylight-reflection effects, radiance is measured through a vertical polarizer, and the instrument is held at 45° from the nadir (near the Brewster angle) and 135° from the principal plane of the Sun. The field of view is 3°. The instrument incorporates a sun photometer to measure aerosol optical thickness, from which the downward solar irradiance in each waveband can be estimated, and the values of remote-sensing reflectance can thus be calculated. The major advantage of this, and other, above-water systems is that the ship does not need to stop for measurements to be taken.
Wood and Cunningham (2001) constructed an above-surface, ship-borne reflectance radiometer, which by means of a grating and a CCD array carried out near-simultaneous measurements of water-leaving radiance (radiometer acceptance angle, 1°) and downward irradiance over the range 380 to 800 nm. With the instrument mounted high on the ship superstructure, at an angle of 53° to the vertical (the Brewster angle), a point on the surface at a distance of about 30 m from the sunnier side of the vessel was viewed through a polarizing filter. The authors point out that the movement of the ship, and the presence of waves on the surface, prevent the Brewster angle being achieved exactly, and some reflected light from the sky will therefore pass through the polarizing filter. Even after applying a correction for this problem, there remained a significant discrepancy between the Lu spectrum obtained with the shipborne instrument, and the true underwater Lu spectrum. However, if the reflectance spectra (Lu[l]/Ed[lj) were normalized to reflectance at 490 nm, there was good agreement between them in the 475 to 750 nm range.
The Satlantic Micro Surface Acquisition System (MicroSAS) is designed for above-water measurements of ocean colour, and is suitable for mounting on ships or on low-flying aircraft. It has a radiance meter pointing at the ocean to measure the water-leaving radiance, another pointed at the sky to provide the information necessary to correct for reflected skylight, and an irradiance sensor pointing upward to measure the downward solar flux so that remote-sensing reflectance can be calculated. In a typical configuration it operates simultaneously in seven wavebands corresponding to those that are used in the spaceborne Sea-WiFS scanner (see later).
In an example of innovative use of inexpensive, off-the-shelf instrumentation, Goddijn and White (2006) used the red-green-blue output from a digital camera to characterize water colour in Galway Bay (Ireland). To eliminate reflection from the water surface a tube was fitted around the camera lens to penetrate the water surface, the end of the tube being about 10 cm below the surface. The data were used to study the distribution of CDOM (see later).
The usual aim in ocean remote sensing is to obtain information from a large area of the sea within a short space of time: in effect, a 'snapshot' of the area of interest as it was at one point in time. An alternative approach is to concentrate on one small location of interest and use remote sensing to follow changes with time. For this purpose towers can be used, but for practical reasons have only so far been constructed in shallow coastal waters. The Aqua Alta Oceanographic Tower, built in 1970, is located in the Adriatic Sea, about 15 km SE of the Venice Lagoon, in a water depth of 17 m. The optics platform is 7 m above the sea surface. Hooker et al. (2004) carried out above-water radiance measurements in five wavebands in the blue-green - 412, 443, 490, 510 and 555 nm. To avoid sunglint the radiometer was pointed 90° or 135° away from the plane of the Sun. To remove the contribution of reflected skylight they measured downwelling sky radiance at the appropriate angle, in the same plane, and calculated reflected sky radiance assuming Fresnel reflection at the surface.
Zibordi et al. (2006) used the Gustaf Dalen Lighthouse Tower, located in the northern Baltic, five nautical miles off the Swedish coast in a water depth of 16 m, from which to monitor water-leaving radiance at regular intervals from June through October 2005. The instrument used was a SeaPRISM radiometer with seven channels (412, 440, 500, 555, 675, 870 and 1020 nm), with 1.5 ° full angle field of view, placed on top of the tower, 25.5 m above average sea level. The sea-viewing angle was 40°, and the relative azimuth with respect to the Sun was 90°. With this instrument, which operated automatically, they were able to monitor the development and subsequent decay of a cyanobacterial bloom during July.
An advantage of above-water monitoring of the submarine light field is that the radiometers are not subject to the problem of biofouling, which bedevils long-term measurement within the water. From towers, such as described above, only a small area of the sea surface can be observed. For long-term, above-water optical monitoring of important medium-scale aquatic ecosystems, such as certain estuaries, the placement of spatially scanning radiometers in tethered balloons might be worth considering.
Gitelson and Kondratyev (1991) sought to combine both 'remote' and sea level sensing by flying 10 to 15 m above the sea surface in a helicopter from which they simultaneously made radiometric measurements and took water samples. Upward and downward radiance and irradiance, in nine l0-nm spectral channels between 430 and 750 nm were measured with a hand-held spectrometer requiring less than 1 s to complete its readings.
Where a spectroradiometer is being used from an aircraft it is important that the spectral scan is completed quickly to minimize the effects of changes in the emergent flux from point to point along the flight path. Clarke et al. (1970) used a spectroradiometer with a 3° x 0.5° field of view, which scanned from 400 to 700 nm in 12 s. The instrument was operated at the Brewster angle in conjunction with a polarizer, to eliminate reflected skylight. Neville and Gower (1977) used a spectroradiometer1426 in which the spectrum produced by a diffraction grating was distributed over an array of 256 silicon diodes, each of which continuously detected the radiant flux in its own narrow region of the spectrum. It was possible to read out the complete spectrum, 380 to 1065 nm, at 2 s intervals.137 This instrument was also operated at the Brewster angle.
At the other extreme, spectrally speaking, useful information can be obtained by airborne remote sensing in two or three wavebands, provided these are carefully chosen. Arvesen, Millard and Weaver (1973) developed a differential radiometer, which carried out simultaneous measurements of upward radiance at 443 and 525 nm, and continuously compared one with the other (in effect, measuring Jerlov's colour index). This had the advantage that changes in incident light intensity or variations in surface water roughness had similar effects on the flux in both wavebands and so were automatically corrected. Changes in phytoplankton concentration, however, which would be expected specifically to affect the ratio of upward flux in these two wavebands (see §7.5), markedly affected the signal.
To address the need for an affordable aircraft ocean colour instrument, the National Aeronautics and Space Administration (NASA) and the National Oceanic and Atmospheric Administration (NOAA) developed the Ocean Data Acquisition System (ODAS).526,1448 This measures nadir radiance in three l5-nm bands centred on 460,490 and 520 nm, these bands being chosen specifically to make possible the use of a previously developed199 spectral curvature algorithm for remote sensing of phyto-plankton. The three radiances are sampled ten times per second and averaged over 1s. With the aircraft flying at 50ms"1 at an altitude of
— 150 m, the footprint of the instrument is about 5 x 50 m. Position is measured accurately at 20 s intervals with Loran-C navigation.
Was this article helpful?