Optical classifications of the seas in relation to phytoplankton ...

Optical classifications o c e a n o l o g i a , N o . 3 1n 1 1 · 1 . · PP· 25—55, 1991.

01 the seas m relation PL i s s n 0078-3234

to phytoplankton Optical and bio-optical1 . . j. classifications of

characteristics natural watersDiffuse attenuation

coefficients for irradiance Distributions of chlorophyll a

concentration in the sea

B o g d a n W o źn iak Institute of Oceanology,Polish Academy of Sciences,

^ % > p o t . N** t ’ovioU-kjuCo* O<łjl0"Ł»SiJfW* 5 5Va d im N . P e le v in

Shirshov Institute of Oceanology AS USSR,Moscow

<&*· >23Manuscript received November 29, 1990, in final form December 31, 1991.

A b stra ct

The paper compares and anlyzes the optical and bio-optical classifications of na­tural marine basins most often quoted in the literature, as well as an attem pt to estimate their accuracy. The authors present two original classifications, worked out from statistical analyses of experimental downward irradiance attenuation spec­tra. The phytoplankton effect and the influence of other optical components on this downward irradiance attenuation is also discussed. A quantitative description of phytoplankton resources in the basin as related to selected components of the light attenuation coefficients in the water column is given.

1. In tr o d u c t io n

T he optical properties of m arine basins vary in a wide range depending on the concentration of individual optically active com ponents and on their original optical characteristics (D era, 1983; Jerlov, 1968, 1976). Neverfeatu- res in common. A num ber of au thors have a ttem p ted to classify the optical properties of w ater masses independently of their depth or general optical properties (i.e. tak ing in to account their changes w ith depth) e.g. Baker

* The investigations were carried out as part of the research programme CPBP 03.10, co-ordinated by the Institute of Oceanology of the Polish Academy of Sciences.

and Sm ith (1982), Jerlov (1951, 1961, 1968, 1976, 1977, 1978), J 0rgensen and Des M arais (1988), M orel (1978, 1982), Pelevin (1985), Pelevin and Rutkovskaya (1977, 1978, 1980), P rieur and S a thyendranath (1981), Rut- kovskaya et al. (1982), Siegel and Dickey (1987), Simonot and Le T ruet (1986), Sm ith and Baker (1978, 1984). The aim of th is paper is to present the existing classifications, to com pare them , and to assess their accuracy. The presen tation has been lim ited to the m ost im p o rtan t and m ost popular classifications quoted in the lite ra tu re two original classifications are also discussed.

Classifications are usually developed w ith respect to the m ajo r apparen t optical properties of the sea, th a t is, the spectra of the diffuse a ttenuation coefficient of the downward irradiance versus dep th Kd{ A, z) - described by the equation (after Preisendorfer (1961))

( i)whereE d ( \ , z ) - downward irradiance by solar rad iation of light wavelength

A a t depth z.The diffuse a ttenuation coefficients K d ( A, z ) are affected on the one hand

by the external conditions of the sun’s rays pene trating th e sea (angular he­ight of the sun, optical properties of the atm osphere and cloud cover, wave m otion a t the sea surface, etc.) , and on the o ther by the inherent optical absorption-diffusion properties of the m arine environm ent - the m ajo r factor influencing the propagation of.light in to the basin. T he influence of ex ternal factors on the K d (A, z ) coefficients are not usually taken in to account. These classifications are therefore less precise as regards the optical properties of surface w ater th an in relation to the deeper w ater layers, where the effect of the irradiance beam struc tu re is not so noticeable (D era, 1983; Jerlov, 1976). The general assum ption, then , is th a t the K d ( ^ , z ) coefficients de­pend m ainly on the content and properties of the individual optically active com ponents of sea w ater. It is common knowledge th a t the m ajo rity of these com ponents (except sea w ater and inorganic suspensions) ensue from direct or indirect biochem ical processes in the phy top lank ton (see V inogradov and Shushkina (1987)). In consequence, a num ber of s ta tis tica l relationships exist betw een phytop lank ton content and the resu ltan t optical properties of the basin (M orel, 1978, 1982; Sm ith and Baker, 1978). Solving this problem constitu tes the second m ajor aim of the paper, i.e. quan tita tive ly charac­terizing the phytop lank ton effect on light a tten u a tio n and the relationship betw een the phy top lank ton content of the w ater colum n and the resu ltan t K d ( A) coefficients. It is usual to express the phy top lank ton concentration in sea w ater in term s of the m ajo r photosynthetic p igm ent, chlorophyll a. This

aim was realized by statistically analyzing the experim ental d a ta collected a t the In stitu te of Oceanology of the PAS. The analysis form ed a basis for W ozniak’s original bio-optical classification of m arine basins, m ore precise and universal th an the previous ones.

It is w orth m entioning th a t the problems considered in this paper, re­lated to the optical classification of m arine basins and their relationship w ith selected phytoplankton characteristics are closely connected w ith the current needs of biology and ecology. For example, m any research team s are concentrating their efforts on developing rem ote optical sensing techni­ques of m easuring prim ary production in various parts of the W orld Ocean from satellites (Pelevin et al., 1991). This task requires a careful analysis of the relationship betw een the m ajor biomass producer - phy top lank ton - and the optical properties of the basin directly influencing prim ary produc­tion or determ ining tlio struc tu re of the radiance beam issuing from the sea and recorded by satellite scanners (M orel and B erthon, 1989; Sm ith et al., 1989a,b).

2. R e v ie w o f o p tic a l c la ss if ic a tio n s o f n a tu ra l w a tersan d m a r in e b a sin s

2.1. J er lo v ’s optical classification o f m arine basins

Jerlov presented his first classification of the apparen t optical p roper­ties of various seas as early as 1951; the final version appeared in 1978 (Jerlov, 1951, 1968, 1976, 1977, 1978). He analyzed the experim ental spectra of the downward irradiance transm ission coefficients in sea w a­ter Tjvf(A,z) = E d ( ^ , z ) / E d ( \ , z = 0), in the wavelength range from 310 to 700 nm , m easured in various regions of the W orld Ocean. Jerlov distingu­ished five types of ‘ optically clean ’ ocean basins characterized by optical indexes I, IA, IB, II and III, and nine optical classes of litto ra l regions w ith indices from 1 to 9. These classes and the optical types of m arine basins differ w ith respect to the absolute values and spectral distributions of the downw ard irradiance a ttenuation coefficients. Exam ples are given in T a­ble 1 and Figure 1 . The d a ta represent the m ean values of K d(X ) for the0-10 m w ater layer for various classes and optical types of seas and oceans. The com plete Jerlov classification (Jerlov, 1978) includes the coefficient of variability w ith depth . In this respect this classification differs from other classifications presented here, because it considers not the apparen t optical properties of certain w ater masses b u t of m arine basins in general. Ho­wever, it also suffers from a num ber of draw backs. It is discrete, because the optical indices (m ore precisely, the num bers of the particu lar types and classes of basins) are discrete, and they are not analytically rela ted to the

9 7 5 3 1 IIIII IB IA I Type

_

Kd(X)

[10 3

m 1], in

various sea

and ocean

types and

classes according

to the

Jerlov optical

classificationTable

1. The

spectra of

the dow

nward

irradiance vertical

attenuation coefficients

in the

0-10 m

water

layer,

350240180 65 37 22 18 15310

390300230170120 32 17.510 7.8

6.2350

300 :

