The purpose of performance assessment is to achieve efficient, productive and effective irrigation and drainage systems by providing relevant feedback to management at all levels. As such, it may assist management or policy makers in determining whether performance is satisfactory and, if not, which corrective actions need to be taken in order to remedy the situation (Bos et al., 2005).
In many countries, accurate evaluation of irrigation system performance and sustainability is hampered by lack of adequate, reliable and timely irrigation statistics. Usually, performance indicators such as yield, cropping intensity and irrigation intensity are measured at an aggregated level, often at the state or national levels. Data at project level are rarely collected. If collected, they are frequently unreliable or not easily accessible (Murray-Rust and Merrey, 1994).
Satellite remote sensing provides the opportunity to capture the information of a large area at frequent intervals. The capability of satellite remote sensing to monitor agricultural and hydrological conditions of the land surface has undergone major improvements in the present decade. The following areas of irrigation water management using remote sensing techniques have been addressed: identification of irrigated crops, storage in reservoirs, irrigated crop areas, crop condition and yield, crop water requirement, soil moisture, water logging and soil salinity irrigation system performance (Ambast et al., 2002), strategic planning and water rights, water use and productivity, impact assessment, precipitation, runoff, crop coefficient and biomass (Karatas et al., 2003).
Crop conditions depend on the amount of soil moisture available to the crop during the growing season. The source for the soil moisture is either precipitation or irrigation. When crops are irrigated and their soil moisture requirements are completely met, the spatial distribution of their health conditions is rather uniform. Various indices have been developed using AVHRR data to monitor crop or vegetation conditions over large areas. Out of these indices, the Normalized Difference Vegetation Index (NDVI) has been widely used for vegetation monitoring (Boken et al., 2004).
In Turkey, data relating to Water User Associations (WUAs) such as crop pattern, irrigation ratio and water supplied to the system at main canal level are kept by the General Directorate of State Hydraulic Works (DSI), but it is impossible to find information on water delivery within the system. In this study, water delivery performance of four selected WUAs in the Lower Gediz Irrigation System, located in the Gediz Basin, one of the countrys important catchment areas, has been assessed according to the indicators of adequacy and equity.
MATERIALS AND METHODS
Material: The Gediz Basin, located in the west of Turkey, covers an area of 17200 km2, with 140000 ha under irrigation. Altitude ranges from sea level to 2300 m. Annual precipitation ranges from 1000 mm on the mountains to 500 mm in the coastal areas and the area has a Mediterranean climate.
Sources of water for irrigation in the basin are the multi-purpose Demirköprü Reservoir and Gölmarmara Lake, along with the Afşar and Buldan irrigation reservoirs. Principal crops are, in order of importance, cotton, grapes, cereals, other fruit and vegetables, olives, tobacco and melons. The natural vegetation is mainly composed of brush and forest (Droogers et al., 2000).
Operation of the Gediz Basin irrigation system was transferred from the DSI to 13 WUAs in 1995. The present study was carried out on the irrigation areas of four selected WUAs of the Lower Gediz Irrigation System. These were, from upstream to downstream, the Salihli Left and Right Bank WUAs at the head, the Gediz WUA in the middle and the Menemen Left Bank WUA at the end.
Irrigation water in the Lower Gediz Irrigation System is delivered from the Demirköprü Reservoir via the Adala, Ahmetli and Emiralem regulators and six main canals connected to them. In months of intensive irrigation, Gölmarmara Lake is brought into use (Fig. 1).
The study has made use of DSI records of planned delivery and actual flow records
along with low-resolution (1.1 km at nadir) NOAA-16/AVHRR images. These satellite
images were obtained from http://www.class.noaa.gov.
Methods: In assessing the water delivery performance of the WUAs regarding adequacy, irrigation ratio (IR), water use ratio (WUR) (Jurriens, 1996) and average NDVI (NDVIave) were used, while equity was assessed using the variation coefficients of NDVI, which showed whether water delivery was uniform in the area served by the WUAs (Sakthivadivel et al., 2001).
IR and WUR were calculated using the following equations:
The normalized difference vegetation index (NDVI) is a measure of the amount
and vigor of vegetation at the surface. It is formulated as follows:
NDVI = (RNIR-Rred) / (RNIR+Rred)
|| Location of the WUAs in the Gediz Basin
Where R is the reflectance value in the indicated band. Index values can range from -1.0 to 1.0. The magnitude of NDVI is related to the level of photosynthetic activity in the observed vegetation. In general, higher values of NDVI indicate greater vigor and amounts of vegetation. The reason NDVI is related to vegetation is that healthy vegetation reflects very well in the near infrared part of the spectrum (Parodi, 2002). As the leaves come under water stress, become diseased or die back, they become more yellow and reflect significantly less in the near infrared range.
