Hydrogeological investigation and numerical modeling of Ergene (Thrace) basin

Ergene Basin is located both close to Istanbul and in a region such as Thrace Region where settlement and industrialisation are dense and of great economic importance. For this reason, there are various studies on many subjects in Ergene Basin such as the quality of water resources, industrial density and its effects, determination of the dimensions of environmental pollution, socio-economic structure and environmental layout plan. In recent years, hydrogeological studies have been conducted at the sub-basin level, and the Çorlu-Çerkezköy sub-basins have been the most studied areas where industrialization is intensive. In this study, the Lüleburgaz sub-basin, where industrialization and population are increasing, was also addressed. Within the scope of the study, current geological and hydrogeological studies were carried out in the Lüleburgaz sub-basin, surface and groundwater quantity and quality were examined, the relationship between surface and groundwater was evaluated, numerical modeling studies were conducted, and predictions were made regarding how groundwater levels in the basin would change in the future. For these purposes, field and laboratory studies were carried out for two periods, physico-chemical properties of a total of 26 water samples were evaluated, the water balance sheet of the basin was prepared, and also a numerical flow model was created to predict what kind of hazards await the basin in different scenarios. The sub-basin under study covers an area of approximately 2150 km2 and is located roughly in the center of the Thrace Region. Morphologically, the basin has an average elevation of 50-100 m. in the region where the Ergene River flows. There are elevations close to 1000 m. towards the north of the basin. When the geology of the basin is analysed, the northern border is bounded by metamorphites and igneous rocks. Among these, Tekedere Group (Pzt) consists of biotite-chlorite gneisses and schists, Kırklareli Group includes gneisses and metagranites, Kocabayır Group comprises metasandstones, metasiltstones, and quartz-feldspathic schists, and Mahya Group is composed of phyllites and schists. Overlying these foundational units, the unconformably deposited İslambeyli Formation (Tei) consists of shales, sandstones, and siltstones, Soğucak Formation (Tes) is primarily composed of limestones, Pınarhisar Formation (Top) consists of sandstones, conglomerates, and occasional limestones, Süloğlu Formation (Tos) comprises shales, Ergene Formation (Tme) includes conglomerates, sandstones, siltstones, and claystones, Sinanlı Formation (Tms) is characterized by limestone with claystone interbeds, and Trakya Formation (Tnt) is made up of conglomerates, sandstones, and claystones in succession. Precipitation data for the study area was obtained from the Turkish State Meteorological Service (MGM), with an average rainfall of 581.4 mm, while the average evaporation and transpiration, calculated using the Thortwaite method, was determined to be 436.4 mm. Flow data was acquired from the General Directorate of State Hydraulic Works (DSI), and the flow data calculated using the Soil Conservation Service Curve Number (SCS-CN) method was compared with actual flow data. As a result, the average annual precipitation in the Lüleburgaz Sub-Basin was estimated to be 1250 x 106 m3/year, while the average evaporation and transpiration was 938 x 106 m3/year. The remaining 312 x 106 m3/year (as measured by flow measurement station data with an average of 335 x 106 m3/year over the last 5 years) was measured as the total flow. The average surface runoff value calculated using the SCS-CN method was 157.6 x 106 m3/year, while it was found to be 135.8 x 106 m3/year from flow observation stations. Since the flow measurement station value is based on measured data, this value was taken as the reference. The remaining 176.2 x 106 m3/year was considered as groundwater recharge. The difference between total discharge and recharge for the study area is approximately only 9 x 106 m3/year. In recent years, the increase in illegal wells and groundwater extraction, along with an annual drop in groundwater levels ranging from 0.6 to 1.2 meters, has led to the apparent overuse of safe groundwater resources for an extended period. A groundwater level map was created using data obtained from groundwater measurement wells in the study area. According to this map, the groundwater flow direction is observed to be from northeast to southwest in north of the Ergene River, and from southeast to northwest in south of the Ergene River. Within the study area, a total of 1 spring, 3 rivers, and 22 well samples were collected during 2 periods (spring and autumn of 2017). Their analyses were conducted, and their suitability was assessed according to the permissible upper limit values specified by TSE-266 (2005) and WHO (2008) standards. Physical parameters such as pH, electrical conductivity, and dissolved oxygen levels were measured at the wellheads. Major anion and cation values along with secondary constituents were analyzed at the ACME Analysis Laboratory. Based on the analysis results, the waters were evaluated using Scholler, Piper, Gibbs, Ion Exchange, Durov, US Salinity Laboratory, and Wilcox Diagrams. As a result of the analyses, it was determined that some samples exceeded the permissible upper limit values for pollution parameters. The Visual Modflow Flex 7.0 program was used in groundwater flow modeling studies. Geological-hydrogeological data (clay hydraulic conductivity 5x10-7 m/s, porosity 0.45; sandy aquifer hydraulic conductivity 0.0001 m/s, porosity 0.35, calculated precipitation and evaporation values, aquifer recharge rates, groundwater levels in monitoring wells, and discharge rates from production wells were used as input parameters for the modeling work. Initial head conditions and river boundaries were defined in the model. Calibration of the model was performed using groundwater level measurements and hydraulic head values obtained from monitoring wells within the model area. The correlation coefficient between the measured groundwater levels in the field and the calculated groundwater levels in the model was found to be 0.9. Additionally, the normalized root mean square error was calculated to be 5.29%. In total, 6 different scenarios were created, and in a model designed for 10 years into the future assuming a 100% increase in water extraction from the wells, it is observed that the groundwater level falls below the sea level.