777
REFERENCES
L. is the deep water radiance, k,- are positive or
1W * 1
negative constants determined from the spectral
signatures of the water-bodies concerned. Obviously,
knowledge on the local conditions, like turbidity
of the water and bottom features will increase the
accuracy of the mapping.
Examples of the dependence of the algorithms on
the bottom depth and composition for both clear
and turbid water regions are shown in Figures 4 and
5.
a z2
(rel.units)
"650 700
wavelength (nm)
il water with
Lcated.
for the
:her in-
:ability of
>ptimal
sitive to the
i composition,
lould be
:e including
>f the
ind the deep-
irs to be
>th dependence
roposed from
sssed as
Figure 4. Bottom depth algorithm as function of
the bottom depth for sandy (full lines), muddy
(dashed lines ) and vegetation (dashed lines
) bottom types. The constants kj = 1, k2 = 0.5.
Longer lines are representative for clear waters,
shorter lines belong to turbid water. Limit values
of the detectable depths are indicated.
A b3
(rel.units)
Figure 5. Bottom composition algorithm as function
of the bottom reflectance for mixed sand-vegetation
bottom types, S = 0 for fully overgrown bottom,
S = 1 for pure sandy bottom. Lower lines are
representative for clear waters with bottom depths
as indicated in metres, upper line belongs to shallow
turbid water region, k^ =-1.05, - 1.
the channel i,
Final tuning of the algorithms, based on the local
seatruth data should be done for each area to be
mapped. A case study for North Sea coastal areas,
using Landsat TM data, will be performed.
Lyzenga, D.R. 1978. Passive remote sensing techniques
for mapping water depth and bottom features. Appl.
Optics. 17: 379-383.
Prieur, L. & S. Sathyendranath 1981. An optical
classification of coastal and oceanic water based
on the specific spectral absorption curves of
phytoplankton pigments, dissolved organic matter,
and other particulate materials. Limnol. Oceanogr.
26: 671-689.
Robinson, I.S. 1985. Satellite Oceanography.
Chichester: Ellis Horwood.