210160110 80 22 12.2

6.65.23.8

375

240160110 78 51 18.5

9.65.13.82.8400

190120 78 54 36 16 8.1

4.23.12.2

425

160 1

89 56 39 25 13.56.83.62.61.9

450

23 71 43 29 17 11.66.23.32.51.8

Wavel

475

99 58 36 22 14 11.57.0

.4.23.22.7

ength500

78 49 31 20 13 11.67.65.44.84.3

A [nm

] 525

63 46 30 19 12 12 8.97.26.76.3550 —

58 46 33 21 15 14.811.59.99.48.9

575

60 48 40 33 30 29.526 24.524 23.5600

65 54 48 40 37 37.533.531.531 30.5625

76 63 54 46 45 44.540 37.537 36 650

92 110

78 92

65 80

56 71

51 65

52 66

46.5 61

43.5 59

43 57

42 56

675 700

wavelength A( r >m]

Fig- 1. The spectra of the downward irradiance attenuation coefficients 7v(j(A); mean values for the surface layer - 0 m to 10 m depth - in various basins, according to the Jerlov classification. Roman and Arabic numerals denote the respective types and optical classes of the seas and the oceans

k d coefficients. Besides this, our experience indicates th a t m odel spec­tra of a tten u a tio n coefficients differ considerably in this classification from experim ental spectra of 7vj(A). These disadvantages lim it the practical ap­plicability of this classification to a significant ex ten t. Considerable progress has been m ade by the continuous classifications presented below.

2.2. T he P elev in and R utkovskaya optical classification o f m arine basins

The Pelevin and Rutkovskaya classification (1977, 1978; Rutkovskaya et al., 1982; Pelevin, 1985) is one of the first a ttem p ts to develop a conti­nuous optical classification. It concerns w ater masses in the euphotic zone irrespective of the depth , not m arine basins in general. It is based on over 500 experim ental 7Q(A) spectra, recorded in various regions of the A tlan tic , Pacific and Indian Oceans. The au thors of this classification studied the re­lationships betw een the 7vd(A) coefficients for various wavelengths from the 400-600 nm region and the light a ttenuation coefficient K d (A = 500 nm ) at

A = 500 nm . Using the linear regression m ethod they found the following relationship fundam ental to the continuous classification:

K j ° ( \ ) = A { A) + B ( A) · m , (2a)

where^.(A), -B(A) - empirical param eters of the relationship, listed in Table 2

(columns 2 and 3),m - optical w ater type index; it is equal to one hundred light a t­

tenuation coefficients a t A = 500 nm , i.e. m = 100 · (A = 500 nm ). This definition transform s the optical index in to

convenient num bers around 1 and m ore,K f ( X ) - downward irradiance a ttenuation coefficients w ith depth , rel­

ative to the base 10 logarithm ic scale of irrad iance.1 Since in this paper the A^°(A) coefficients are relative to the n a tu ra l

logarithm scale (see equations (3 )-(5 )), expression (2a) takes the form

K \ ° = (A )2 .3 -[A (A )+ J3 (A )-m ], (2b)

where the optical index is m — 100 · K d {A = 500 nm )/2 .3 .Equations (2a) and (2b) m ake it possible to determ ine the spectral func­

tions of light a ttenuation coefficients in the 400-600 nm range. As regards

the long wave range of the K d( A) spectrum , the differentiation of the coeffi­cient values is m uch smaller in practice th an in the short wave region. The reason for such behaviour comes from light absorption by w ater molecules, affecting m ainly the a ttenuation coefficient of red light. For this reason P e­levin and R utkovskaya suggest th a t when A > 700 nm the coefficients A'd(A) should be assum ed equal for all w ater types:

A'd(A) « 1.2 · aw( \ ) , (6)

whereaw( \ ) - generally accepted values of light a tten u a tio n in pure w ater (Popov

et a i , 1979); the num erical factor 1.2 is the ra tio of A'd(A)/a(A) typical for this spectral range, approaching the value of the so- called distribu tion function of the lum inous flux (according to the definition quoted by Preisendorfer(1961)).

1 According to equation (1), the downward irradiance attenuation coefficient Kd can be expressed as a depth derivative of the natural logarithm of irradiance:

Kd( z ) = —d \ n E d ( z ) / d z . (3)The Kd° coefficients presented by Pelevin et al. are defined as derivatives of the base 10 logarithm:

K 1d°(z) = - d^ f / zK (4)Thus the relationship between the coefficients is:

ÜTd = In 10 · ÜT50 « 2 . 3 - K i ° . (5)

Table 2. The list of param eters ^4(A), 5(A), applied in the optical classification of water masses by Pelevin and Rutkovskaya (eqs. (2a) and (2b)) and standard errors of attenuation coefficient K]°

A[nm]

A[10-2 m -1]

B[10~2 n ,"1]

[10“ " m "1]m < 2.5 2.5 < m < 5 5 < m < 10 10 < m < 20

1 2 3 4 5 6 7400 -0.88 1.60 0.0 1.0 1.0 2.0410 -1.08 1.69 0.4 1.0 1.0 1.5420 -1.05 1.64 0.4 0.5 0.5 1.2440 -0.85 1.48 0.2 0.3 0.4 0.8460 -0.67 1.31 0.18 0.25 0.25 0.5480 -0.48 0.15 0.09 0.15 0.25 0.3490 -0.12 1.05 0.06 0.2 0.2 0.2500 0 1 - _ _520 0.93 0.88 0.08 0.1 0.15 0.3530 1.1 0.83 0.1 0.2 0.25 0.4540 1.38 0.78 0.15 0.2 0.4 0.6560 2.34 0.64 0.3 0.3 0.5 0.8580 3.6 0.67 0.5 0.5 0.7 1.1590 5.2 0.71 0.9 0.9 1.2 2600 7.0 0.89 2 2 2 2

To determ ine K d ( A) in the 600-700 nm range, the au thors used linear in terpo lation betw een the relevant values of K d {A = 600 nm ) obtained in equation (2) and K d {A = 700 nm ) given by equation (6 ). Exam ples of spec­t r a of K d{A) obtained according to this procedure for various types of sea w ater of different optical indices m and for all the spectral intervals discus­sed are given in Figure 2.

Unlike Jerlov’s classification, the one proposed by Pelevin and R utkov­skaya is a continuous classification because the optical index m , and in consequence the K d( A) coefficients, m ay have a rb itra ry values. It is useful because the index m is easily determ inable, as is the full spectrum of K d ( A), from the a tten u a tio n coefficient m easured in the field - for example: ^ ( A i ) of one wavelength Ai from the 400-600 nm range. The values of m are then calculated from the relationships (after transform ation of equation (2a) or (2b))

A.) or /Q (A ,) - 2.3 · .4(A,)B ( A,) ° r m ~ ' (7)

In practice, the Pelevin and Rutkovskaya classification gives a m uch b e tte r spectral approxim ation of the K d{A) coefficients. This has been verified in experim ents in which the m ean square deviations of m odel coefficients and experim ental values were calculated (see columns 4 -7 in Table 2). The deviations were relatively small.

w avelength Al n m]

Fig. 2. The spectra of the downward irradiance attenuation coefficients /’Cd(A) for water of various optical indices m according to the Pelevin and Rutkovskaya classification

