The relationship between land use change and water quality in Kucukcekmece Lake watershed. 4th GIS Days in Turkey

THE RELATIONSHIP BETWEEN LAND USE CHANGE AND WATER QUALITY IN KÜÇÜKÇEKMECE LAKE WATERSHED

Ali DEMĐRCĐ1, Michael Andrew MCADAMS1, Omar ALAGHA2, Mehmet KARAKUYU1

ABSTRACT

Remote Sensing and GIS can be powerful tools in analyzing land use change. In watersheds, the land use has a direct effect in determining water quality. In this study, land use changes are analyzed by using Remote Sensing and GIS from 1963 to 2005 for Küçükçekmece watershed (N40.59"24; E 28.45"57) in Istanbul. Land uses classification was done by utilizing a Corona 1963 satellite image, an aerial photography from 1996 and a 2004 IKONOS high resolution image. In parallel to that, water quality samples taken at 19 locations around the lake were compared with the land use changes. Water quality analysis and land use investigation showed that the unplanned urban and industrial expansions and domestic and industrial waste discharge are having a direct impact on the water quality of the lake. This study demonstrates the usefulness of the combined tools of GIS and Remote Sensing to measure land use changes in watersheds.

Keywords: GIS, Remote Sensing, Land Use Change, Water Quality, Urban Development, Watershed Management.

1. INTRODUCTION

Remote Sensing and GIS are excellent vehicles to monitor land use changes in watersheds. The key factors in monitoring watershed land uses would be the amount of impervious surfaces (roads, buildings, parking lots etc.) and the amount/types of land uses (residential, commercial, industrial and agricultural). Obviously, the best condition for a watershed is for there to be no development. This may be unfeasible due to market forces. In this light, some land uses can be allowed that will not significantly degrade the watershed. The effect of different land uses on a watershed can be estimated by the modeling; thus giving the ability to ascertain the amount of pollution and its impact on water quality of the watershed (Tang, 2005; Tong, 2002). The monitoring of watersheds is particularly aided by high-resolution images (i.e., IKONOS). Analyzing land use change over a long period of time is possible through images collected through various international programs (Clapham, 2003). The combined use of GIS and Remote Sensing with other statistical techniques, such as modeling offers much promise in watershed analysis.

The protection of watersheds and their associated reservoirs is crucial to providing suitable and safe drinking water for urbanized areas. Reservoirs that are on the fringe of urbanized areas are at a particular risk as market forces clash with the public interest (Beck, 2005). While drinking water degradation due to inadequate watershed regulations does occur in developed nations, the situation may be more severe in developing nations which have little or no environmental controls (Lee, 1996). Once a reservoir is degraded, remediation can be difficult, if not impossible. The water authorities will then be forced to cope with the existing water supply and establish new reservoirs.

This degradation of a reservoir happened in the case of Küçükçemece reservoir in Istanbul, Turkey (see Figure 1). Küçükçemece reservoir was originally an estuary of the Marmara Sea. It was a clean natural habitat before the 1980’s (Demirci, A., 2001). In the 1980s, Istanbul started to have a tremendous growth spurt. The development quickly transformed this previously sparsely populated semi-rural/exurban area of Küçükçemece to an urbanized one. There were no regulations governing the protection of the watershed. All types of development were allowed to build in this sensitive area. Industries directly dumped into streams their waste. High density residential development occurred on the edges of the reservoir. Raw sewage from these developments was allowed to be put into the streams and directly into the lake. The result was two fold: first, the greater Istanbul area lost a valuable reservoir; and second, the watershed continued to be degraded by unregulated development so that it became an ecological disaster.

