# M Ostrowska - Baltic new mathematical expressions part 1 total - страница 1

Remote sensing of vertical phytoplankton pigment

OCEANOLOGIA, 49 (4), 2007.

pp. 471-489.

distributions in the

© 2007, by Institute of Oceanology PAS.

Baltic: new mathematical expressions. Part 1: Total

KEYWORDS

Baltic Sea

chlorophyll a distribution[1] ChbrophyU a concentration

1 ^ Vertical distribution

Miroslawa Ostrowska1 Roman Majchrowski2 Joanna Ston-Egiert1 Bogdan Wozniak1'2 Dariusz Ficek2 Jerzy Dera1

1 Institute of Oceanology, Polish Academy of Sciences,

Powstancow Warszawy 55, PL-81-712 Sopot, Poland; e-mail: ostra@iopan.gda.pl

2 Institute of Physics, Pomeranian Academy,

Arciszewskiego 22B, PL-76-200 Slupsk, Poland; e-mail: majchrowski@apsl.edu.pl

Received 6 September 2007, revised 29 November 2007, accepted 3 December 2007.

Abstract

This article is the first in a series of three describing the modelling of the vertical different photosynthetic and photoprotecting phytoplankton pigments concentration distributions in the Baltic and their interrelations described by the so-called non-photosynthetic pigment factor. The model formulas yielded by this research are an integral part of the algorithms used in the remote sensing of the

Remote sensing

-Ф-

Baltic ecosystem. Algorithms of this kind have already been developed by our team from data relating mainly to oceanic Case 1 waters (WC1) and have produced good results for these waters. But their application to Baltic waters, i.e., Case 2 waters, was not so successful. On the basis of empirical data for the Baltic Sea, we therefore derived new mathematical expressions for the spatial distribution of Baltic phytoplankton pigments. They are discussed in this series of articles.

This first article presents a statistical model for determining the total concentration of chlorophyll a (i.e., the sum of chlorophylls a + pheo derived spec-trophotometrically) at different depths in the Baltic Sea Ca(z) on the basis of its surface concentration Ca (0), which can be determined by remote sensing. This model accounts for the principal features of the vertical distributions of chlorophyll concentrations characteristic of the Baltic Sea. The model's precision was verified empirically: it was found suitable for application in the efficient monitoring of the Baltic Sea. The modified mathematical descriptions of the concentrations of accessory pigments (photosynthetic and photoprotecting) in Baltic phytoplankton and selected relationships between them are given in the other two articles in this series (Majchrowski et al. 2007, Wozniak et al. 2007b, both in this volume).

1. Introduction

The 'light-marine photosynthesis' models that we have been developing for the remote sensing of marine ecosystems (e.g., Wozniak et al. 2003) require, among other things, the determination of the vertical distributions of the concentrations Cj (z)1 of the various phytoplankton pigments in the sea: the principal plant pigment chlorophyll a, Ca(z), and accessory pigments - photosynthetic pigments like chlorophylls b, Cb(z), chlorophylls c, Cc(z) and phycobilins Cphyc(z), photosynthetic carotenoids (PSC), Cpsc(z) and photoprotecting carotenoids (PPC), CPPC(z).Knowledge of the vertical distributions Cj(z) of all these pigment groups, and also of their mutual proportions as given by the non-photosynthetic pigment index fa (Ficek et al. 2000), is essential for estimating, for example, the absorptive properties of phytoplankton in the sea and the quantum yield of photosynthesis at different depths in the sea; from these magnitudes the vertical distributions of the primary production of organic matter in the marine environment can be calculated. The model formulas presented in this series of articles form an integral part of the algorithms permitting the efficient monitoring of the Baltic ecosystem by remote sensing.

In our earlier 'light-marine photosynthesis' model for determining the vertical distributions of pigment concentrations in the ocean Cj(z) (Wozniak et al. 2003, Ficek et al. 2003) we used model formulas derived from empirical research and the modelling of the photo- and chromatic acclimation of

phytoplankton. These formulas enable the concentrations of pigments at different depths z in the sea Cj(z) to be determined from two remotely measured parameters - the total surface chlorophyll a concentration Ca(0), and the spectral downward irradiance at the sea surface Ed(X, 0).They are as follows:

• Ca(z) = f (Ca(0)) - the dependence of the chlorophyll a concentration (Ca) atdifferentdepths z in the sea on its surface concentration. We derived this formula for oceanic waters (see Wozniak et al. 1992a,b).

• Cb(z) = f (Ca,Fb), Cc(z) = f (Ca,Fc), Cpsc(z) = f (Ca,Fpsc) - the respective dependences of the concentrations of chlorophylls b, c and of PSC on the chlorophyll a concentration (Ca) spectral fitting functions (Fb, Fc, Fpsc), which are determined from known irradiance conditions in the sea and the absorption properties of these pigments. We derived the relationships for the concentrations of these pigments in Case 1 oceanic waters (see Majchrowski & Ostrowska 1999, 2000, Majchrowski 2001, Wozniak et al. 2003).

• CPPC(z) = f (Ca,PDR*) - the dependence of photoprotecting carotenoids on the chlorophyll a concentration (Ca) Potentially Destructive Radiation (PDR*), which depends, in turn, on the irradiance conditions in the sea and the specific coefficients of light absorption by chlorophyll a. We established this relationship for oceanic waters (see, e.g., Majchrowski & Ostrowska 1999, 2000, Majchrowski 2001).

