Sundarbans (Bangladesh and India)
From the 1970se2000s, mangrove forest in the Sundarbans
decreased by 1.2%. The rate of change, however, was not uniform
from the 1970se1990s and from 1990s to 2000s. From the 1970se
1990s, mangrove forest area increased by 1.4%, and from 1990s to
2000s, the area decreased by 2.5%. These changes are nonsignificant
in the context of errors associated with classification
and the dynamic nature of mangrove ecosystems. In other words,
these changes are well within the margin of error. For example,
because of the fluctuation of tide, certain areas in flooded areas,
barren lands, and water bodies could be misclassified from one
class to another. Small changes less than 3x3 pixels were not
detected from this study as this was the minimum mapping unit
used. This is expected to minimize the errors arising from missregistration
of satellite imagery.
While the measured net loss of mangrove forest was not
considerable, the change matrix (Table 5) shows that turnover was
much greater than net change. For example, 7% of the 1970s-era mangrove forest had changed to non-mangrove, flooded, water
bodies, or barren lands by 2000. The largest category of mangrove
forest change was loss to flooded (4.6%). The change matrix also
revealed that during the same period approximately 37% of flooded
areas, 21% of barren lands, 8.3% of non-mangrove, and 2.2% of water
bodies were converted to mangroves. Similar patterns of change
were observed from the 1970se1990s and from 1990s to 2000s
(Table 5).
In all three classifications, 93e95% of mangrove forests, 93e96%
of water bodies, and 69e79% of non-mangrove areas did not
change. During the same period, the turnover for flooded areas and
barren lands was, however, quite high, only 30e35% of flooded and
15e50% of barren lands remain unchanged. The large change between
flooded and barren lands may possibly be due to variation in
tidal inundation at the time of satellite data acquisition. Major
change areas were concentrated either in the outer periphery or
near the shoreline (Fig. 4), caused by anthropogenic and natural
forces, respectively.
The high turnover between mangrove and non-mangrove is due
primarily to encroachment, erosion, aggradation, and mangrove
rehabilitation programs. The rate of erosion is highest at the
southern edges of Mayadwip, Bulcherry Island, and Bhangaduni
Island. For example, Bhangaduni Island lost one-fourth of its land
area (25.1%) and just less than one-fourth of its mangrove area to
erosion between the 1970s and 2000s. The majority of this loss in
this island occurred between 1989 and 2000s, which is evident
from the following illustrations (Fig. 5).
Due to aggradation, land continues to be made afresh in the
Sundarbans, offsetting a large part of the loss to erosion. This
process has increased the land and mangrove forest areas. Once the
new land is formed, such lands are typically colonized by a
sequence of plant communities, culminating in the establishment
of mangrove forests. Examples of aggradation can be seen in Fig. 5.
Between 1970s and 1990s, mangrove forest gained from
aggradation (2925 ha) nearly equals mangrove forest lost to erosion
(3157 ha). From the 1990se2000s, however, the rate of erosion
claimed seven times as much mangrove forest (4151 ha) as aggradation
created (592 ha). Erosion was concentrated along the banks
of major river channels and at the landewater interface with the
Bay of Bengal. Approximately half of the mangrove forested land
lost was at the extreme southern edge of the Sundarbans where
almost no compensating aggradation took place.
On the India side of the Sundarbans, the most dramatic area of
change is located approximately 14 km east of Kisoripur. In the
1970s image, 1085 ha of mangrove forest, interspersed with open
flooded areas, extended approximately 4 kminland from the Matla/
Bidya River. By 1990s, the classification shows that 13.27% of the
mangrove forest had been lost, and the boundary between development and mangroves had receded approximately 1 km to
the east. By 2000s (ETMþ),
Only a ring of mangrove at the shoreline remained. The evidence
of development is apparent with the building of diked areas and
canals as the forest was removed. This area falls outside of the
managed forest reserves and contrasts sharply with the mangrove
forested areas to the south and east, which remained generally
unchanged during the same period.
Again, the net mangrove loss over thewhole of the Sundarbans is
about 1% as the numerous areas of loss are counter-balanced by
areas of gain. Most of this gain is found in areas where new land
formed through deposition has become vegetated. One of the exceptions
is an area of afforestation located in the Jilla forest block on
the northern forest boundary of the India side. This area of
approximately 400 ha was completely degraded in 1975, but had
been re-vegetated by 1989 andwas generally indistinguishable from
surrounding forested areas in a remote-sensing image by 2000s.