2.3. T he S m ith and Baker b io-optical classification o f natural w aters

Despite the considerable usefulness of the Pelevin and Rutkovskaya clas­sification, m uch b e tte r from the biological point of view are classifications combining the biological characteristics and optical types of w ater m as­ses. An early a tte m p t was presented by Sm ith and Baker, who applied the relationships between the spectra of the downward irradiance a ttenuation coefficient 10(A ) and chlorophyll a concentration in a sea w ater B a (Baker and Sm ith, 1982; Sm ith and Baker, 1978,1984). The to ta l coefficient K d (A) can be expressed by elem ents characterizing the particu lar optically active com ponents of sea w ater in accordance w ith the approxim ate dependence

I ( d( A) Si K w( A) + A'a(A) + I ( pl(A), (8a)

or, tak ing in to account the dependence of K pi on chlorophyll concentration

K d( A) * I ( w{A) + K a(A ) + k c( A) · B a , (8b)

whereK w(A) - contribution to downward irradiance a tten u a tio n due to pure

w ater,K pi(A) - contribution resulting from phytop lank ton pigm ents,

A a (A) - contribution from optically active adm ixtures, w ith the exception of photosynthetic pigm ents,

&c(A) - specific light a ttenuation coefficient of photosynthesic pigm ents, i.e. k c( A) = K pi ( \ ) / B a .

For the sake of simplicity, the au thors took K w( A) to be the results of m easurem ents of to ta l downward irradiance a ttenuation in the Sargasso Sea, one of the least polluted n a tu ra l basins, where optically active substances are practically non-existent. The spectral function of I ( w(A) is given in Table 3 (colum n 2). In the next stage, Sm ith and Baker statistically analysed the relationship K d ( A) — K W(X) = f ( B a ) a t various wavelengths from 350 to 700 nm . The empirical courses of these relationships were approxim ated by the following broken lines using linear regression m ethods:

• for B a < 1 mg · m ~3

K d {A) - K W(X) = A.’c,i(A) · B a , (9a)

• for B a > 1 mg · m -3

k-d(A) — I i w(X) = A A,2(A) + k c^ ( X ) · B a , (9b)

where^c.i> &c,2 ~ specific coefficients of downward irradiance a tten u a tio n by p hy to ­

p lankton pigm ents for small ( B a < 1 m g -m - 3 ) and high ( B a > 1 m g -m - 3 ) chlorophyll concentrations respectively,

-ft A,2 - contribution of optically active adm ixtures to the to ta l light a ttenuation ; the au thors in troduce it for w aters w ith a high chlorophyll content ( B a > 1 m g -m - 3 ). In w aters of low chloro­phyll concentration ( B a < 1 m g -m - 3 ) the effect of adm ixtures has been neglected, i.e. K a ,i — 0 .

The param eters of equation (9) are presented in Table 3. This rela­tionship forms the basis of a continuous optical classification in which the chlorophyll concentration B a is adopted as the optical index of the w ater type. Exam ples of Kj,( A) spectra determ ined for a num ber of basin classes (i.e. chlorophyll concentrations) by the application of equation (9 ) together w ith experim ental curves are given in Figure 3.

Chlorophyll a concentration applied as the optical w ater type index is an im portan t advantage of the Sm ith and Baker classification. Chlorophyll concentration conveys a lot of inform ation about the complex biological processes tak ing place in the ecosystem. Thus, the Sm ith and Baker classi­fication becomes a bio-optical one. However, it has a num ber of draw backs;

T able 3. The list of parameters applied in the bio-optical classification of water masses by Smith and Baker (eqs. (9a) and (9b))

A K w K * . 2 kc,i ^c, 2[nm] [m -1] [m-i] [m2 (mg Chi)-1] [m2 (mg Chi)1 2 3 4 5350 '0.059 0.177 0.249 0.066355 0.055 0.177 0.249 0.066360 0.051 0.177 0.249 0.066365 0.045 0.178 0.248 0.063370 0.044 0.179 0.245 0.061375 0.043 0.179 0.240 0.058380 0.040 0.179 0.237 0.055385 0.036 0.179 0.232 0.053390 0.031 0.177 0.277 0.051395 0.029 0.175 0.223 0.050400 0.027 0.172 0.216 0.049405 0.026 0.167 0.210 0.048410 0.025 0.162 0.205 0.047415 0.024 0.156 0.200 0.046420 0.024 0.150 0.194 0.045425 0.023 0.145 0.187 0.044430 0.022 0.137 0.181 0.042435 0.022 0.132 0.175 0.041440 0.022 0.125 0.168 0.039445 0.023 0.121 0.163 0.038450 0.023 0.116 0.158 0.037455 0.023 0.112 0.150 0.036460 0.023 0.110 0.146 0.034465 0.023 0.104 0.141 0.033470 0.023 0.100 0.135 0.031475 0.022 0.095 0.130 0.030480 0.022 0.091 0.125 0.029485 0.024 0.087 0.120 0.027490 0.025 0.084 0.115 0.026495 0.027 0.080 0.110 0.025500 0.029 0.077 0.105 0.024505 0.033 0.074 0.102 0.022510 0.037 0.071 0.096 0.021515 0.043 0.069 0.093 0.020520 0.048 0.066 0.088 0.019525 0.050 0.064 0.085 0.017530 0.050 0.061 0.084 0.016535 0.052 0.060 0.080 0.015540 0.055 0.059 0.076 0.014545 0.059 0.056 0.073 0.013

T able 3. (continued)

A K w ^ A ,2 kc} 1 &c,2[nm] [m-i] [m-i] [m2 (mg Chi)-1 ] [m2 (mg Chi)-1 ]1 2 3 4 5550 0.063 0.055 0.070 0.012555 0.067 0.054 0.070 0.011560 0.071 0.053 0.070 0.011565 0.074 0.052 0.071 0.010570 0.077 0.053 0.072 0.009575 0.082 0.054 0.074 0.009580 0.088 0.056 0.077 0.008585 0.099 0.059 0.085 0.008590 0.107 0.066 0.095 0.007595 0.121 0.091 0.110 0.007600 0.131 0.131 0.125 0.007605 0.146 0.150 0.148 0.007610 0.170 0.159 0.168 0.007615 0.188 0.165 0.184 0.006620 0.212 0.167 0.195 0.006625 0.244 0.169 0.205 0.006630 0.277 0.161 0.213 0.006635 0.300 0.137 0.222 0.007640 0.327 0.117 0.227 0.007645 0.339 0.095 0.231 0.008650 0.336 0.061 0.225 0.009655 0.337 0.037 0.205 0.011660 0.390 0.015 0.180 0.012665 0.425 0.002 0.156 0.014670 0.460 0.0 0.118 0.015675 0.485 0.0 0.088 0.016680 0.510 0.0 0.068 0.015685 0.540 0.0 0.045 0.014690 0.570 0.0 0.028 0.011695 0.600 0.0 0.015 0.008700 0.630 0.0 0.008 0.004

The m ost im portan t ones are:

• The relationship of kd(A) vs. chlorophyll concentration, expressed in equation system (9), are non-infinitesimal. M oreover, th e /^ (A ) values for B a = 1 m g · m -3 ( th a t is the boundary betw een the assum ed chlorophyll concentration ranges) determ ined from equations (9a) and (9b) are divergent.