1 Fatih University, Department of Geography, Turkey, [email protected], [email protected], [email protected] 2 Fatih University, Department of Environmental Engineering, Turkey, [email protected]

2. OVERVIEW ABOUT STUDY AREA

The study area, Küçükçekmece Lake watershed (see Figure 1), is located twenty-four kilometers (24 km) southwest of the center of Istanbul and adjacent to the Marmara Sea. It has a surface area of approximately fifteen square kilometers (15.22 km²). The watershed is one hundred and sixty-five square kilometers (165 km²). A narrow channel connects it to the Marmara Sea in the southeast. The lake is ten kilometers (10 km) in length and six kilometers (6 km) at its widest part. The average depth of the lake is approximately ten meters (10 m). The deepest part is twenty meters (20 m) near the southern edge of the lake. Three stream systems feed the lake: Nakkaşdere, Sazlıdere and Ispartakule (Demirci, 2001). The Sazlıdere stream output into the lake is much less due to the damming of this stream in 1995, which formed Sazlıdere Lake. The construction of a dam on this stream caused Küçükçekmece Lake to lose almost half of its watershed area. The lack of fresh water which was coming from the Sazlıdere stream did not affect Küçükçekmece’s water level due to its connection with the Marmara Sea. However, its salinity has increased dramatically. Since the discharge of Nakkaşdere stream was stopped and diverted offshore to the Marmara Sea by a new pipeline system in 2005, the lake has been fed by the Ispartakule stream from the northwest, surface runoff from the surrounding areas and by the sea water from the south (ĐSKĐ, 2005a).

Figure 1: The Location of the Study Area

3. METHODS AND ANALYSIS

3.1. Water Quality Samples

In March and April of 2003, surface water samples from nineteen different locations were collected and analyzed (Güney, K., 2003) (See Figure 2). Samples for chemical analysis (COD, nitrogen and phosphorus) were collected in pre-washed HDPE bottles and brought to lab on the same day for analysis. Microbiological parameters (total coliform) samples were collected in sterilized glass bottles that brought to laboratory for analysis within the same day. All the analysis was carried out according to the standard method of analysis (Franson, 1998). Direct on-site measurements were made for pH, temperature, turbidity, and dissolved oxygen (DO) using a HORIBA multi-probe meter. In 2005, the data was inserted into a GIS (ArcGIS 8.3). Different prediction maps were produced by using the Geostatistical Analyst Extension of ArcGIS for each of the different parameters including pH, DO, COD, coliform, conductivity, turbidity; phosphate and Kjeldahl total nitrogen (see Figure 2 for sampling locations and land use).

Figure 2: Land Uuse Map of the Küçükçekmece Watershed, 2004.

3.2. Land Use Analysis

Remote sensing images were used to document the land use changes between the years of 1963, 1996 and 2004 as provided by the Istanbul Water Authority (ISKI). The 1963 image was an NASA aerial photograph under the Corona initiative. The 1996 image was an orthophoto taken by ISKI. The 2004 image was taken by an IKONOS satellite. Each of the images was manually inspected and the land use classified and digitized in ArcGIS 8.3. The digitized classifications were further converted to raster so that the amount of the land use could be determined. Several land use types were identified for the 2004 land use map including: agricultural, residential, industrial/commercial, mining, construction areas, institutional, open land, bushes, marsh, forest, pond and major roads (e.g., Trans European Motorway-TEM). The residential land uses were divided into three different categories, Residential-1, 2 and 3. Residential-1 includes the densest residential areas with no open spaces. Residential-2 includes areas of open spaces still to be developed. Residential-3 includes new residential areas. To more clearly discern the changes of urbanization and industrialization in the watershed, the number of land use classes was reduced to six for each land use map. At this point of the project, the water quality data of 2003 and some previous studies

were been used with the three different land use maps in order to see the effects of land use changes on the water quality.

4. RESULTS AND DISCUSSIONS

Over the last thirty years, the combination of different factors affected the water quality of Küçükçekmece Lake and turned it into a polluted unnatural habitat and unhealthy ecosystem. Among the factors that polluted the lake, the most important ones are: rapid, unplanned urban and industrial expansions; domestic and industrial waste discharge to the lake; agricultural pollution, the leakage from a closed garbage dump area; and other non-point resources such as run-off from streets and highways

4.1. Land Use Analysis

There was a dramatic shift in the land use composition within the watershed from 1963 to 2005 (See Figure 3). In 1963, most of the land use was agricultural and open space (98%).