Overall accuracy of 86%, 85%, and 79% were achieved for
2000s, 1990s, and 1970s classification with the Tau coefficient of
0.85, 0.83, and 0.76, respectively. The tau coefficient for the year
2000, for example, indicates that our classification systems produce
a map on which 85% more pixels were classified correctly
than would be expected by random assignment. This means that
for this classification, we were correct 85% of the time. Confusion
arose in discriminating flooded and water bodies, and nonmangrove
and barren lands classes. Mangrove class was relatively
well classified.
The canopy closure layers derived from NDVI measurements for
the three mosaics showchanging patterns of forest condition in the
Sundarbans. The pattern of healthy upper-story vegetation is
different in the different era classification results. Therefore, the
least healthy areas in 2000s are different from the least healthy
areas of 1990s. Furthermore, the pattern of relatively unhealthy vegetation in 2000s corresponds to areas of reported top dying. As
explained above, the lack of multiple images for each era, the
different seasons of acquisition for images of different eras, and
variation in the degree of tidal inundation in the various images
prevents comparison of absolute values derived from each of the
canopy closure layers. While the absolute values for canopy closure
that the model is designed to generate are not reliable, patterns of
relative canopy closure are confirmed as generally valid. Visual
confirmation of the validity of the canopy closure layer comes from
two sources: the 1985 (1983 data) Chaffey et al. inventory maps and
QuickBird high-resolution remote-sensing images from 2002. The
Chaffey et al. (1985) maps from 1983 aerial photography, while
compiled approximately 6 years later, support the validity of the
1970s-era canopy closure layer. The 1983 maps show roughly twothirds
of this area as having canopy closure above 70% and little or
none of this area to be below 30% canopy coverage. These areas
correspond well to the high and low canopy closure areas in the
1970s-era canopy closure layer. The largest change in the pattern of
canopy closure is between the TM and ETMþ eras, when a large
corridor of reduced canopy closure appears between the Bal and
Sibsa Rivers. This corresponds to forest compartments that have
high rates of top dying
Sundarbans (Bangladesh and India)From the 1970se2000s, mangrove forest in the Sundarbansdecreased by 1.2%. The rate of change, however, was not uniformfrom the 1970se1990s and from 1990s to 2000s. From the 1970se1990s, mangrove forest area increased by 1.4%, and from 1990s to2000s, the area decreased by 2.5%. These changes are nonsignificantin the context of errors associated with classificationand the dynamic nature of mangrove ecosystems. In other words,these changes are well within the margin of error. For example,because of the fluctuation of tide, certain areas in flooded areas,barren lands, and water bodies could be misclassified from oneclass to another. Small changes less than 3x3 pixels were notdetected from this study as this was the minimum mapping unitused. This is expected to minimize the errors arising from missregistrationof satellite imagery.While the measured net loss of mangrove forest was notconsiderable, the change matrix (Table 5) shows that turnover wasmuch greater than net change. For example, 7% of the 1970s-era mangrove forest had changed to non-mangrove, flooded, waterbodies, or barren lands by 2000. The largest category of mangroveforest change was loss to flooded (4.6%). The change matrix alsorevealed that during the same period approximately 37% of floodedareas, 21% of barren lands, 8.3% of non-mangrove, and 2.2% of waterbodies were converted to mangroves. Similar patterns of changewere observed from the 1970se1990s and from 1990s to 2000s
(Table 5).
In all three classifications, 93e95% of mangrove forests, 93e96%
of water bodies, and 69e79% of non-mangrove areas did not
change. During the same period, the turnover for flooded areas and
barren lands was, however, quite high, only 30e35% of flooded and
15e50% of barren lands remain unchanged. The large change between
flooded and barren lands may possibly be due to variation in
tidal inundation at the time of satellite data acquisition. Major
change areas were concentrated either in the outer periphery or
near the shoreline (Fig. 4), caused by anthropogenic and natural
forces, respectively.
The high turnover between mangrove and non-mangrove is due
primarily to encroachment, erosion, aggradation, and mangrove
rehabilitation programs. The rate of erosion is highest at the
southern edges of Mayadwip, Bulcherry Island, and Bhangaduni
Island. For example, Bhangaduni Island lost one-fourth of its land
area (25.1%) and just less than one-fourth of its mangrove area to
erosion between the 1970s and 2000s. The majority of this loss in
this island occurred between 1989 and 2000s, which is evident
from the following illustrations (Fig. 5).