• It has been assum ed th a t the specific coefficients of light a tte n u ­ation by pigm ents k c>\ = const for B a < 1 m g - m -3 and k c>2 = const for B a > 1 m g · m -3 for bo th chlorophyll concentration ranges, B a < 1 m g - m -3 and B a > 1 m g -m - 3 , and for given light waveleng­ths. In fact the coefficients significantly change as the chlorophyll content in w ater does so (see eq. (13) below).

• A nother simplification in troduced, sim ilar to the above, was the con­tribu tion of o ther optically active substances K& to the to ta l light a tten u a tio n (yellow substances and suspensions w ith the exception of phytop lank ton pigm ents). Especially in the range of small chlorophyll concentrations it was assum ed th a t K a , i = 0 . In fact, the photosyn- thetic pigm ents are always accom panied by optically active adm ix tu­res, and their concentrations, and in consequence their contributions to the to ta l light a ttenuation , gradually increase when m oving from poor basins to biologically flourishing ones.

w avelength A [ n m ]

Fig. 3. The spectra of the downward irradiance attenuation coefficients A'd(A) - experimental (broken line) and model (solid line), according to the bio-optical classification by Smith and Baker in water masses of different chlorophyll a con­centration Ba. The concentration Ba serves as the optical index of the water type. The respective spectra are for: 1 - Ba = 9.32 m g-m -3 , 2 - Ba = 2.25 m g-m -3 , 3 - Ba — 0.44 mg · m -3 , 4 - Ba = 0.03 mg · m-3

These disadvantages have been partly rem oved in the second version of the Sm ith and Baker classification (Baker and Sm ith, 1982). Here the relation­ship betw een the optical coefficients and the chlorophyll concentration is described in infinitesim al functions. However, ano ther variable has been in­troduced a p a rt from B a , i.e. the concentration of dissolved organic m atte r (DOM ). T he au thors adm it th a t this version is based on m odest experim en­ta l m ateria l and requires fu rther verification. M oreover,the practical use of this version is considerably lim ited by the need to know DOM values.

For these reasons, to fulfil the dem ands of this study, an original ‘ single­index ’ bio-optical classification of n a tu ra l w aters has been prepared by W ozniak, to a large ex ten t free of the drawbacks of the Sm ith and Baker classification.

2.4. W ozniak’s b io-optical classification o f natural w aters

This classification has not been widely published yet. It is based on the sta tistica l analysis of experim ental d a ta (obtained by W ozniak and his co- workers a t the In stitu te of Oceanology of the Polish A cadem y of Sciences, Sopot) and considers various w ater masses in n a tu ra l basins (oceans, seas and lakes) w ith chlorophyll a concentrations ranging from about 0.03 to approxim ately 60 m g -m - 3 . Thus, it includes nearly all trophic basin types, from oligotrophic to supereutrophic, and it is continuous. T he spectral range is from 350 nm to 750 nm. As in the Sm ith and Baker classification, the chlorophyll a concentration B a is taken to be the w ater type index. The proposed bio-optical classification is also based on empirical relationships betw een the coefficients of a tten u a tio n and B a . Several assum ptions are similar to those in the Sm ith and Baker classification: in particu lar, the general form of the A'd(A) = f ( B a ) relationship in equation (8 ) and the spectrum of light a ttenuation coefficients in pure w ater K w( A) (colum n 2 in Table 4) are the light a tten u a tio n spectra in the Sargasso Sea. By contrast,^ however, the approxim ation function (8) in W ozniak’s classification has an infinitesim al form (i.e. it is based on a single m athem atical equation). The param eters K& and k c have been m ade variable in relation to B a . Besides the sum m ary K d( A) coefficients, the classification is also extended to the spectra of the contributing coefficients: K pi(X) = k c( A) · B a - a tten u a tio n caused by phytop lank ton pigm ents, and A'a(A) - a tten u a tio n resulting from other optical adm ixtures.

The relationships of the respective light a ttenuation coefficients vs. the chlorophyll concentration are described by the following equations

• sum m ary coefficient

K d{A) = K w + B a · [Z?i(A) · e~ a^ Ba + k d<n( A)], (10)

• phy top lank ton pigm ent coefficient

Kpi(X) = B a ■ [C2(A) · e - a^ Ba + k e,n (A)], (11)

• optically active adm ixtures

K A (l·) = K d( \ ) - [ K W( \ ) + K pl( \ ) } = B a ■ [ C 1( \ ) ■ e - a'M■B a -

+ C 2(X) ■ e~a^ Ba + k d>n(X) - k c<n(X)]. (12)

The em pirical param eters in these relationships are C i(A ), C 2(A), ai(A ), a2(A), fcd n (A) and A:c>n(A). The last and the last b u t one param eters are

■

550540530520510500490480470460450440430420410400390380370360350I lnm

JA Table

0.0630.0550.0500.0480.0370.0290.0250.0220.0230.0230.0230.0220.0220.0240.0250.0270.0310.0400.0440.0510.0592 [m

XJK

w

4. T

he

0.5120.6040.6910.5870.5920.6070.5100.5040.5030.4890.5550.6210.5850.5350.4970.4220.4190.3570.2940.2930.3013 [m

^mg

Chi)-1]

a i

list of

parametei0.0437

0.05060.05750.06050.05980.06700.07720.07790.08660.09460.1050.1100.1180.1310.1350.1390.1460.1630.1810.2170.2604 [m

^mgC

hl)-1]

•s applied

in the

0.02860.03000.03030.03390.03860.04190.04640.05080.05270.05890.06250.06690.06780.06960.07140.07500.08030.08210.08390.08660.08755 [m

2(mgC

hl)'kd,n

Wozniak

hin-

1.891.912.002.001.841.621.541.491.601.621.701.721.731.821.851.952.002.092^573.923.836

-1] [m

a(mgC

hl)-1]“2

•optical classification

0.02390.03180.03700.04310.05280.06270.06830.07270.07980.08660.08620.08990.08560.08880.08230.06030.04760.03250.02670.01770.01177 [m

2(mg C

hi)-1]C

l

V*

>« vvvux

xu

uu

uv

unf

maccoc

0.01220.01370.01670.01970.02130.02330.02640.02930.03200.03440.03730.03830.03750.03600.03330.03070.02800.02330.01870.01710.01618 Fm

2(mgC

hl)-1lkc.n

(eqs. (10)—(12))

7507407307207106906806756706606506406306206106005905805705601 [nm

jλ

3.0202.4361.8521.3741.0020.5700.5100.4850.4600.3900.3360.3270.2770.2120.1700.1310.1070.0880.0770.0712 Im

"1]K

w

- - - - - - - 0.1730.1730.3470.3460.4210.4210.3130.3010.3310.3620.3950.3810.4623 [m

J(mg C

hi)-1]a\

0 0 0 0 0 0 0 0.004250.009380.01410.01640.02160.01850.01520.01610.01730.01860.02380.02900.04264 [m

^mgC

hl)-1]C

i

0.000010.000070.00040.00140.00440.01900.02580.02690.02670.02580.02410.02330.02320.02320.02230.02320.02410.02410.02500.02585 [m

2(mgC

hl)_1]kd,n

- - - - - - - ■

- - - 5.405.074.094.013.943.893.583.472.561.596 [m

^mgC

hl)-1Û2

0 0 0 0 0 0 0 0 0 0 0.000500.000640.001480.001900.03700.004920.006810.009010.01260.01587 [m

2(mgC

hl) *]

c2

0.000010.000070.00040.00140.00440.01820.02500.02600.02510.02130.01500.01200.01070.01060.01040.01070.01110.01130.01120.01158 [m

2(mgC

hl)-1lkc.n

the boundary values of the specific light a tten u a tio n coefficients ( /^ „ (A ) - the a tten u a tio n due to all com ponents except w ater, and k c,n ( A) - th a t due to phytop lank ton pigm ents). The real values of the coefficients in super- eutrophic basins ( B a > 1 m g -m -3 or B a —> oo) approach these boundary values.

