Water

Pure water, although it appears colourless in the quantities we handle in everyday life, is in fact a blue liquid: the blue colour is clearly apparent under sunny conditions in oceanic waters or in coastal waters that are infertile and have little input from rivers. The colour of pure water arises from the fact that it absorbs only very weakly in the blue and the green regions of the spectrum, but its absorption begins to rise as wavelength increases above 550 nm and is quite significant in the red region: a one-metre thick layer of pure water will absorb about 35% of incident light of wavelength 680 nm.

Because of the very weak absorption it is very difficult to measure the absorption coefficient of pure water in the blue/green spectral region, and the values reported in the literature, determined by normal spectrophotometric procedures, vary widely. I rely here predominantly on the measurements of Pope and Fry (1997) using their integrating cavity absorption meter, supplemented with those obtained by Quickenden and Irvin (1980) and Litjens et al. (1999) using long-pathlength cells.

The extent to which pure water absorbs light in the ultraviolet (UV), including the ecologically relevant UVB (280-320 nm) region, is a matter of some controversy. Liquid water does have an intense absorption band at 147 nm due to an electronic transition, but this, on theoretical grounds, would be expected to tail away to very low values in the 200 to 300 nm range. An anticipated exponential diminution of absorptivity with decreasing photon energy (the Urbach rule) implies, for example, an absorption coefficient of only ~ 0.02 m_1 at 207 nm, and still lower values at higher wavelengths.1096 While some workers over the period 1928 to 1976 found absorption coefficients (200-300 nm) that appeared to be orders of magnitude higher than this, more recent studies by Quickenden and Irvin (1980), and Boivin et al. (1986) using very carefully purified water, indicate that absorption is indeed very low in the near UV. The previously reported high absorption coefficients are attributable to dissolved oxygen, which does absorb UV quite strongly (and which is, of course, present in sea and lake water anyway), and to trace organic materials.

Literature values for the absorption coefficient of pure water over the range 280 to 800 nm are listed in Table 3.1, and Fig. 3.3 shows the absorption spectrum of pure water from the near ultraviolet, through the photosynthetic range, to the near infrared. It can be seen that the absorption by water at the red end of the visible spectrum is really the tail end of a series of much stronger absorption bands in the infrared: there are even more intense absorption bands beyond 1000 nm. The two shoulders in the visible spectrum - the distinct one at ~604nm and the weak one at ~5l4nm - have been identified as corresponding to the fifth and

300 400 500 600 700 800 900 1000 Wavelength (nm)

Fig. 3.3 Absorption spectrum of pure water. Absorption coefficient values have been taken from Table 3.1 for the range 310 to 790nm, and from the data of Palmer to Williams (1974) for the range 790 to 1000 nm.

300 400 500 600 700 800 900 1000 Wavelength (nm)

Fig. 3.3 Absorption spectrum of pure water. Absorption coefficient values have been taken from Table 3.1 for the range 310 to 790nm, and from the data of Palmer to Williams (1974) for the range 790 to 1000 nm.

sixth harmonics, respectively, of the O-H stretch vibration of liquid water;1341 the peaks at ~960 and ~745nm in the infrared correspond to the third and fourth harmonics (the fundamental is at about 3 mm). Light absorption by pure water shows no significant temperature dependency at wavelengths below 550 nm. but does increase slightly with temperature at longer wavelengths.1324,1038 A plot of CT (da/dT, m"1 "C"1) shows small peaks centred at wavelengths of ~604 and 662 nm, and a large peak at ~740nm. The peaks in CT correspond to shoulders in the absorption spectrum of pure water. The CT maximum at 740 nm has a peak value of -0.015 m"1 °C"1, corresponding to an absorption coefficient change of 0.3 m"1 over a 20°C change in water temperature. Sullivan et al. (2006) found pure NaCl at a concentration of 10% (w/w) to have little effect on water absorption below -600 nm, a small positive peak between 600 and 650 nm, a negative peak between 680 and 730 nm, and a marked increase in absorption between 735 and 750 nm. They attribute these changes to effects of the salt ions on the molecular vibrations of the water molecules rather than to intrinsic absorption by NaCl itself. The implications for sea water (salinity 36) are that the additional absorption induced by salt in the 400 to 680 nm spectral range is very small, with a maximum magnitude of <0.002 m"1.

Table 3.1 Absorption coefficients for pure water: 280 to 320 nm, Quickenden and Irvin (1980); 366 nm, Boivin et al. (1986); 380 to 720 nm, Pope and Fry (1997); 730 to 800 nm, Smith and Baker (1981).

I (nm)

a (m-1)

I (nm)

a (m-1)

280

0.0239ab

560

0.0619

290

0.0140ab

570

0.0695

300

0.0085ab

580

0.0896

310

0.0082ab

590

0.1351

320

0.0077ab

600

0.2224

366

0.0055a

610

0.2644

380

0.0114

620

0.2755

390

0.0085

630

0.2916

400

0.0066

640

0.3108

410

0.0047

650

0.340

420

0.0045

660

0.410

430

0.0050

670

0.439

440

0.0064

680

0.465

450

0.0092

690

0.516

460

0.0098

700

0.624

470

0.0106

710

0.827

480

0.0127

720

1.231

490

0.0150

730

1.799

500

0.0204

740

2.38

510

0.0325

750

2.47

520

0.0409

760

2.55

530

0.0434

770

2.51

540

0.0474

780

2.36

550

0.0565

790

2.16

800

2.07

a Absorption coefficients derived from the published attenuation coefficients by subtracting estimated values632'832 of scattering coefficients. b These values were obtained using deoxygenated water.

The contribution of water itself to the attenuation of PAR by absorption of quanta is of importance only above about 500 nm. While the salts present in sea water appear to have no significant effect on absorption in the visible/photosynthetic range,1113,1244 the nitrates and bromides do cause a marked increase in absorption below 250 nm.161,1006,1011

Understanding the optical properties of water in its frozen, solid forms - ice and snow - has taken on greater importance as ice at high altitudes and latitudes (glaciers, Arctic Ocean, the Greenland and Antarctic ice sheets) melts at increasing rates with the advent of global warming. This information is essential for an accurate assessment of the overall energy balance of the planet, but is much harder to acquire than for liquid water. The first measurements of absorption by ice were carried out by Sauberer (1950) on blocks of ice cut from a lake in Austria. He found that absorption was high, as it is in liquid water, in the near-infrared and red regions but fell markedly with decreasing wavelength, arriving at a minimum at about 400 nm. Snow itself is so highly scattering that the light field becomes diffuse within a few centimetres of the surface, and clean, fine-grained snow reflects 97 to 99% of incident light in the 300 to 600 nm range, although reflectivity falls off, due to increasing absorption, as wavelength increases into the red and near-infrared.1440 On the basis of their measurements within clean Antarctic snow, Warren et al. (2006) conclude that the absorption coefficient of pure ice has a minimum at ~390nm, and is everywhere below 0.1 m"1 between 300 and 600 nm.

0 0

Post a comment