Due to aggradation, land continues to be made afresh in the
Sundarbans, offsetting a large part of the loss to erosion. This
process has increased the land and mangrove forest areas. Once the
new land is formed, such lands are typically colonized by a
sequence of plant communities, culminating in the establishment
of mangrove forests. Examples of aggradation can be seen in Fig. 5.
Between 1970s and 1990s, mangrove forest gained from
aggradation (2925 ha) nearly equals mangrove forest lost to erosion
(3157 ha). From the 1990se2000s, however, the rate of erosion
claimed seven times as much mangrove forest (4151 ha) as aggradation
created (592 ha). Erosion was concentrated along the banks
of major river channels and at the landewater interface with the
Bay of Bengal. Approximately half of the mangrove forested land
lost was at the extreme southern edge of the Sundarbans where
almost no compensating aggradation took place.
On the India side of the Sundarbans, the most dramatic area of
change is located approximately 14 km east of Kisoripur. In the
1970s image, 1085 ha of mangrove forest, interspersed with open
flooded areas, extended approximately 4 kminland from the Matla/
Bidya River. By 1990s, the classification shows that 13.27% of the
mangrove forest had been lost, and the boundary between development and mangroves had receded approximately 1 km to
the east. By 2000s (ETMþ),
Only a ring of mangrove at the shoreline remained. The evidence
of development is apparent with the building of diked areas and
canals as the forest was removed. This area falls outside of the
managed forest reserves and contrasts sharply with the mangrove
forested areas to the south and east, which remained generally
unchanged during the same period.
Again, the net mangrove loss over thewhole of the Sundarbans is
about 1% as the numerous areas of loss are counter-balanced by
areas of gain. Most of this gain is found in areas where new land
formed through deposition has become vegetated. One of the exceptions
is an area of afforestation located in the Jilla forest block on
the northern forest boundary of the India side. This area of
approximately 400 ha was completely degraded in 1975, but had
been re-vegetated by 1989 andwas generally indistinguishable from
surrounding forested areas in a remote-sensing image by 2000s.
Overall accuracy of 86%, 85%, and 79% were achieved for
2000s, 1990s, and 1970s classification with the Tau coefficient of
0.85, 0.83, and 0.76, respectively. The tau coefficient for the year
2000, for example, indicates that our classification systems produce
a map on which 85% more pixels were classified correctly
than would be expected by random assignment. This means that
for this classification, we were correct 85% of the time. Confusion
arose in discriminating flooded and water bodies, and nonmangrove
and barren lands classes. Mangrove class was relatively
well classified.
The canopy closure layers derived from NDVI measurements for
the three mosaics showchanging patterns of forest condition in the
Sundarbans. The pattern of healthy upper-story vegetation is
different in the different era classification results. Therefore, the
least healthy areas in 2000s are different from the least healthy
areas of 1990s. Furthermore, the pattern of relatively unhealthy vegetation in 2000s corresponds to areas of reported top dying. As
explained above, the lack of multiple images for each era, the
different seasons of acquisition for images of different eras, and
variation in the degree of tidal inundation in the various images
prevents comparison of absolute values derived from each of the
canopy closure layers. While the absolute values for canopy closure
that the model is designed to generate are not reliable, patterns of
relative canopy closure are confirmed as generally valid. Visual
confirmation of the validity of the canopy closure layer comes from
two sources: the 1985 (1983 data) Chaffey et al. inventory maps and
QuickBird high-resolution remote-sensing images from 2002. The
Chaffey et al. (1985) maps from 1983 aerial photography, while
compiled approximately 6 years later, support the validity of the
1970s-era canopy closure layer. The 1983 maps show roughly twothirds
of this area as having canopy closure above 70% and little or
none of this area to be below 30% canopy coverage. These areas
correspond well to the high and low canopy closure areas in the
1970s-era canopy closure layer. The largest change in the pattern of
canopy closure is between the TM and ETMþ eras, when a large
corridor of reduced canopy closure appears between the Bal and
Sibsa Rivers. This corresponds to forest compartments that have
high rates of top dying
การแปล กรุณารอสักครู่..