Fig. 4. The spectra of the downward irradiance attenuation coefficients according to the Woźniak bio-optical classification: a - total attenuation K d(A), b - attenuation by phytoplankton K pi(X), c - attenuation by other optically active substances, i.e. yellow substances and suspensions K a (A), with the exception of phytoplankton. O - oligotrophic sea (Ba(0) < 0.2 m g-m -3 , M - mesotrophic sea (0.2 < B a (0) < 0.5 m g-m -3), P - intermediate sea: meso-eutrophic (0.5 < 5a(0) < 1.0 m g-m -3 ), E - eutrophic sea (Ba(0) > 1 .0 m g-m -3 )

These em pirical param eters were determ ined after a s ta tis tica l analy­sis of

• approxim ately 850 experim ental K d ( ^ ) spectra in relation to the chlo­rophyll concentration B a ,

• abou t 250 approxim ated2 K pi( \ ) spectra in relation to the chlorophyll concentration B a .

T he approxim ation was carried out by non-linear m ethods. The param eters calculated for various light wavelengths are presented in Table 4. The spec­tra l courses of the to ta l A) coefficient and the contributing K pi( \ ) and

^ a (A ) coefficients in w ater masses of various chlorophyll concentrations B a , derived from equations (10) - (12), are illustrated in Figure 4. The variabi­lity ranges of the particu lar coefficients in the given trophic types of basins are also presented.

Fig. 5. Experimental interrelations between chlorophyll a concentration Ba, in water masses of various and oceans: a - total coefficient of downward irradiance attenuation Kd( A) = 440 nm), b - component K pi( A = 440 nm) - coefficient of light attenuation by phytoplankton. Solid lines denote functions approximated by equ­ations (10) (Fig. a) and (11) (Fig. b). O - oligotrophic sea (B a(0) < 0.2 m g-m -3 , M - mesotrophic sea (0.2 < B a (0) < 0.5 m g-m -3), P - intermediate sea: meso- eutrophic (0.5 < 5a(0) < 1.0 m g-m -3), E - eutrophic sea (B a(0) > 1.0 m g-m -3)

Practice has proved th a t equations (10) - (12) give a good approxim ation of the experim ental functions of the particu lar coefficients versus chlorophyll a concentration a t a particu lar light wavelength: this is shown in Figure 5 for the K j and K pi coefficients displayed a t A = 440 nm . This emerges from the justified physical sense of these approxim ation functions. It is particularly obvious when K pi(A) = f ( B a ) . According to equation (11), the specific coefficient of light a tten u a tio n by phytoplankton pigm ents k c = K pi / B a is equal to

*c(A) = = C 2(A) · e~ a^ - Ba + k c>n( A). (13)

This m eans th a t the k c(X) coefficients have the highest values in oligotrophic w ater masses of low chlorophyll concentration, where th e pigm ent accom ­panying chlorophyll (i .e . o ther chlorophylls, carotenoids and phycobyli- nes) m ake a significant contribution to the general pigm ent concentration in the pho tosynthetic appara tu s of oligotrophic phytocoenoses, which was proved sta tistically by W oźniak and O strow ska (1990a). Following the in­crease in chlorophyll a concentration in w ater, the contribu tion of o ther pigm ents decreases, and so do the values of k c( A). T hus, the sm allest coef­ficient is lowest in eutrophic phytocoenoses and is approxim ately equal to k c(X) « k c<n(A). The averaged relationship betw een the specific coefficient of light absorption by phy top lank ton k c( A) = api ( X ) / B a and the chlorophyll a content (W oźniak, 1989; W oźniak and O strow ska, 1990b) reveal a similar behaviour. As regards the form of the function approxim ating the rela­tionship of the sum m ary coefficients K d(A) to the concentration B a (see equation (10)), it was assum ed th a t the specific coefficients of downward irradiance a tten u a tio n k d(X) of all the optically active com ponents except w ater, qualitatively identical to the phy top lank ton coefficients k c( A), are equal to

. . . . _ K d(X) - K w{ A) _ K pl(X) + K A (X) d ̂ > B a B a

= C 1{ X )e -a^ Ba + k d,n ( A). (14)

The in troduction of this analogy results from the correlation betw een the concentration of various optical com ponents and chlorophyll, a content (W oźniak, in p repara tion). The s ta tis tica l d istribu tion of these compounds gives rise to the significantly scattered d istribu tion of experim ental points in the diagram of K d(A) versus B a , as com pared to the case of K pi(A) versus B a (Figs. 5a and 5b). Finally, the approxim ation function of the /v'a(A) = f ( B a ) relationship (see equation (12)) emerges as the difference betw een equations (10) and (11).

The precision and accuracy of the W oźniak classification were verified experim entally by using 200 sets of experim ental d a ta on the m easured spectra of the to ta l downward irradiance a tten u a tio n coefficients K d( A), the com ponent spectra of the phy top lank ton a tten u a tio n coefficients de­term ined indirectly (from m easurem ents of absorption in vivo or in acetone ex tracts) K pi(A), and the m easured concentrations of chlorophyll a B a . The d a ta recorded in various sea types and a t different depths of the euphotic zones in these basins. Using the experim ental B a results, the m odel spectra of to ta l light a tten u a tio n and a tten u a tio n by phy top lank ton were obtained

VkNB: <

£k

>

Kpl(X)

Kd{ A)

typeSpectrum

Table 5.

E:- statistical error.- system

atic error,

Eutrophic >

1

Intermediate

0.5 -j-1

Mesotrophic

0.2 0.5

Oligotrophic

< 0.2

Eutrophic >

1

Intermediate

0.5 -i-1

Mesotrophic

0.2 -ż- 0.5

Oligotrophic

< 0.2

Bio-optical

Ba range

type of

basin [mg

· m-3]

(cperimental

verification

CTc [%

]<

sk>

[%]

[%]

<£*> [%

][%

]<

£k >

[%]

[%]

<€k>

[%]

[%]

< e*

> [%

]<re

[%]

< £k

> [%

][%

]<

£k> [%

][%

]<

£k >

[%]

Parameter

of the

Woznić

+14 +14

±18 ±28

±45 ±28

-11 -8

±42 ±29

-10 -10

±35 ±30

+9 +9

±28 ±30

+9 +9

±48 ±48

-7 +7

±41 ±23

-5 -10

+16 +14

±30 ±28

460<

420 4-

420

ik bio-optical

cl;+16±20

±26 -4±16 -8±21 +9

+10±29

±41 -7±20-10

+12±22

500 -r460

issific;±20 +3±23 -4±16 0±20 +8±19 +7±37 -3±21-11

+12±20

540 -j-500W

a>

ition

±20 ±22

+4 +7

±25 ±22

-2 -4

±19 ±14

-2 -4

±19 ±21

+3 +4

±21 ±14

+6 -4

±31 ±14

-3 +1

±13 ±10

-8 +7

±18 ±25

+8 +2

580 620

-7- 4

540 580

relength [nm

]

of w

ater m

as

±28 +6±26 -3±18 +1±18 +3±26 0±26 -3±18 +2±24 0660 -Î-620

ises

±17 +6±15 -3±16 -3±18 -1±19 +4±16 —6±18 -7±20 -7700 4660

±41 +5±60 -4±42 -3±45.+7 - - - - - - - -

> 700

from equations (10) and (11). These spectra were com pared w ith the expe­rim ental curves, which led to the respective relative errors £jt(A) a t various wavelengths

ckW = W , (15)

where _K m ea s (X )- m easured values of the K d(X) or K pi(A) coefficients,Kcal(X) ~ calculated values of th e coefficients.

