The Lower Euphrates Project is one of the seven GAP
(Southeastern Anatolia Project) sub-projects on the Euphrates
River. It consists of the most important schemes of GAP like
the Atatürk Dam and Hydroelectric Power Plant (HEPP),
Sanliurfa Tunnels, Sanliurfa -Harran irrigation, Mardin-
Ceylanpinar irrigation, Siverek-Hilvan pumped irrigation and
Bozova pumped irrigation (Fig 1).
The Atatürk dam has a great contribution on the agricultural and
industrial development in the GAP Region. The irrigation
system of Lower Euphrates Project is based on the Atatürk dam
reservoir. The construction of the dam was initiated in 1983 and
the diversion of water through the tunnels started in June 1986.
The reservoir was filled in August 1990 and the power
generation was started in 1992.
= sut
tts
E
e .
cM MCI,
Qmm "X.
An
e: ze
^il NA Hasena Bucher ox es e
torri: xar ET > Ts
de
2 Fu
M Bose n
po i etapa
Ve Arf Ear N à 3* tege Seteoret
a X ker
Fig. 1 Location of planned irrigation schemes of GAP. (Striped
areas will be irrigated from Atatürk Reservoir; others will be
irrigated by other sources) (SPO, 1989).
The Landsat image archive represents an opportunity to assess
agricultural monitoring over time through time-series analysis.
Landsat, with a spatial resolution of 30 m and spatial extent of
185x185 km per scene, is used widely for mapping biophysical
vegetation parameters (Cohen & Goward, 2004) and has proven
useful for monitoring land cover (Wulder et al, 2008) and
ecosystem disturbance (Healey et al., 2005; Masek et al., 2006).
Remote sensing is a critical data source for observing and
understanding the effects of landscape disturbance (e.g. Potter et
al., 2003; Linke et al., 2008; Masek et al., 2008). Landsat-based
detection of disturbances (e.g. Cohen et al., 2002; Franklin et
al, 2001; Seto et al, 2002) commonly use image
transformations such as the Tasseled Cap transformation (Crist
& Cicone, 1984; Kauth & Thomas, 1976) to consolidate
multispectral reflectance measurements and enhance the
detection of disturbance events.
The Tasseled Cap transformation reduces the Landsat
reflectance bands to three orthogonal indices called brightness,
greenness and wetness, and is a standard technique for
describing the three major axes of spectral variation across the
solar reflective spectrum measured by Landsat (Kauth &
Thomas, 1976). Once an image is transformed into its Tasseled
Cap data space, image arithmetic and thresholding techniques
can be used to automatically identify and classify land cover
changes and land cover disturbance (e.g. Cohen et al., 2002;
Franklin et al., 2001; Healey et al., 2005).
Up-to-date and objective information on the spatial distribution
of irrigated crops as well as changes in their areal extent over
time can help achieve the goal of efficient water resource
management. The objective is to map the extent of irrigated
agricultural fields and assess land-cover trends at the object
level within a selected study area in GAP region using Landsat
images from 1984 to 2011.
Ozdogan (2004) indicates that in Southeastern Turkey, the best
time to distinguish irrigated lands from other land cover types is
mid-to-late summer. In the study, Landsat imagery for these
months were used in order to determine the land cover/use
changes especially on agricultural fields under the Atatürk Dam
Lake and also irrigated agricultural fields with water reserve
changes for ~30 years.
2. STUDY AREA
The first focus of this study is the Atatürk Dam Lake with a
catchment area of 92,240 km? while the impounded land is 817
km?. The maximum reservoir capacity of the Atatürk dam is
48.7 x10? cubic meters with the total irrigation area of 8,724
km“. The minimum level of water in the reservoir is 526 m. The
second focus is the Harran Plain, which is irrigated from the
Atatürk Dam Lake and is located in the south-central part of the
GAP project within the Sanliurfa - Harran Irrigation District.
The Plain is 2756 km? and is located in a region of rolling hills
and a broad plateau that extends south into Syria (Fig 2). Two
meteorological stations that have been used in the study are
shown as red stars in Fig. 2.
387350"E Won "E
a M
z
:
5
38°0
31930'0"N
STSFIFOUN
IHN
38°30 39°0'0"E
Fig. 2 Landsat TM images of the study area (August 201 1).
Inset shows mean annual temperature of Turkey with the study
area outlined by a black rectangle.
The study area has the major climatic features of the Eastern
Mediterranean with a strong continental influence. The annual
averag
mm. 1
most |
area r
irrigat
The S
scherr
broug
syster
with
m° /se
in 19
are al
Sever
veget
variet
any 8
Rem
irriga
locati
requi
chan;
and 1
2001
Mon
irrige
data.
agric
have
Rese
ident
imag
Afte
reser
anal
Lake
last
seas
chan
on t
appt
beer
dete
clim
incli
tran:
refle
data
Tas:
Lan
(Cri
coe;
to E
to v
Lan
into
ET!
bet
Vo