GISdevelopment.net ---> AARS ---> ACRS 1990 ---> Poster Session

Remote determination of the sea-surface chlorophyll a concentration along the coastal water by FY-1 satellite

Lin Shouren, Han Along and Zhang Gongwei
Zehijanna Remote Sensing Center,
The Second Institute of Oceanography,
SOA, Hangzhou, PRC


Abstract
This paper introduces a result of studying ocean colour experiment of "FY-1" satellite. A principle of spectral equivalent albedo is defined as FY-1 data processing. The ratio of FY-1 VHRSR's channe 1-3 and channe 1-4 equivalent albedo is employed to estimated the sea-surface chlorophy11 a concentration. Chlorophy11 a concentration is calculated from following formula:

Nc = a * R-b (a = 3.3336 and b = 5.2316)


Where a and b are both contant values.

Finally, the distribution map of sea-surface chlorophyll a concentration retrieved form FY-1 data (Sep. 21, 1988) demonstrates that sea-surface chlorophy11 a concentration is between 0.6 and 5.0 mg/m3 at Bohai Bay. This result agrees with those of surveying along the East and South China Sea.

Introduction
The change of ocean color is due to the suspended matter, such as the particulate matter, phytoplankton, and the dissolved material in the water. Fishing ships need the ocean color data to locate fish population, and to optimize ship routes, thereby, improving catch efficiency, and reducing costs. The research user group needs only continuous availability of ocean color data to specify quantitatively the ocean's role in the global carbon cycle and to determine the magnitude and variability of annual primary production by marine phytoplankton on a global scales. Recently, remote measurement of the change of ocean color has become very important.

In the oceanography from space, the remote sensing of ocean color focuses on the measurement of chlorophyll concentration. From 1978 to 1986, the Coastal Zone Color Scanner (CZCS) aboard the Nimbus-7 satellite provided the first ocean color data. During this period researchers demonstrated that:
  1. The technology of remote sensing can be used to survey and map the distribution of chlorophyll on the global scale. (1) (2) (3)

  2. An algorithm of G/B ratio for estimating sea surface chlorophyll concentration is available. Ratio of CZCS' band-1 (443nm) is used for open ocean in which the chlorophyll concentration is between 0.03-1.50mg/m3. And the ratio of band-2 (520nm) and band-3 for coastal zone in which the chlorophyll concentration is higher than 1.5mg/m3. (4) (5)

    FY-1 satellite is a polar orbiting satellite which was designed and launched by China. Its main mission is observing meteorological phenomena from space. Besides this, a mission of remote sensing of ocean color experiment was carried out. (6)
FY-1 Data Processing
  1. FY-1 Satellite System and Senior.
    FY-1 satellite is in a relatively low orbit around the Earth. Its altitude is 900 km approximately. Table 1 lists the main orbital parameters. A main sensor aboard the FY-1 is a Very-High Resolution Scan Radiometer (VHRSR) It has 5 spectral channels, including 2 ocean channels in the blue-green region. Table 2 lists the wavelength region and the use of VHRSR.

    Table. 1 FY-1 satellite orbital parameters
    Altitude h=901 km
    Orbital Inclination i=99°
    Orbital Period T=102.86 min
    Eccentricity e=0.005


    Table. 2 VHRSR's vavelength region and the use
    Channel Wavelength Region (µm) Use
    1 0.58-0.68 Daytime cloud and surface
    2 0.725-1.1 Daytime cloud image and water,ice,snow and vegetation observation
    3 0.48-0.53 Ocean colour image
    4 0.53-0.58 Ocean colour image
    5 10.5-12.5 Diurnal cloud image, surface observation and sea surface temperature

  2. Defintion of Specfral Equivalent Albedo
    The VHRSR four visible channels data are related to the spectral equivalent albedoes. Generally, the spectral equivalent albedo is defined as a ratio of the object upwelling irradiance and the downwelling solar irradiance of the top of the atmospheres. (7) In fact, most of the objects on the earth surface do not have an ability of the visible emission. They only reflect or scatter sun light. So that, we have to assume that each object on the earth surface is an equivalent source of visible mission. According to above definition, the spectral equivalent albedo As can be written as:

    As = Es / Eo
    ---------------(1)