In the next stage of verification, the m ean error < £*(A) > and s ta n ­dard deviation <re(A) were calculated, separately for each trophic type. The results are given in Table 5. The m ean error < £&(A) > is the m easure of the system atic error of the classification verified. On the o ther hand , the s tan d ard deviation Cfc(A) is the m easure of tlie random (sta tis tical) errors of th is classification. T he values presented in Table 5 are relatively small and com parable w ith the m easurem ent accuracy of the coefficients K d(A) and Kpi(X). This m eans th a t W ozniak’s classification is to a great extent correct and practicable.

3 . C o m p a r iso n o f s p e c tr a l o p tic a l p r o p e r t ie s o f n a tu ra lw a te r s a cco r d in g to v a r io u s c la ss if ic a tio n s

The spectral optical properties of n a tu ra l w aters determ ined on the ba­sis of various classifications differ in some respects and are sim ilar in others. This is illustrated in Figure 6 . The curves represent m odel spectra of the coefficients of to ta l downward irradiance a tten u a tio n in sea w ater I ( d( A) obtained from W ozniak’s classification versus the relevant spectra ob tai­ned from the classifications by Jerlov (Fig. 6a), Pelevin and Rutkovskaya (Fig. 6b) and Sm ith and Baker (Fig. 6c), as well as experim ental spec­t r a (Fig. 6d). To facilitate the com parison, the K d( A) spectra of similar K d(X = 500 nm ) values were selected in the case of the W ozniak and the Pelevin and Rutkovskaya classifications, or the spectra were chosen for the sam e chlorophyll a concentrations (when com paring the K d( A) spectra from the W ozniak classification w ith the Sm ith and Baker classification or w ith experim ental K d(A) spectra).

T he p icture shows sim ilar shapes of the spectra from all four classifica­tions analyzed, and bears a considerable resem blance to the experim ental spectra. This is evidence for their considerable correctness, th e m ore so th a t the discrepancies betw een the m odel curves in various classifications ensue from the individual selection of th e tested experim ental m aterial, experim ental errors and different m ethods of s ta tis tica l approxim ation of the da ta .

o b

c d

w avelength A l n m ]

Fig· 6 . Comparison of the spectra of the total coefficients of downward irradiance attenuation K d(A) obtained by Wozniak’s classification (solid lines) with the fol­lowing spectra (broken lines): a - Jerlov’s classification, b - Pelevin and Rutko- vskaya’s classification, c - Smith and Baker’s clasification, d - experimental exam­ples. The respective optical indexes are: in Figure a: l-j-3, B a = 0.11 m g-m -3 , Ba = 1.8 m g-m -3 , B a = 13 m g-m -3 ; 1' -f- 3' Jerlov indices IB, 1, 7; in Figure b:l-i-3 Ba = 0.45 m g-m -3 , Ba = 1.6 m g-m -3 , Ba = 7.4 m g-m -3 ; l '-f-3 ' Pelevin and Rutkovskaya indices m = 2, m = 5.3, m = 15; in Figure c in both cases: l-i-3 and 1' -j- 3' Ba — 0.1 m g-m -3 , B a = 1.0 m g-m -3 ; in Figure d in both cases: 1-^3 and I ' a 3' B a = 0.5 m g-m -3 , B a = 1.0 mg - m-3 and Ba = 10 m g-m -3

T he greatest analogy occurs betw een the m odel spectra of W oźniak and those of Pelevin and Rutkovskaya (Fig. 6b). They also closely resemble the experim ental spectra (Fig. 6d).

There is also a considerable analogy betw een the m odel A'd(A) spectra from W ozniak’s classification and those from Sm ith and B aker’s (Fig. 6c). They agree particu larly well a t small concentrations of chlorophyll a. The sim ilarities result from the adopted identical K w(A) spectrum , i.e th a t of light a tten u a tio n by pure w ater, because in oligotrophic basins K w{A) is the dom inating factor in the to ta l light a tten u a tio n K d ( A). T he differences appear a t higher concentrations of B a . However, the differences concern th e absolute values, no t the relative spectral functions, i.e. the form of the light a tten u a tio n spectrum . The absolute values in the Sm ith and Baker classification are either higher (for m oderate chlorophyll a concentrations, e.g. B a = 1 m g - m -3 - see curves 2 and 2 ’ in Fig. 6c) or lower (for high chlorophyll a concentrations, e.g. B a = 10 m g - m -3 - see curves 3 and 3’ in Fig. 6c) th an the K d {A) values estim ated by the W oźniak classification. We believe th a t these discrepancies are the result of the draw backs of the Sm ith and B aker classfication discussed in Section 2 .

The Jerlov classification shows the g reatest disagreem ent (Fig. 6a) with· o ther classifications. The discrepancies in the form of the /0 (A ) spectrum appear m ainly in optically contam inated seas. Jerlov’s spectra also indicate higher values of the Jvd(A) coefficients in the short wave region.

4. E x te n s io n o f W o z n ia k ’s b io -o p t ic a l c la ss fic a tio n o f w a term a sse s to b a sin s

The classifications presented in Section 2 - the optical one by Pelevin and Rutkovskaya, the bio-optical one by Sm ith and Baker, and the bio- optical by one W oźniak - characterize w ater masses of certain optical indices regardless of their depth . The changes in chlorophyll a concentration serve as the optical index in bio-optical classifications. It is well established th a t the chlorophyll a concentration alters w ith depth , so the light a tten u a tio n coefficients correlated w ith B a a lter accordingly. Experim ental examples of such changes in chlorophyll a concentration 2?a(z), and in the K d (A = 500 nm , z ) coefficient a t a selected wavelength of 440 nm are illu stra ted in Figure 7.

To m ake the optical classifications of w ater masses applicable to the cha­racterization of light transm ission in to the sea it was necessary to consider the changes of th e relevant optical indices w ith depth , i.e. chlorophyll a concentration changes w ith depth.

a

c h lo r o p h y l l a c o n c e n t r a t i o n B a l r r i g - m 3]

b

Pig· 7. Examples of experimental vertical profiles: a - chlorophyll a concentration Ba: 1 - central Indian Ocean, 2-r-5 -central Atlantic Ocean, 6-7-7 - Atlantic Ocean- the Ezcura Gulf, 8-j-ll, 13 - the Baltic Sea and the Gulf of Gdansk, 12, 14, 15 - the Black Sea and the Gulf of Burgas; b - coefficient of downward irradiance atte­nuation at 440 nm Kd(X = 440 nm): 1 , 2 - the Indian Ocean (around Mauritius),3 ~ the Atlantic Ocean (the Canary Islands), 4 - the Atlantic Ocean (Antarctic), 5 , 8 - the Baltic Sea (the Gdańsk Deep), 6 , 9 - the Black Sea (the Gulf of Burgas), 7 - the Baltic Sea (the Gotland Deep), 10, 11 - the Gulf of Gdansk, 12 - Puck Bay. K d values averaged in 5 m water layers

distributionsexperim

entalanalyzedN

umber of

0.10.5-0.11-0.53-15-310-520-1030-2060-3090-60

TM [%]

Table 6.

