The total absorption coefficient of a natural water at a given wavelength can be determined from a measurement made within the water body itself. To make accurate measurements at those wavelengths where marine, and some inland, waters absorb only feebly, quite long pathlengths (~ 0.25 m or more) are needed. The simplest arrangement is to have, immersed in water, a light source giving a collimated beam and, at a suitable distance, a detector with a wide angle of acceptance. Most of the light scattered, but not absorbed, from the beam is detected: diminution of radiant flux between source and detector is thus mainly due to absorption and can be used to give a value for the absorption coefficient.
An alternative approach is to have, not a collimated beam, but a point source of light radiating equally in all directions.78,850 The detector does not have to have a particularly wide angle of acceptance. For every photon that is initially travelling towards the detector and is scattered away from the detector, there will, on average, be another photon, initially not travelling towards the detector, which is scattered towards the detector. Thus diminution of radiant flux between source and detector is entirely due to absorption. An absorption meter based on this principle has been used to measure the absorption coefficient of water in Lake Baikal (Russia).107 A problem with this and the previous type of absorption meter is that although scattering may not directly prevent photons reaching the detector, it increases the pathlength of the photons and thus increases the probability of their being absorbed, i.e. spuriously high values of absorption coefficient may be obtained. This error can be significant in waters with high ratios of scattering to absorption, but a correction can be made using the values of the scattering coefficient and average cosine of the scattering phase function if these are available for the water in question.719
To avoid the necessity of comparing readings in the water body with readings in pure water or other standard medium, a variant on the second type of absorption meter has been developed that still uses a point light source but has two detectors at different distances from the source: from the difference in radiant flux on the two detectors, the absorption coefficient may be obtained.414 By means of interference filters, this particular instrument carries out measurements of absorption coefficient at 50 nm intervals from 400 to 800 nm.
With all these absorption meters, if the light source is modulated and the detector is designed to measure only the modulated signal, then the contribution of ambient daylight to radiant flux on the detector can be eliminated.
In the case of an absorption meter using a collimated light beam, one possible solution to the problem of distinguishing between attenuation of the signal that is genuinely due to absorption from that which is due to photons in the measuring beam simply being scattered away from the detector, is to shine the beam down an internally reflecting tube235,1496 containing the water under investigation. The principle is that photons that are scattered to one side will be reflected back again, and so can still be detected at the other end. In such an instrument, however, there is still some residual loss of photons by non-reflection at the silvered surface, and by backscattering, and so a correction term proportional to the scattering coefficient needs to be applied.713 Current commercial instruments6 have dispensed with the silvered surface and rely on the internal Fresnel reflection (100% for scattering angles up to 41 °) of photons that takes place in water within a glass cylinder surrounded by air.1497 The light is collected by a diffused large area detector at the far end of the flow tube. These instruments can be used for vertical profiling of absorption coefficients.163
Gray et al. (2006) have described an in-principle design of a cylindrical flow-through absorption meter, which makes use of the integrating cavity principle. Musser et al. (2009) have constructed a prototype instrument, based on this design. The sample fluid flows through a 122 cm long quartz tube, internal diameter 2.54 cm, surrounded by two cylindrical diffuse reflectors, 102 cm long, separated by an air gap. Light from light emitting diodes (LEDs) operating at six wavelengths is introduced into the air gap, leading to the establishment of a diffuse light field within the sample cylinder. The device has an operating range from 0.004m-1 to 80m-1, and like other instruments based on the ICAM principle, is essentially unaffected by high levels of scattering within the sample.
Absorption meters and other optical instruments are sometimes attached to moorings at various depths to obtain long-term time series of optical data for periods up to several months. Unfortunately, instruments submerged for long periods will - unless preventative measures are taken - soon have their measurements compromised by fouling -the settlement and growth of marine organisms on their surfaces.
Davis et al. (1997) arranged for a concentrated solution of bromine to diffuse into the flow tube between sampling periods, and in this way were able to prevent biofouling for 3.5 months at 11 m and 6 months at 40 m, in the Bering Sea.
Absorption coefficients of natural waters can also be determined in situ from measurements of irradiance and scalar irradiance. The instrumentation for such measurements we shall discuss in a later chapter (see §5.1). Determination of absorption coefficient makes use of the relation (see §1.7)
Eo where E0 is the scalar irradiance, E is the net downward irradiance (Ed - Eu) and KE is the vertical attenuation coefficient for E. Thus a can be obtained from measurements of net downward and scalar irradiance at two, or a series of, depths (measurements at more than one depth are required to give KE). Absorption meters based on these principles have been built.570,1283 By making use of empirical relationships derived from numerical modelling of underwater light fields, estimates of absorption (and scattering) coefficients can be derived from in situ irradiance,714,145 or irradiance and nadir radiance,1307,885,890 measurements.
Was this article helpful?