    Where Es is the upwelling object irradiance reached to sensor. Eo is the downwelling solar irradiance on the top of the atmosphere. in our study, the Es is divided into two components: (1) the contribution from the scattering of solr irradiance by the atmospheric constituents, and (2) the contribution form the irradiance which is diffusely reflected by the object on the earth's surface. The contribution from (1) is usually referred to as the path irradiance, which will be denoted by Ea. Let Ew denote the contribution from (2) at the earth's surface, then the irradiance reaching the sensor is Ew x Ta, where Ta is the diffuse transmittance as given by Gordon et al. (8) Therefore

    Es = Ew * Ta + Ea
    ---------------(2)

    Then, Equation (1) can be rewritten as:

    As = Ew / Eo * Ta + Ea / Eo = Aw * Ta + A a
    ---------------(3)

    Where Aw = Ew / Eo and Aa = Ea / Eo are called the spectral equivalent abed of the object and atmosphere, respectively.

  3. Estimate the Atmospheric Equivalent Albedoes of Two Ocean Channels
    The percentage spectral equivalent albedo As, which can be calculated from the VHRSR visible channel digital number DN has the form

    As = G * DN + I
    ---------------(4)

    Where G is the perecntage spectreal albed per count and I is the percentage intercept albedo.

    For remote sensing ocean colour experiment, the sea-surface equivalent albedo can be estimated from equation (3) if the transmittance and the equivalent albedo of atmosphere are known. Unfortunately, we could not directly obtain the transmittance and the equivalent albedoes of atmosphere for the visible channels. But we well known that the infrared radiance was almost absorbed by water body, especially, by the open seawater. We shall assume that determined equivalent albedo by channel 2 on the orbit equate to its atmospheric equivalent albedo approximately, i.e

    As (2) = Aa (2)
    ---------------(5)

    On the other hand, according to Duntly's method (9) of measuring atmospheric parameter from ground, we measured the atmospheric transmittance and path radiances using a four visible channels radiometer in which the center wavelength of each channel is the same as VHRSR's. A relationship of atmospheric path radiances between channel 2 and other channel is showed as.

    La (1) / La(2) = (l1 / l2)B
    ---------------(6)

    Where l1 is the centre wavelength of two channels and B=3.1665. Using above result and La = Ea/p , As = Ea/Eo, the atmospheric equivalent albedoes of two ocean channels could be estimated.

    Aa(3) = As(2) * Eo(2) / E0(3) * ( l3 / l2 )B

    and

    Aa(4) = As(2) * ( Eo(2) / Eo(4) ) * ( l4 / l2 )B
    ---------------(7)

    Where Eo(2), Eo(3), Eo(4) are the solar irradiances on the top of the atmosphere of VHRSR's channe 1,2,3,4 respectively. Its unit is mw/cm2 their values are given as following:


    Eq....8

    Where Eo (l) is the recommended value of the average solar irradiance on the top of atmosphere. (10) Dl is the wavelength width band of each channel. f(l) is the relative response function of the VHRSR. Fig. 1 shows variation of the wavelength of four visible channels. (11)

    Fig. 1 Relative spectral response curves of VHRSR

  4. Retrieving of the Sea-surface Chlorophyll a Concentration
    Based on the option properties of chlorophyll a in the seawater, the chlorophyll a has a strong absorbance at blue (443nm) and a maximum reflectance at green (550nm). In the ocean color remote sensing experiment, usually the ratio of blue and green reflectance is developed to estimate the sea-surface chlorophyll a concentration.

    From equation (3) and (7), the sea-surface equivalent albedoes of ocean channels can be easily obtained.

    Aw(i) = (As (i)) / Ta(i) = (As(i) - As(2) * Eo(1) / o(2) * (l1 / l2)B / Ta(i)
    ---------------(9)

    (I = 3.4)

    and the ratio two ocean channel's equivalent albedoes is as follows

    R = Aw(3) / Aw(4) = As(3) - As(2) * (Eo(2) / Eo(3)) * ( l3 / l2 )B / As(4) - As(2) * (Eo(2) / Eo(4)) * ( l4 / l2 )B * Ta(4) / Ta(3)
    ---------------(10)


    Because of the centre wavelength of channels 3 is very close to that of channel,4 we shall assume that ratio of two average atmospheric transmittances equate to 1. Equation (10) could be rewritten as

    R = As(3) - As(2) * (Eo(2) / Eo(3)) * ( l3 / l2 )B / As(3) - As(2) * (Eo(2) / Eo(4)) * ( l4 / l2 )B
    ---------------(11)


    Then, the sea-surface chlorophyll a concentaion can be retrieved form a power function of sea-surface equivalent albedoes. Generally, we used the following formula:

    Nc = a * P-b
    ---------------(12)