Vertical

distribution of

chlorophyll a in

various sea

types (after

Wozniak, in

preparation)

Water

layer in

Relative’concentration

of chlorophyll a B

a(T\f)/Ba(z

= 0)

[dimensionless]

transmission

scale O

ligotrophic sea

Mesotrophic

sea Interm

ediate sea

Eutrophic sea

0.440.651,312.062.062.021.541.451.091.06

Mean

Ba(0)

<

165 ±0.38±0.41±0.85±

1.16±1.25±1.46±0.76±0.65±0.39±0.20

deviationStandard

0.2 mg

· m-3

0.791.171.261.571.631.661.371.311.211.07

Mean

0.2 <

Ba{0

105 ±0.40±0.85±0.98±0.91±0.74±0.78±0.46±0.48±0.34±0.17

deviationStandard

) <

0.5 mg

· m

8!

0.831.091.141.211.301.391.451.271.221.18

dM

ean £

0.5 <

Ba(0)

«3

meso-

and

5 ±0.26±0.29±0.39±1.00±0.79±0.44±0.45±0.42±0.42±0.28

leviationStandardC

1 mg

· m~

eutrophic

0.670.810.941.031.081.161.201.241.181.12

Mean

3B

a(0) >598 ±0.51

±0.51±0.50±0.44±0.49±0.60±

0.49±0.39±0.38±0.33

deviationStandard

1 mg

· m-3

Fig· 8. Statistical distribution of vertical profiles of relative chlorophyll a con­centration Ba{TM ) / B a ( z = 0) in typical basins: O - oligotrophic sea (Ba(0) < 0.2 m g-rn-3 , M - mesotrophic sea (0.2 < Ba(0) < 0.5 m g-m -3), P - inter­mediate sea: meso-eutrophic (0.5 < £a(0) < 1.0 m g-m “ 3), E - eutrophic sea (Ba(O) > 1.0 m g-m -3). Optical depth expressed on the scale of PAR irradiance transmission into the basin. Horizontal segments denote the standard deviation of the relative chlorophyll a concentrations (a); examples of hypothetical vertical pro­files of absolute chlorophyll a concentrations Ba in the marine basins of different trophisms (Wozniak and Ostrowska, 1990a) (b)

T he bio-optical classification of n a tu ra l w aters by W oźniak has been extended to basins using the s ta tis tica l regularities of the vertical d istribu­tion of chlorophyll a in the sea published earlier by W oźniak and Ostrowska (1990a) and W oźniak (in p reparation) (see Table 6 and Fig. 8). These papers present m ean vertical profiles of chlorophyll a concentration in various m a­rine basins in relation to depth or optical depth. O ptical depth in the la tte r case is expressed in term s of the transm ission in the sea of the downward

Fig. 9. Vertical profiles of the coefficient of downward irradiance attenuation: total (Kd) and by phytoplankton (Kpi), for light of a wavelength A = 440 nm in various bio-optical basins characterized by the surface chlorophyll a concentration B a (0) according to the Woz'niak bio-optical classification. Kd and K pi changes presented in relation to the optical depth expressed by the irradiance transmission coefficient in the sea Tm and in relation to the actual depth z: a - Kd versus Tm , b - Kd

Kpi versus Tm, d - K piversus z, c versus z

irradiance, in tegral in the PA R3 spectral range. Applying these d a ta and using equations ( 1 0 ) - ( 12) the vertical d istributions of the to ta l K j ( A, z ) and of com ponent K pi(A, z ) and A 'a (A , z ) - the coefficients of light a tten u a tio n - are determ ined for various bio-optical sea types. The surface concentration of chlorophyll a B a ( z = 0) is the m ain input d a ta and the index of sea type. Examples of vertical profiles of K d {A) = 440 nm , z ) in various seas, related to actual and optical depths, are shown in Figure 9.

The figure shows th a t the model vertical profiles m atch the experim en­ta l profiles of light a ttenuation coefficients (Fig. 7b). I t is obvious th a t the coefficient changes w ith depth follow the chlorophyll changes (Figs. 7a and 8 ). The figure indicates the occurrence of certain levels w ith m axim um light a ttenuation , corresponding to the levels of m axim um chlorophyll concentra­tion. Above and below these m axim a the values of the light a tten u a tio n coefficient decrease. Hence, as w ith chlorophyll, the location of the a tte n u ­ation m axim um changes according to the trophic type of the basin. The a ttenuation m axim a a tta in the greatest depths in oligotrophic seas, and as the trophism of the basin increase they approach the surface.

5 . F inal rem arks

For biological practice the au thors of this paper recom m end Pelevin and R utkovskaya’s optical classifications or W ozniak’s bio-opt ical classification. The argum ents presented in Sections 2 and 3 against the o ther classifi­cations show their m ore lim ited applicability. The Jerlov classification is discrete while the m ajo r draw backs of the Sm ith and Baker classification

are the discontinuity of the approxim ation functions of the light a tten u a tio n coefficients K d ( to ta l) , K& (a com ponent related to yellow substances and suspensions, b u t no t pigm ents), K vi and k c (com ponents related to ordinary and specific a ttenuation by phytoplankton pigm ents) in relation to chloro­phyll a concentration. These classifications do not agree so well w ith the experim ental da ta .

Particu lar a tten tion is draw n to W ozniak’s bio-optical classification. It takes in to account the spectra of the to ta l light a ttenuation coefficient K d ( A), and its com ponents K W(X) K vi{ \ ) and A 'a (A ). In the version ex ten­ded to m arine basins it additionally includes the coefficient changes w ith depth. This classification reveals a feature im portan t from the biological point of view, i.e. its double bio-optical character. The index of the basin type, i.e. the surface concentration of chlorophyll a - B a ( z — 0), enables

PAIl - photosynthetic available radiation. In oceanological measurements PAR ranges from 400 nm to 700 nm.

Fig. 10. Vertical profiles of the relative downward irradiance in the PARrange (400- 700 nm) T m ( z ) = E d tPAR(z ) /Ed„PAR(z = 0) in various types of seas according to the Wozniak bio-optical classification of basins (a). A diagram for the approximate determination of the optical index (Number - No.) with the application of various parameters: Ba(0 ) - surface concentration of chlorophyll a; 70,p a r - coefficient of Ed,p a r irradiance attenuation near the surface; K&,p a r - mean coefficient of E d , p a r irradiance attenuation in the 0-30 m layer; z, - Secchi disc visibility; J - basin type according to Jerlov; m - Pelevin and Rutkovskaya’s optical index (b)

bo th the vertical profiles of the light a ttenuation coefficient (and consequen­tly an estim ation of the underw ater irrad iation in the PA R range) and the vertical profiles of chlorophyll a concentration to be obtained. T he features presented have played an im portan t role in the creation of th e algorithm for determ ining-prim ary production by satellite sensing4. In addition, the charts of chlorophyll a d istribution in the surface w ater layers of various seas and oceans published by Krey and Babenerd (1976) and M ordasow a (1976) (see also W ozniak and O strow ska, 1990a) m ay also be applied as indicators of a particu lar bio-optical basin type in the W orld Ocean.

