PS =O OO OW
y
la
al
te
to
ite
he
by
er
tal
he
ote
ral
tic
jas
ich
1S-
sity
ind
his
vas
od
(13
the
lds
rest
thin
nce
with
ent
tion
ere
was
and
the
0m
irect
Class Overall Area in
area in 250 m
[26] buffer
in [%]
Water 12.4 not
consider.
Shadows 1.2 3.6
Coniferous forest 29.9 24.6
Mixed forest — predominant 7.4 8.8
conifers
Mixed forest 10.0 16.5
Deciduous forest 4.5 6.7
Clearings within forest 7.4 12.3
Meadows / pastures 1.9 2.0
Cultivated areas - predominant 7.6 8.5
vegetation cover *
Cultivated areas — intermediate 9.8 9.4
vegetation cover *
Cultivated areas - low vegetation 2.0 1.3
cover *
Bare soils 3.1 0.7
Settlements 2.8 5.4
*during the observed period between February and September 1997
Tab. 8: Overall percentage of surface cover types
and percentage in 250 m surroundings of
the lakes
Many lakes are almost completely surrounded by
forest, which is in accordance with forest as the
predominating landuse class within the 250 m buffer
zone (see tab.8). In case of settlements the percentage
within the buffer increases in comparision to the overall
area which reflects the fact that many settlements are
build closely to lakes. Further analysis will focus on
lakes with predominant agricultural landuse and high
percentage of settlements within the buffer zone since
these are indicators for high anthropogenic impact on
the lake ecosystem.
Additional aspects which have not been studied yet are
interconnection between lakes, ground water flow as
well as substrats forming lake bottoms and soils. Such
information have to be incorporated into the GIS from
thematic maps and other sources.
5.3 Possible relationships
An example of possible relationships between lake
water properties and influencing factors is shown for
one way of grouping the lakes. In this case the
determining factors are water depth and position of the
lakes in the drainage system which led to the following
groups :
a) extreme shallow lakes without surface tributaries
b) lakes without surface tributaries or situated
upstream at the beginning of a chain of
interconnected lakes
C) other interconnected lakes within the upstream part
of the watershed of Müritz-Havel channel system
(northwestern part of study area)
d) other lakes within the downstream part of the
watershed of river Rhin (central and southern parts
of study area)
e) other lakes within the watershed of river Havel
(northeastern part of study area).
About two thirds of the total 53 lakes fall into these
groups. The other lakes are not considered in this
analysis due to missing information.
These groups were investigated in their relationship to
the differentiation of lake water properties which was
obtained by multitemporal analysis of satellite data (see
fig. 7). For this purpose a statistical analysis of the first
principal component values was performed for each
group (see tab.9).
Table 9 shows a significant lower variation of the first
PC values within each group in comparison with the
results of the first PC for all lakes. These results
indicate that the grouping based on main influencing
factors is also reflected in the multitemporal remote
sensing data. The largest deviation from the overall
mean can be observed for group a) and b) which
include single lakes of varying depth and extreme
shallow lakes. Group a) represents lakes of highest
bioproduction whereas group b) includes lakes of
lowest bioproduction.
1.PC GROUP | GROUP | GROUP | GROUP | GROUP | Total
a) b) c) d) e)
(3) (13) (6) (7) (5)
Mean 543 16,1 271] 23.21 219 230
Min 49,0 9,81 21,51 21,2]. 20,2 9,1
Max 61,2 23,21 207 254 24,9 61,7
Stdev 6,5 3,8 3,2 1,3 1,7 9,5
Tab. 9: Summary statistics of first principal
component calculated from Means
(IRS1,IRS2) of all datasets for different lake
groups
Group c) through e) show small deviations from the
overall mean indicating that these groups represent the
dominating lake type which is characterized by medium
bioproduction. For each of these individual groups
standard deviation is low which means that they
contain lakes of similar seasonal behavior.
Lake Plátlinsee as one of the large lakes in the study
area (see fig. 2) allows a differentiation of lake water
properties within its water body using multitemporal
satellite data. A significant differentiation between the
northern and southern part of the lake can be
observed based on the calculated indices (see 4). This
differentiation can be related to a higher chlorophyll-a
content during the whole season of bioproduction
within the northern part which is much shallower (depth
of 4-5 m) than the southern part (more than 20 m).
6. CONCLUSIONS AND OUTLOOK
The relationship between ground measurements and
spectral properties of lakes allows to derive relative
chlorophyll-a content. It can be summarized that the
relative changes of chlorophyll-a within lakes can be
well observed and analyzed by multitemporal IRS-1C
satelite images. The spatial variation of relative
differences in chlorophyll-a content can be mapped for .
each date. The relative temporal changes of
bioproduction during the period between August 1996
to September 1997 reflect typical phenomena of algae
development , such as algal blooms which are known
for the study area from field observations.
Intemational Archives of Photogrammetry and Remote Sensing. Vol. XXXII, Part 7, Budapest, 1998 135
