Coral reefs change detection using remotely sensed data L a case study in Bintan Beach International Resort, Riau Islands, Indonesia

P

ABSTRACT

e

p p p

n

ZggZ

:::z

s ,I SYAWALUDIN ALISYAHBANA HARAHAP (2006).

Coral Reefs Change
c. 5 Q Detection Using Remotely Sensed Data: A Case Study in Bintan Beach
2 %- %
5 =s t International Resort, Riau Islands, Indonesia. Under the direction of
% 3 c g g VINCENTIUS P. SIREGAR and IDUNG RISDIYANTO.


J Q Q 3
* 3 3 m n

s p g y
Q 9 . ap
~

~ E X S S

Bintan Beach International Resort is located on North of Bintan Island

Islands province), this area is placed on 104°17'50" to 104°38'36"E and
$ P g " s (I@
$ g. s % f r m 1°6'10" to 1°15'29"N, has soft relief with slope 5 - 15%. Mostly, beach
n 3 3 3 =
m
4 a r k is covered by sandy beach with the depth of waters is about 1 - 20 meters.

2


o.5 BS a bi tan

'CQtDQC
2133

Beach International Resort region represent a coastal area that high
a-((I
0
3 0.
r c
* s d$amic change. It has competence in conserve, protect as well as develop this
8 E*g .
art& and it have concern and responsive about ecosystem change. In this research,
5 . 2.
&&oral
reef ecosystem become main target because it very essential matter that
;2 3 g
sh&ld be monitored for planning or to avoid may further change that can damage
K P

ii
ij Z
g 4Q
;6 5 3
or*
the environment.
b5 8
J
S
The Objectives of this research are to detect and mapping coral reefs using
f,
g kg
2 ~telysensed data and image processing technique, further to analysis the coral
$ 5%
E
:
5
change during two periods (1992 and 2003). A two periods of Landsat-5 TM
k GQ
P)

3il 14,1992) and Landsat-7 ETM+ (March 04,2003) images with the path/row
-.
2. g. 9
' 359 was used for coral reef mapping in the Bintan Beach International Resort.
Q
2

""
Q

5

rC

-

B

i.


;I F
$r

i5
w

7<

Y *

2

F
fQ

E

gg
3s


To properly use these data, a minimum set of correction needs to be made:
)metric, geometric and water column correction as well as ground-truthing.
unsupervised classification was used for image classification and postifzcation comparison method was used for change detection.

The result of training area processing was produced attenuation coeficient
: (k,/kj) are 1.526129 for image of 1992 and 0.879188 for image of 2003.
feature classes of coral reef performed in 5 (five) classes were the live coral
, dead coral (DC), dead coral with algal (DCA), sand (S) and macro algal
). The overall accuracy was found in this research is 76.66 %. LC has been
m:asing equal to 16.84% or 271.08 ha where in 1992 the area is 422.87 ha to
9 ha in 2003. Also by DC and DCA, in 1992 the area of DC is 535.70 ha,
while in 2003 it has been decreasing equal to 17.20% or 258.69 ha in 2003,
DCA has been decreasing equal to 15.96% or 256.94 ha from 478.53 ha in
to 221.59 in 2003. The improvement of coral reefs class is viewed only on
and macro algal, even sand has been increasing equal to 48.3 1% or 777.90 ha
630.70 ha in 1992 to 1408.60 ha in 2003, while macro algal has been
:masing equal to 1.69% or 27.13 ha from 145.77 ha in 1992 to 172.90 ha in

-"Fa
3-


words: Coral reefs, change detection and remotely sensed data.

3
-.

<

0

z-.

u"