    In our calculation, a and b are taken 3.3336 and 5.2316 respectively, which are obtained form solving eq. (12) with two known chlorophyll a concentration data only.
Distribution of sea-surface chlorophyll a concentration at Bohai Bay
Fig. 2 is a four visible channel rude image which was obtaind form FY-1 satellite at its orbit on Sept. 21, 1988. the distribution of sea surface chlorophyll a concentration retrieved by above method is showed in Figure. 3 The grey shades of the image will correspond to definite levels of chlorophyll a concentration as indicated by the grey wedge shown on top of Fig. 3. It is obvious that the concentration is lower than 0.6 mg/M3 between the peninsulas of Liao Dong and Shan Dong, as well as at middle of the Bohai Bay. Generally, it is about 2.5 mg/m along the coastal zone. There is a higher concentration patch at the Laizhou bay. Near the river estuary, such as Yellow River, Haihe and Liaohe River's, the concentration is the highest, usually, over 5.0 mg/m3. That is because those rivers discharge a large of nutrient materials into the Bohai Bay, and make the phytoplankton rich and blooming.


Fig. 2 FY-1's rude images of four visible channels


Fig. 3 The distribution of sea surface chlorophyll a concentration at Bohai Bay

Discussion
FY-1 satellite is the first meteorological satellite launched by China. It was designed and made by the Ministry of Aeronautics and Astronautics and is operated and managed under the State Meteorological Administration (SMA). There is a very High Resolution Scan Radiometer (VHRSR) aboard the satellite, which is very similar to NOAA's AVHRR but the wavelength design. There are four visible channels on the VHRSR sensor and one thermal infrared channel for measuring the surface temperature. There is not internal radiance calibration on both satellites. The spectral equivalent albedoes can be calculated from the visible channel digital numbers.

In our ocean colour experiment, we calculated the sea surface chlorophyll a concentration using the spectral equivalent albedoes values. The result is satisfactory, even through the VHRSR's wavelength bands are wider (l=50nm) and the sensitivity is not high enough. At least, following results agree with those of surveying along the East and South China sea.
  1. The chlorophyll a concentration is increased form the middle Bohai Bay to the coastal. The highest appears near the river estuary.

  2. In the most of our coastal zone sea water, the chlorophyll concentration is between 4.0 to 5.0 mg/m3 inshore and is below .5mg/m3 offshore. (12)(13)(14)
Even though, the first FY-1 satelite's life is too short, we have made a good beginning in the ocean color remote sensing. This study has proved that FY-1' data is useful for the ocean applications.

Acknowledgment
The authors express their appreciation to the meteorological satellite centre, S.M.A. for providing FY-1 satellite image. We also acknowledge helpful discussion with Prof. Pan Delu and Prof. Chen Gancheng. We also thank Mr. Shao Ronghu and Ms. Jiang Xiulan for their assistance in supplying figure and typing for the report.

References
  • Clark, D.K., In: J.F.R., Gower(ed.), Oceanography from space, Plenum Press, New York, 1981, p: 227-237
  • Gordon, H.R., and Clark, D.K., Boundary-layer Meteorol.. 18,1980, p: 299-313
  • Sturm, B., In: Arthur. P. Cracknell (ed.), Remote sensing applications in marine science and technology, Plenum press, Oxford, 1984, p: 137-168
  • Sturm, B., In: J.F.R. Gower (ed.), Oceanography from space, Plenum press, New York, 1981, p: 267-279.
  • Sathyendranath, S.and Morel, A., In: Arthur, P. Cracknell (ed.), Remote sensing applications in marine science and technology, Plenum press, Oxford, 1984, p: 323-357.
  • Lin Shouren, Chinese Journal of infrared research, 9,1990. p: 162-177
  • Singh, S.M. and Cracknell, A.P. INT. J. Remote sensing, 7,1986 p: 3610367
  • Gordon, H.R. et al, Appl. Opt. 22,1983
  • Colvocoresses, A.P., Manual of remote sensing, American society of photogrammetry. 1975
  • Grum, F. and Becherer, R.J., Radiometry, Acdemic press, INC. 1979
  • Zheng Qinbo, et al., Chinese Journal of infrared research.9,1990. p:91-98
  • Ning Siuren, et al., Acta Oceanologica Sinica. 7,1985. p: 751-762
  • Cen Xingqun., Acte Oceanologica Sinica. 11,1989
  • Huang Liangmin., Acte Oceanologica Sinica. 11,1989. p: 769-779