In calculating NDVIave and CVNDVI values, an equal number of pixels (35 pixels) forming a rectangle was determined within the irrigation area of each WUA on the NDVI maps which were drawn. NDVIave and CVNDVI values were calculated with reference to the NDVI values of these pixels. NOAA-16/AVHRR images used in the study had undergone pre-processing in order to correct the geometric distortions, calibrate the data radiometrically and eliminate the noise and clouds that present in the data (Gautam et al., 2006). WinCHIPS software was used for this process and NDVI maps were drawn using ILWIS software.
RESULTS AND DISCUSSION
Adequacy of water delivery: The adequacy of delivery of water to the irrigated areas of the four WUAs in the Lower Gediz Irrigation system is evaluated below according to the indicators of IR, WUR and NDVIaverage.
Irrigation ratio (IR): It can be seen from Table 1
that IR values varied from 48 to 89%, with averages of 77% in 2004 and 76% in
2005. The lowest IR values for each year came from Gediz WUA. The main factor
causing these low values is that the Gediz WUA is at the end of a main canal
which also serves three other WUAs and thus it gets insufficient water. Apart
from this WUA, there is no important difference in irrigation ratios between
WUAs according to their distance from the source.
Water use ratio (WUR): The WUR values of the selected WUAs show variations
between the two years and from month to month, but seasonal averages were generally
close to 1 (Table 2).
|| IR values (%) for the irrigation seasons 2004 and 2005
If WUR value was equal to 1, this showed that the amount of water delivered
per unit of irrigated area was same to the planned amount; a value less than
1 shows that less water was delivered than planned, while a value greater than
1 indicates that more water was delivered than planned (Jurriens, 1996). Thus
it can be seen that in 2004 the amount of water delivered per unit of irrigated
area in the selected WUAs was closer to the planned amount than in 2005, but
in both years water was not delivered homogeneously month by month over the
season. Water was delivered to the system in excess of the plan in both years,
particularly in August. This can be seen especially clearly in the case of the
Salihli Right Bank and Salihli Left Bank WUAs, which are close to the source.
When monthly WURs are compared, it can be seen that 2004 was a more successful
year from the point of view of adequacy.
Average of NDVI (NDVIave): NDVIave values were calculated for the four selected WUAs for the 2004 and 2005 irrigation seasons for five images, three in July and two in August, when irrigation was most intensive and also vegetative development was at a maximum for the season (Fig. 2 and 3) and results are given in Table 3. Monthly NDVIave values show an increase from the beginning of July to the end of August. Seasonal values varied between 0.32 and 0.42 for 2004 and between 0.26 and 0.42 for 2005. NDVIave values for all WUAs were higher in 2004. High NDVI represents good crop condition and hence better productivity (Sakthivadivel, 2001). Thus it can be seen that crop condition was better in 2004 than in 2005; in other words, water delivery in the selected WUAs was better as regards adequacy in the 2004 irrigation season.
Equity in water delivery: The equity of delivery of water supplied to
the irrigated areas of the four selected WUAs is assessed below using the CVNDVI
|| WUR values for the irrigation seasons 2004 and 2005
|| NDVIave and CVNDVI values for the irrigation
seasons 2004 and 2005
||NDVI maps for the irrigation season 2004
||NDVI maps for the irrigation season 2005
Variation coefficients of NDVI (CVANDVI): CVNDVI values for the 2004 and 2005 irrigation seasons of the four WUAs studied are presented in Table 3. In 2004, the WUAs season average coefficients of variation ranged from 0.14 to 0.23 and in 2005 from 0.14 to 0.30. When the two years are taken together, it can be seen that the coefficient of variation was low particularly in the irrigated area of the Gediz WUA and crop condition and, related to it, water delivery, were more homogeneous; that is, this WUA was more successful from the point of view of equity of water delivery. The reason why the coefficient of variation of the Salihli Left Bank WUA was higher is that it was less successful in terms of equitable water delivery.
This study assesses the water delivery performance of four selected WUAs in the Gediz Basin irrigation system from the point of view of adequacy and equity for the years 2004 and 2005, using remote sensing techniques. Adequacy of water delivery was determined according to irrigation ratio, water use ratio and average of Normalized Difference Vegetation Index and equity according to values of the coefficient of variation of NDVI. The study showed that irrigation ratio values for a selected WUA were related to its location on the main canal rather than their distance from the source. According to this indicator, a WUA located at the end of the main canal is in a worse condition with regard to adequacy. According to WUR and NDVIave indicators, WUAs were more successful in terms of adequacy in 2004. Also, there was a parallel between the two years regarding equity. That is, a WUA which was unsuccessful in 2004 was also unsuccessful in 2005 and a WUA which was successful in 2004 was successful in 2005 too.
Only if studies like this are carried out over many years in the Basin will more comprehensive assessment of the WUAs water delivery performance according to the selected indicators be possible.