Figure 3: Land Use Changes in Küçükçekmece Watershed, 1963-2004

Residential areas were sparse (1%). There were no industrial buildings in the area. By 1996, the residential areas increased dramatically but still only occupied a small amount of the watershed area (16.92 %). There was considerable increase in residential growth from 1996 to 2004 (42%). This percentage of area for residential land uses increased to twenty-four percent (24%) of the total areas in 2004. The new construction was mainly in the northern portion of the watershed (see Figure 2). In 2004, the watershed consisted of agricultural land (42%); residential areas (24%); open land including bushes and marsh around the lake (23%); industry and commercial land uses (4%); institutional lands including military, customs house, and a football stadium (3%); construction areas for new houses (2 %); a major expressway, Trans European Motorway (TEM), (1 %); mining (0.4%) and forest (0,2%).

4.2. Pollution Parameters

In Figure 4, the concentrations of measured parameters are shown for March 2003. By current water quality standards, the lake’s water does not even approach safe drinking water standards and can be regarded as a heavily polluted lake in terms of many parameters like total coliform, turbidity, temperature, phosphate, DO and COD (European Union, 2006; ĐSKĐ, 2005b; U.S. E.P.A., 2004). The average pH

Large Changes

value is 8.7 for the lake. This value is approximately the same as open sea pH, which indicates the intrusion of saline sea water from the southern part into the lake converting it to brackish one. Some of the highest concentrations of pH value is noticed near the connection channel (see Figure 3, points 5 and 18) and thus supporting this theory. The average high conductivity values (11.4 mS) is further evidence of the intrusion of sea water. The uniform distribution of salinity and pH within the lake is attributed to the prevailing wind systems (Lodos winds) which increase the intrusion of sea water into the lake. Dissolved oxygen (DO) values are between 5.5 and 21 mg/l. in the lake with the lowest level on the southeast portion. This would indicate areas where some fish species may be under stress. The DO supersaturated surface water (unpolluted water DO is about 8.6 mg/L) is a clear indication of level approaching eutrophication. This phenomenon can be further explained by inspecting the eutrophication precursors of nitrogen and phosphorus. Phosphate concentrations are far exceeding and unpolluted surface water (range 0.1-9.8 mg/L).

Figure 4: Distribution of Pollutants in Küçükçekmece Lake

The Phosphates and Nitrates levels found are the result of sewage discharge and the application of agricultural fertilizers. Excess Phosphorus assists algal growth in freshwater lakes. Chemical Oxygen Demand (COD) values in the lake show a large value spread (between 4.4 and 732 mg/l). The highest values of COD are found near the Ekşinoz and Nakkaşdere streams. These streams receive a large amount of pollution from the nearby Hadimköy and Đkitelli industrial areas. The total coliform values are too high for either drinking water or recreational waters with average 563 per 100 ml. It is the highest in the north at the mouth of streams and over the southern portion of the lake with values above 1600 per 100 ml. This is because the pollutants are entering the lake from the streams and then concentrate in its southern portion. Turbidity is widespread across the lower levels in the southwest portion and higher near some of the streams

There is no doubt that the land use has affected the water quality of Küçükçekmece. To directly relate particular areas that are responsible for the distribution of the water quality indicators is sometimes difficult. This is because the lake surface is small and affected easily by wind patterns and surface-subsurface currents. The water quality parameters of the stream waters show a wide variation due to the different industrial facilities discharging their waste at different times. It takes industrial facilities a few

minutes to discharge their waste to the streams and within a short time period pollutants reach the lake. A new data set is expected to provide more insight between the relationship of land use and water quality.

