248 
3. MEASUREMENTS 
3.1 Determination of inherent optical properties 
However, additio 
types are requirec 
shows the calcula 
The IOP of 18 water bodies were determined from spectrophotometric measurements. For each water sample two types 
of measurements were carried out with a Perkin-Elmer UV-VIS 55IS spectrophotometer. First the apparent beam 
coefficient c' (Eq. 2) was measured. Then the sample was placed in an integrating sphere which captured the light in 
the range of 0-40°, i.e: 
= a + b t 
(5) 
In principle the measurements were carried out within hours of the sampling. In some cases concentration of the 
samples was applied following Kirk, 1980. Kirk estimated the loss of particulate material to be ca. 10%. The results 
from this study were corrected by this percentage. In order to extract a, b } and from c' and c' m two assumptions 
were made: that between 700-800 nm the absorption equals the absorption of pure water which gives the ratio bjb i 
and that this ratio is constant from 400 to 800 nm (see Dekker, 1993). The a and b } , averaged over each water type, 
are depicted in Fig. 1. The values of b } integrated over PAR are given in Table 2, ranging from 0.6 to 25 m 1 , 
considerably higher than the values of Petzold (1972). 
The absorption curves are shown in Fig. l.a were calculated by summing the absorption curves for humus, 
tripton, phytoplankton and water (see Dekker, 1993) and by subsequent averaging per water type. At short wavelengths 
humus, tripton and the first chlorophyll a absorption peak cause high total absorption. This is the reason for the low 
reflectance observed in this spectral area (Fig. 2.b). Beyond 500 nm absorption decreases and reflectance increases cor 
respondingly allowing a better discrimination of spectral features in reflectance spectra. The lowest total absorption 
values occur at 550 to 600 nm, at 650 nm and at 700 - 710 nm, conversely coinciding with maxima in co 4 and R(O-) 
(Fig. 2.a & b). The scattering spectra in Fig. l.b are spectra obtained by averaging the scattering spectra per water 
type (Dekker, 1993). No discrimination into scattering by water, phytoplankton and tripton was made. The most 
remarkable feature in Fig.l.b is the slope of the spectral scattering for the shallow eutrophic lake samples. This is 
probably due to the (complex) scattering behaviour of the filamentous prokaryotes abundantly present in these waters. 
Subsequently, B' needs to be estimated. Since no volume scattering function was available the results of Petzold 
(1972) were used; in combination with the measured b } and b m , B' was obtained by 
( 6 ) 
wavelength (nm) 
shallow shallow 
eutrophic mesotrophic 
(a) 
wavelength (nm) 
deep river&canal 
(b) 
3.2. Determinatii 
The underwater ir 
downwelling irrac 
For remotely sen 
converted to irrad 
from the upwellin 
(= 0 . 021 ) of the 
where L 0 ( 0 ) was i 
panel. In case of 
upwelling radianc 
Often the value c 
literature (Dekker 
component ¿^(Oq 
downward reflect 
where rfQj is the 
and <o 4 have a sin 
4. RESULTS 
A regression anal) 
r, are independen 
along with an opti 
at the time of the / 
ve error of 1 % tc 
Whitlock et al. (IS 
of the coefficients 
by a masking effei 
for the coefficient 
characterization i: 
In 15 ca 
approximated by i 
Rijn Kanaal, whic 
Figs. 3.a 
types. The figures 
that a low correlat 
Fig. 1 a) The measured absorption a and b) scattering spectra b } for the four water types