• fa(z) = f (apippp, apl>psp, Cppp, Cpsp, PAR(0), т)- thedepen-dence of the non-photosynthetic pigment factor on the following: a) the irradiance at the sea surface by the photosynthetically available radiation (PAR); b) the total concentration of all photosynthetic pigments (PSP) - Cpsp and photoprotecting pigments (PPP) - Cppp, and their specific absorption coefficients in vivo, apl psp,and apl ppp; c) the optical depth т in the sea. We derived this relation for oceanic

waters (see Ficek et al. 2000, Ficek 2001, Wozniak et al. 2003).

Earlier we had also developed a preliminary model description of the vertical distributions of the chlorophyll a concentration in the Baltic (Wozniak et al. 1995a,b), based on mathematical formulas resembling those for oceanic waters, but which took account of the seasonal changes occurring in the Baltic. Unfortunately, that description failed to live up to expectations.

We also recently attempted to adapt the oceanic 'light-marine photosynthesis' model to the remote sensing of the Baltic ecosystem. Again, the earlier formulas for determining depth profiles of pigment concentrations in the clear, Case 1 waters of the oceans failed to produce results of a similar

quality when applied to the algorithms for remotely sensing primary production in the Baltic. Also, the precision of the formula for calculating the depth profiles of chlorophyll a, modified for the Baltic to allow for its seasonal variations, was poor. The reasons for this are to be sought in the specifics of Baltic waters. These are brackish (Baltic Proper surface waters ~ 6 — 8 PSU) and contain considerable amounts of anthropogenic substances - dissolved and suspended yellow substances as well as other optically active substances. Any description of the adaptation and acclimation of algae to the conditions prevailing in these waters therefore appears to be a much more formidable task than for Case 1 waters.

In response to these arguments, our objective was to derive more precise, though not necessarily more complicated, mathematical formulas for determining vertical concentration profiles of chlorophyll a, Ca(z), accessory pigments Cj(z) and the factor fa in the Baltic. To this end, we accumulated a bank of suitable empirical data from 1978-2005. These data were subjected to statistical analysis: this enabled us to derive new formulas for the Baltic, the utility of which we then tested in satellite algorithms for determining primary production in the Baltic. The subsequent empirical verification of these formulas showed them to be of a far superior precision than the earlier ones, which were mentioned above.

The present paper, the first in a series of thematically linked articles, presents the modified mathematical description of the vertical distributions of the total chlorophyll a concentration in the Baltic. The other two papers in the series will deal with the modified mathematical descriptions of the accessory pigment concentrations in Baltic phytoplankton (part 2, Majchrowski et al. 2007, this volume) and the non-photosynthetic pigment factor fa characteristic of Baltic waters (part 3, Wozniak et al. 2007b, this volume).

2. Characteristics of the empirical material

Our statistical analysis is based on numerous empirical data sets, systematically collected over many years (1978-2005) and stored in the Oceanographic Data Bank at IO PAS. Most of this work was funded by the Committee for Scientific Research and the Ministry of Scientific Research and Information Technology through project PBZ-KBN 056/P04/2001 (The study and development of a satellite system for monitoring the Baltic ecosystem).

Chlorophyll a concentrations were measured at different depths in the sea using the traditional spectrophotometric method (Strickland & Parsons

1968) over very many years (1978-2005) and at all seasons, mainly from r/v 'Oceania', but also from r/v 'Baltica' and other research vessels. For the purposes of our analysis some 5000 vertical profiles of chlorophyll a were gathered, measured in Baltic Sea basins of different trophic index, i.e., in different regions of this sea, but mostly in its southern part. We may therefore make the assumption that the results analysed here are representative of all situations encountered in the southern Baltic, but to a certain extent also in the adjacent regions. Table 1 lists the numbers of vertical profiles of total chlorophyll a concentration, Ca(z),estimated spectrophotometrically in samples of water drawn from different depths in the sea. Each Ca(z) profile specified in Table 1 consists of at least 5 measurement points at particular depths z. The table also shows the number of Ca(z) profiles measured in different trophic types of Baltic water and in each month of the year. In some cruises chlorophyll a concentrations were measured not only spectrophotometrically, but also with an in situ fluorescence technique (see, e.g., Ostrowska et al. 2000a,b, Ostrowska 2001) using a PumpProbe fluorimeter (Ecomonitor, Moscow) calibrated in total chlorophyll a concentration units [mg tot. chl a m-3] (Ostrowska et al. 2000a,b, Ostrowska 2001). Table 2 lists the vertical

Table 1. Number of vertical profiles of the total chlorophyll a concentration, consisting of no less than 5 measurement points at different depths, measured spectrophotometrically in 1978-2005, and classified according to trophic type of water and month of the year (season)

reb re re

ts m ber mb mb su e b me m

"3

1

2

3 4 5 6

7

8

9

10

11

12

13

14

15

O1

Ca < 0.05

0 10 0

1

0

0

0

0

0

4

0

6

O2

0.05 <Ca <=0.1

0 110

8

0

1

0

0

0

0

0

11

O3

0.1 <Ca <=0.2

1

0

2

5

16

5

0

2

0

0

4

3

38

M

0.2 <C\ <=0.5

41

32

60

10

42

65

2

4

0

3

20

1

280

I

0.5 <Ca <=1

112

69

67

101

95

40

26

33

4

33

130

19

729

E1

1 <Ca <=2

16

9

46

101

263

96

84

111

## Похожие статьи

M Ostrowska - Baltic new mathematical expressions part 1 total