Finally, we present a practical diagram for a rough estim ation of the bio-optical type of a basin and the relative vertical d istribu tion of PA R do­wnw ard irradiance in the sea - Figure 10. Figure 10a illustrates th e vertical distributions of the to ta l downward irradiance from 400 nm to 700 nm in various bio-optical basin types. They were obtained from the spectral and vertical d istributions of A’<i(A,z) obtained from the extended W ozniak clas­sification. Figure 10b shows the possibility of determ ining the bio-optical basin type from factors o ther th an chlorophyll concentration B a ( z = 0), e.g. the Jerlov or Pelevin and Rutkovskaya indices and o ther param eters. The diagram has been constructed from analogies betw een the various op ti­cal classifications (see Section 3) and the relationship betw een the optical param eters.

Acknow ledgem entsThe au thors wish to thank M r. Slawomir K aczm arek, M .Sc., of the

Biophysics L aboratory of the In stitu te of Oceanology PAS for the necessary com puter calculations.

R eferen ces

Baker K. S, Smith R. C., 1982, Bio-optical classification and model of natural waters, 2, Limnol. Oceanogr., 27 (3), 500-509.

Dera J., 1983, Marine physics, PWN, Warszawa, 431 pp., (in Polish), [1992, Elsevier Sci. Publ. Comp., Amsterdam - Oxford - New York, 516 pp. (in English)].

Jerlov N. G., 1951, Optical studies of ocean water, Rep. Swedish Deep-Sea Exped.,3, 1-59.

Jerlov N. G., 1961, Irradiance in the sea in relation to particle distribution. Symposium on radiant energy in the sea, IUGG Monograph, 10, Helsinki, 3-8.

Jerlov N. G., 1968, Optical oceanography, Elsevier Sci. Publ. Comp., Am sterdam - Oxford - New York, 194 pp.

4 See our article: Pelevin et al. (1991) in this volume.

Jerlov N. G., 1976, Marine optics, Elsevier Sci. Publ. Comp., Amsterdam - Oxford- New York, 231 pp.

Jerlov N. G., 1977, Classification of sea water in terms of quanta irradiance, J. Constr. Explor. Mer., 37, 181-287.

Jerlov N. G., 1978, The optical classification of sea water in the eutrophic zone, Report No. 36, K0benhavns Univ. Instit. Fys. Oceanogr., Copenhagen.

J0rgensen B. B., Des MaraisD. J., 1988, Optical properties o f benthic photosynthetic communities: Fiber-optic studies of cyanobacterial mats, Limnol. Oceanogr., 33(1), 99-113.

Krey J., Babenerd B., 1976, Phytoplankton production - atlas of the international Indian Ocean expedition, Inst, fur Meereskunde - Kiel Univ., 70 pp.

Mordasowa N. V., 1976, Raspredeleniye khlorofilla a v mirovom okeane, 01, 9, 3, CNIITEIRH SSSR, Moskva, 49 pp, (in Russian).

Morel A., 1978, Available, usable and stored radiant energy in relation to marine photosynthesis, Deep-Sea Res., 25(8), 673-688.

Morel A., 1982, Optical properties and radiant energy in waters of the Guinea dome and Mauritanian upwelling area in relation with the primary production, J. Constr. Explor. Mer., 180, 94-107.

Morel A., Berthon J. F., 1989, Surface pigments, algae biomass profiles, and potential production of the euphotic layer. Relationships reinvestigated in view of remote sensing applications, Limnol. Oceanogr., 34(8), 1545-1562.

Pelevin V. N., 1985, Opticheskiye yavleniya. [In:] Monin A. S., Yavleniya na poverkhnosti okeana, Gidrometeoizdat, Leningrad, 299-343, (in Russian).

Pelevin V. N., Rutkovskaya V. A., 1977, Ob opticheskoy klassifikatsyi okeanskikh vod po spektralnomu oslableniyu solnechnogo izlucheniya, Okeanologiya, XVII, 1.

Pelevin V. N., Rutkovskaya V. A., 1978, Ob osłabieni fotosintecheskoy radiatsyi v vodakh Tikhogo okeana, Okeanologiya, XVIII, 4, 619-626.

Pelevin V. N., Rutkovskaya V. A., 1980, Ob osłabieni potoka solnechnoy energii s glubinoy v vodakh Indiyskogo okeana. [In:] Pelevin V. N., Kozlaninov M. M. (Eds), Svetovyye polya v okeane, AN SSSR, Moskva, 73-85, (in Russian).

Pelevin V. N., Woźniak B., Koblentz-Mishke O. J., 1991, Algorithm fo r primary production estimation in the sea from satellite sensing, Oceanologia, 31, 57-72.

Popov N. I., Fedorov K. N., Orlov V. M., 1979, Morskaya voda, Nauka, Moskva, 327 pp, (in Russian).

Preisendorfer R. W., 1961, Application of radiative transfer theory to light measurements in the sea, Int. Union Geod. Geophys. Monogr., 10, 11-30.

Prieur L., Sathyendranath S., 1981, A n optical classification of coastal and oceanic waters based on the specific absorption curves of phytoplankton pigments, dissolved organic matter and other particulate materials, Limnol. Oceanogr., 26(4), 671-689.

Rutkovskaya V. A., Pelevin V. N., Voitov V. I., 1982, On penetration of photosynthetically-active solar radiation into the Atlantic ocean waters, Okeanologiya, XXII, 3, 398-403.

Siegel D. A., Dickey T. D., 1987, Observation of vertical structures of the diffuse attenuation coefficient spectrum, Deep-Sea Res., 34(4), 547-563.

Simonot J. Y., Le Truet H., 1986, A climatological field of mean optical properties of the World Ocean, J. Geophys. Res., 91(C5), 6642-6646.

Smith R. C., Baker K. S., 1978, Optical classification of natural waters, Limnol. Oceanogr., 23 (2), 260-267.

Smith R. C., Baker K. S., 1984, The analysis of ocean optical data, Proc. Soc. Photo-optical Instrum. Energ., 489, Ocean Optics, VII, 119-126.

Smith R. C., Prézelin B. P. Bieligare R. R., Baker K. S., 1989a, Bio-optical modeling o f photosynthetic production in coastal waters, Limnol. Oceanogr., 34(8), 1524-1544.

Smith R. C., Merra J., Perry M. J., Baker K. S., Swift E., Buskey E., Kiefer D. A., 1989b, Estimation of a photon budget for the upper ocean in the Sargasso Sea, Limnol. Oceanogr., 34(8), 1673-1693.

Vinogradov M. E., Shushkina E. A., 1987, Funktsyonirovaniye planktonnykh soobshchestv epipelagyali okeana, Nauka, Moskva, 240 pp, (in Russian).

Woźniak B., 1989, Semi-empirical, mathematical modelling of phytoplankton absorption properties o f the World Ocean, CBO 16th Conf. of Baltic Oceanogr., Kiel, Proc. 2, Inst. Mar. Res., 1160-1169.

Woźniak B., Energetyka morskiej fotosyntezy (Energy balance of marine photosynthesis), (in preparation).

Woźniak B., Ostrowska M., 1990a, Composition and resources of photosynthetic pigments of marine phytoplankton, Oceanologia, 29, 91-115.

Woźniak B., Ostrowska M., 1990b, Optical absorption properties of phytoplankton in various seas, Oceanologia, 29, 117-146.