The most recent industrial areas are in the northwestern part of the region. There are a wide variety of industrial firms in the areas (i.e., textiles, metallurgy, chemistry, wood and leather). Liquid wastes from nearly all these industrial facilities are being directly discharged into the streams and then carried to the lake without any treatment. Obviously, the streams have been adversely affected by this discharge. The lake is not only being polluted by the industry in the watershed, but by industrial facilities outside of it. Approximately 600 industrial firms in the upper northwest of the watershed discharge their waste to the Ekşinoz stream. These industries, which include textile and leather processing firms, discharge a significant amount of toxic chemicals into the stream

5. CONCLUSION

The combination of Remote Sensing and GIS proved to be powerful in documenting the amount of growth, spatial distribution and the changing land use composition that led to the degradation of the Küçükçekmece watershed. The use of high resolution IKONOS images enabled detailed classification of land uses for the present period. The exact year when the watershed reached a point where it was unusable for drinking water is unknown. Nevertheless, it is obvious that the industries that are polluting the streams entering into the lake and the run-off from impervious surfaces are directly related to the state of water quality. This is generally observed as the concentration of certain pollutants such as Phosphates and Nitrogen near the streams entry into the lake.

The recognition of the problems is just one step. It is hoped that the public will become more aware of its environmental problems and the opportunities open to reclaim this lake as a recreational and urban wildlife resource and urge governmental officials to direct significant funds to make this lake a success story and not a tragedy.

REFERENCES

Avcı, C. B., Güler, E., Toğrol, E., Karpuzcu, M., & Kılanç, B. (1995). Düzensiz Katı Atık Depolama Alanlarının Yeraltı Su Kalitesine Etkileri: Đstanbul Halkalı Katı Atık Alanı Örnek Çalışması. Teknik Dergi, cilt 6, s. 1, TMMOB, Đnşaat Mühendisleri Odası

Ardos, M. (1994). Jemorfoloji Sözlügü, Đstanbul: Đst. Ünv. Ed. Fak. Yayınları.

Alparslan, E., & Yüce, H. (2003). Monitoring Urban Growth Around Küçükçekmece Lake Through Remote Sensing Technology. Küçükçekmece ve Yakin Çevresi Teknik Kongresi Bildiriler Kitabi, 2003. Deprem ve Planlama (p. 62-72), vol (1), Đstanbul, 08-10 Ekim 2003.

Bird Life International (2004). BirdLife IBA Fact sheet for Küçükçekmece, Retrieved May 30, 2007, from http://www.birdlife.org/datazone/sites/?action=SitHTMDetails.asp&sid=780&m=0

Beck, M.B. (2005). Vulnerability of water quality in intensively developing urban watersheds. Environmental Modeling and Software, 20, 381-400.

Carpenter, S. R., & Cottingham K. L. (1997). Resilience and restoration of lakes. Conservation Ecology [online]. Retrieved May 30, 2007, from http://www.ecologyandsociety.org/vol1/iss1/art2/ and http://www.ecologyandsociety.org/vol1/iss1/art2/inline.html

Chapman, W.B. (2003). Continuum-based classification of remotely sensed imagery to describe urban sprawl on a watershed scale. Remote Sensing of Environment 86, 322-340.

Demirci, A. (2001). Types and distribution of landslides in the Eastern Part of Büyükçekmece Lake by Using GIS. Unpublished graduate thesis. Fatih University, Đstanbul, Turkey.

Demirci, A., & Butt, A. (2001). Historical Overview and Current Trends in Istanbul’s Water Supply Development, Globalization and Water Resources Management: The Changing Value of Water, AWRA/IWLRI-University of Dundee International Specialty Conference, August 6-8, 2001, Dundee, Scotland.

Ercan, A. (1990). Gürpinar Beldesi Yer Mühendislik Projesi. Yeraltı Aramacılık Billimsel Araştırma Kuruluşu, cilt 1, 2, Istanbul.

Erol, O., & Altın, B. N. (1991). Binkılıç-Karacaköy Dolayının Jeomorfolojisi, Istranca Dağları Güneydoğusu, Trakya. Atatürk Kültür, Dil ve Tarih Yüksek Kurumu Cografya Arastirmalari Dergisi, sayı 3.

Franson, M. (1998). Standard methods for the examination of water and wastewater. American Public Health Association.

