پديد آورندگان :
حسيني، سيدموسي نويسنده استاد يار دانشكده جغرافيا، دانشگاه تهران , , عيوضي، اكرم نويسنده دانشجوي كارشناسي ارشد ژيومورفولوژي، دانشگاه تهران ,
كليدواژه :
آبخوان كارستي , چشمه‌علي دامغان , روش Coutagne , روش Mangine , منحني فروكش , مدل‌سازي جريان
چكيده فارسي :
در اين پژوهش، مدل تجربي Mangin و مدل Coutagne به منظور مدل‌سازي دبي روزانه چشمه‌علي دامغان در طول دوره فروكش مورد استفاده قرار گرفت. براي اين منظور داده هاي روزانه دبي چشمه مورد مطالعه طي نه سال متوالي (از سال 1380 تا 1388) اخذ و بكار گرفته شد. روش اول، بر اساس تيوري جداسازي جريان سريع كه در قسمت غيراشباع آبخوان به صورت موقت ناشي از تغذيه سطحي ذخيره‌شده از جريان پايه كه در تشكيلات كارستي آبخوان مطابق با تيوري Mailletوجود دارد، استوار است. درحالي‌كه روش دوم، منحني فروكش را پاسخ سيستم تك مخزني آبخوان به محرك ورودي (بارندگي) درنظر مي گيرد. نتايج روش Mangin در مورد چشمه‌علي دامغان نشان داد كه درصد مشاركت حجم جريان سريع در مقايسه با جريان پايه آبخوان كارستي از حجم كل جريان، بسيار اندك (%5/0) مي باشد كه نشان‌دهنده درجه بالاي توسعه‌يافتگي بخش غيراشباع اين آبخوان كارستي نسبت به بخش پاييني و غالب بودن سيستم افشان بخش اشباع آبخوان است. حجم جريان ذخيره‌شده در آبخوان كارستي كه توسط روش Mangin (با متوسط خطاي 5%) به دست آمده است، از خطاي كمتري نسبت روش Coutagne (با متوسط خطاي 23%) برخوردار است. درحالي‌كه مقادير پيش‌بيني‌شده دبي جريان در طول منحني فروكش توسط روش Coutagne(80/0=RMSE و 08/0= R2)، درجه دقت بالاتري نسبت به روش Mangin (44/0=RMSE و 45/0= R2) دارد. نتايج اين تحقيق نشان داد كه به منظور پيش بيني جريان روزانه چشمه كارستي طي دوره فروكش، روش Coutagne نسبت به روش Mangin از اين نظر كه لازم به تفكيك جريان هاي سريع و پايه چشمه طي دوره فروكش نمي باشد و همچنين به دليل سادگي ساختار مدل و اينكه تنها به يك پارامتر وابسته است و در نتيجه به محاسبات كمتري نيازمند است، داراي ارجحيت مي باشد.
چكيده لاتين :
Introduction
Karstic aquifers are important sources of water supply around the world. Approximately 25% of groundwater is stored in karstic formations. In Iran, karstic formations cover 11% of the land. More than 80% of municipal water demands in karst-dominated areas (e.g. Fars and Kermanshah provinces) are supplied from springs or wells in karstic aquifers. Karst aquifers are exceedingly heterogeneous in properties due to having double or triple porosity structures, mixed flow nature and varying conduit permeability (Quinlan et al., 1996). Hydrographs of spring flows show overall response of these aquifers to input precipitation (Padilla et al., 1994). The shape of recession curve of a hydrograph for a karstic spring is influenced by the aquifer size and hydrodynamic characteristics such as infiltration rate and water flow rate passed through the aeration and saturated zones. There are three main approaches for analyzing hydrologic functions of karstic aquifers and simulation of spring hydrographs: empirical relationships, conceptual and physical based models (Estrela and Sahuquillo, 1997). In Iran, the researches on the modeling of karstic spring hydrographs during recession period are very limited. In this study, two different type of empirical and conceptual models, respectively developed by Mangin (1975) and Coutagne (1968) were used to simulate the daily karst spring flow of Cheshmeh-Ali (located in Damghan city, Iran). Daily flows data for studied spring from 2003-2010 were collected from the Iranian Water Resources Management Co.
Methodology
Models description
The Manginʹs empirical method states that the recession curve of a spring hydrograph is composed of two parts: quick flow and delayed or base flow. This equation incorporates the Mailletʹs exponential formula to explain the base flow. The quick flow in this model estimated by an empirical function which two parameters of ? and ? play main role. The parameter of ? denotes to inverse of recession curve duration, while the parameter of ? indicate the capacity of infiltrated zone to pass the flow.
The Coutagneʹs method assumes the recession curve is the result of a single reservoir drainage to the input pulse (precipitation). Coutagneʹs expression for the recession curve of a karstic spring flow is a simple function depends on a parameter, n which determines the physical conditions of aquifer. For example, if n=0, the aquifer empties in a constant rate (without recession). If n=1 adopted, it means a thick aquifer discharges with constant groundwater velocity. When n=2, the hydrographs comes out from a stratified aquifer. For natural aquifer conditions, 1? n ?2.
Despite simulation of discharge ordinates (Q) during the recession period, the considered methods are also capable to compute the initial and total water volume stored in the karstic formations (in unsaturated and saturated zones) by integrating from discharge equation respect to time. Cheshmeh-Ali watershed with 148 km2 drainage area, average height of 1450 m, average pan evaporation, temperature, and precipitation respectively equal to 1900 mm/year, 16°C, and 155 mm located in Daryacheh-Namak basin. Computing the SPI for the monthly precipitation data during 2003 to 2010 shows that the 93% of total months lay in the normal region (-1 < SPI < 1). The Cheshmeh-Ali spring discharges the stored groundwater in karst aquifer averagely 600 lit/s (or 27 million cubic meters per year). Daily precipitation and spring flow data during 2033 to 2010 were used for assessing the efficiency of two considered models in simulation of recession curve ordinates and also stored water volume.
Results and Discussion
Two Manginʹs and Coutagneʹs methods were applied for simulation of recession curves of Cheshmeh-Ali during nine years (20033-2010). The recession curve of the hydrographs in the first year (2003) was considered for models calibration and other recessions for models verification. The beat fitted values of parameters including each models were obtained equal to n=1.36 (in Coutagneʹs method) and ?= 0.1 1/day, ?=0.77 1/day (in Manginʹs method). Using these fitted values, the discharge ordinates in recession period and also volume of storage water in karstic formations for years of 2004 to 2010 were simulated and compared with the corresponding observed values. The results of both models stand for domination of diffuse flow in compared to the conduit flow in the karst formations. The results of Manginʹs methods indicate that the 99.5% of total stored water in the karst aquifer is related to diffuse flow (base flow). In addition, the Manginʹs formula outperforms to the Coutagneʹs method to estimate the water storage, whereas the Coutagneʹs methods has higher efficiency in simulation of discharge ordinates during recession period in terms of R2 and RMSE.
Conclusion
According to results of two used methods, we conclude that the Coutagneʹs model has more efficiency from practical viewpoints in compared to the Manginʹs formula due to model simplicity (this method is depends on a parameter, n), time and cost consuming aspects, gives more accurate results, and doesnʹt require to base- and quick-flow separation. However, the Manginʹs methods yields the more precise water storage in karst aquifer.