European Union, European Communities (Drinking Water) Regulations. (2000). Retrieved May 30, 2007, from http://www.fsai.ie/legislation/food/eu_docs/Water/SI439_2000.pdf (put into effect 2004).

Güney, K., Sülün, A. D., Üstündağ, R. N., & Alagha, O. (2003). Küçükçekmece Gölündeki Bazı Çevresel Parametrelerin Analizi ve Kirlilik Haritaları, Üniversite Öğrencileri Çevre Sorunları Kongresi 14-16 Nisan 2003, Fatih Üniversitesi, Đstanbul.

ĐSKĐ. ( 2000). Đstanbul Su ve Kanalizasyon Đdaresi 1999-2000 Annual Report.

ĐSKĐ. (2005a). Đstanbul Su ve Kanalizasyon Đdaresi 2004-2005 Annual Report.

ĐSKĐ. (2005b). The Adventure of Water in Istanbul. ĐSKĐ Yayınları, Đstanbul

Klapper, H. (2005). The assessment, management and reversal of eutrophication’ in The Lakes Handbook. vol, 2. Blackwell Publishers.

Lee, M.D. (1996). Multiple resource needs and multiple conflicts in urban watersheds in developing countries: the case study of the Guacerique watershed, Tegucigalpa, Honduras. Proceedings of the Integrated Management of Surface and Groundwater. UCOWR Annual Meeting, July 30-August 2, 1996. San Antonio, Texas. Retrieved May 30, 2007 from http://www.isis.csuhayward.edu/ALSS/Geography/mlee/guacerique.html,

Living Lakes. (2005). Habitat Restoration. Retrieved May 30, 2007 from http://www.livinglakes.org/issues/restoration.htm#briggs

Orhon, D., Çetiner, A., Zeren, N., Giritlioğlu C., ve diğerleri. (1988). Küçükçekmece Atiksu Toplam Alani Içindeki Endüstri Tesislerinin Konumu ve Kirleti Yüklerinin Belirlenmesi, ĐSKĐ Gn.Md., S.190.

Pehlivan, R., & Yılmaz, O. (2001). Heavy Metal Content of the Lake and Stream Water in Küçükçekmece Lake Basin, 54th Geological Congress of Turkey, Ankara.

Tang, Z., Engel, B.A., Pıjanowskı, B.C., & Lim, K.J. (2005). Forecasting land use change and its environmental impact at a watershed level. Journal of Environmental Management, 76, 35-45.

Tanık, A., Beler, B. B., & Gonenc, I.E. (1999). The impact of agricultural pollutants in six drinking water reservoirs. Water Science and Technology, 40:2, 11-17.

Tong, S., & Chen, W. (2002). Modeling the relationship between land use and surface water quality. Journal of Environmental Management, 66, 377-393.

Topçuoğlu, S., Güngör, N., & Kırbaşoğlu Ç. (1997). Küçükçekmece Gölü'nün Saptanan Bazı Çevresel Parametreleri. Su Ve Çevre Sempozyumu Bildiriler Kitabi (s, 89). Đstanbul, 2-5 Haziran 1997: TMMOB Jeoloji Müh. Odası, Bakırköy Belediyesi, TMMOB Đstanbul Şubesi Yayını, No.: 1.

Tuncer, M. (1999). Doğal Çevre Koruma Öncelikli Bir Eylem Alanı: Đstanbul Küçükçekmece Gölü. Türkiye’de Çevre Kirlenmesi Öncelikli Sempozyumu III, 18-19 Kasım 1999.

U.S. E.P.A. (2004). Drinking Water Standards for Regulated Contaminants. Retrieved May 30, 2007, from http://www.epa.gov/safewater/therule.html,

Üstün B., Đnce N., Cansever G., Ülger E., Ağcıoğlu B., Şapçı Z., et al. (2002). The Development of Environmental Management Model in Küçükçekmece Basin. Joint Research and Development Project. (GSRT- TÜBĐTAK).