کاربرد روشهای استوکاستیک و مدلهای ریاضی در پیش بینی تغییرات سطح آب های زیرزمینی شهر شیراز

نوع مقاله : مقاله پژوهشی

نویسندگان

1 استادیار گروه مهندسی آب، واحد شیراز، دانشگاه آزاد اسلامی، شیراز، ایران

2 دانشجوی کارشناسی ارشد، گروه مهندسی آب، دانشکده علوم و کشاورزی و فناوری نوین، واحد شیراز، دانشگاه آزاد اسلامی، شیراز، ایران

چکیده

مدل‌های استوکستیک به عنوان یک روش جهت بررسی تغییرات داده‌های سری زمانی در آینده مورد استفاده قرار می‌گیرند. هدف از پژوهش حاضر تحلیل وضعیت سطح آب زیرزمینی چاه‌های پیزومتری شهر شیراز، شبیه‌سازی و پیش‌بینی کوتاه‌مدت وضعیت آن در آینده با استفاده از روشهای استوکستیک می‌باشد. در این تحقیق جهت مدل‌سازی، از آمار سطح آب 14 عدد چاه پیزومتری که آمار کامل‌تری داشتند از سال 1372 تا 1398 استفاده شد. سپس بر اساس مدل آریما و روش خودهمبستگی و خودهمبستگی جزئی و با ارزیابی تمامی الگوهای احتمالی به لحاظ ایستا بودن و بررسی پارامترها و انواع مدل‌ها، مدل مناسب جهت پیش‌بینی سطح آب زیرزمینی در هرچاه به صورت جداگانه بدست آمد. پس از اعتبارسنجی و ارزیابی مدل براساس معیارهای ضریب تعیین همبستگی خطی (R2)، معیار اطلاعاتی ضریب آکاییک (AIC)، جذر میانگین مربعات خطا (RMSE) و ضریب کارایی مدل (EF)، پیش‌بینی مقدار سطح آب زیرزمینی در سالهای 1399 تا 1405 انجام گرفت. نتایج حاصل از پیش‌بینی سطح آب زیرزمینی در چاه‌های سطح شهر شیراز با استفاده از مدلهای سری زمانی نشان داد که با فرض ثابت ماندن الگوی مصرف و عدم تغییر محسوس در روند تغذیه سفره آب زیرزمینی در طی 7 سال آینده به طور متوسط با کاهش حدود 5/3 متر در سطح آب زیرزمینی نسبت به وضعیت فعلی مواجه خواهد شد. بنابراین با توجه به افت قابل ملاحظه سطح آب زیرزمینی در آینده و محدودیت منابع تأمین آب شرب شهر شیراز و لزوم برداشت آب بیشتر در آینده، تصمیم‌گیری مناسب جهت مدیریت تأمین و مدیریت مصرف در این شهر ضروری به نظر می‌رسد.

کلیدواژه‌ها


عنوان مقاله [English]

The application of stochastic methods and mathematical models in forecasting of groundwater fluctuations in Shiraz city

نویسندگان [English]

  • Seyed Amir Shamsnia 1
  • Nazanin Pirnia 2
  • Reza Afshin Sharifan 1
1 Department of Water Engineering, Shiraz Branch, Islamic Azad University, Shiraz, Iran
2 M.Sc. Student of Water Engineering, Shiraz branch, Islamic Azad University, Shiraz, Iran
چکیده [English]

Stochastic models will be used as a method, to study changes in time series data in the future. The purpose of this study is to analyzing the groundwater level of piezometer wells in Shiraz, simulating and short-term forecasting of its status in the future using stochastic methods. for modeling In this research, water level statistics of 14 piezometer wells that had more complete statistics from 1993 to 2019 were used and based on ARIMA model and partial autocorrelation and autocorrelation method and with evaluation of parameters and types of models and all possible patterns in terms of being static, suitable model for predicting groundwater level in each well was obtained separately. After validation and evaluation of the model, the groundwater level was predicted in the future years 2020 to 2026. The results of groundwater level forecasting in wells in Shiraz using time series models show that with assuming the consumption pattern remains constant and no noticeable change in the groundwater recharge process, during the next 7 years, on average, we will face a decrease of about 3.5 meters in the groundwater level compared to the current situation. Therefore, due to the significant drop in groundwater level in the future and the limited resources of drinking water supply in Shiraz and the need for more water in the future, appropriate decision-making for supply management and consumption management in this city, seems necessary.It is necessary to study the consumption and demand management, in order to save water and prevent its loss in the city as a basic solution.
Extended Abstract
 
Introduction
Shiraz city, has faced with the problem of water shortage due to the increase in population in recent years. Currently, the most demand of drinking water is supplied by groundwater, which has been declining in recent years. In this research, quantitative status of groundwater will be analyzed according to the available information of the piezometric wells in city, during the last three decades. Then, by using time series models, its fluctuations are investigated and by representing a suitable ARIMA model, water level situation in the future can be predicted. Considering the importance of Shiraz plain aquifer in drinking water supply and the need to predict groundwater level changes to make decisions of relevant managers, for better management of water resources on the one hand and accuracy and ease of using time series models, what has been done in Iran and the world, led to the study of ARIMA time series models in this research.Finally, the purpose of this study is to predict the groundwater level in piezometric wells in Shiraz city by using stochastic methods and time series analysis.
 
Methodology
In the study area, 14 piezometric wells have been investigated. For predicting future conditions of water level, time series models based on Box-Jenkins, in ITSM software, were used. Model were calibrated with 4 parametes like R2, AIC, RMSE and EF. In the next phase, the suitability of the model is investigated and in short term (7 years), the groundwater level is predicted.
 
Results and Discussion
Results show that during past 27 years, water level has faced an avreage of 12.7-meter decrese and the general trend of groundwater unit hydrograph is downward and indicates a continuous drop in water level. Groundwater level prediction values obtained from Arima models, show that by assuming the current trend of water withdrawal in groundwater, in the next 7 years, we will face an average of 3.54 meters decrease above groundwater level in piezometric wells of Shiraz city.
 
Conclusion
The results of this study show that the trend of groundwater level is decreasing and will face water level drop in future.
According to the results of this study, the following can be mentioned:
1-  The proposed ARIMA models predict the groundwater level well in short-term. However, the application of the proposed model is not recommended for long-term forecasts.
2-  Since the trend of groundwater level changes in the past and in the future according to the results of the predicted model is declining, it is expected that this trend will continue in the following years.
3-         By considering the trend of groundwater level changes, it can be concluded that the most important reason for this downward trend is seasonal changes and fluctuations due to reduced rainfall and uncontrolled withdrawal of groundwater resources. Both of which, Have significant affect on the process of water level changes in the future. It is necessary to study the consumption and demand management, in order to save water and prevent its loss in the city as a basic solution.

کلیدواژه‌ها [English]

  • Model
  • Groundwater
  • Shiraz
  • Stochastic
  • Time series
  • Afruzi, A., & Zare Abyaneh, H. (2017). Groundwater level modeling and forecasting using time series models: Case study of the plains of Hamedan province. Journal of Watershed Management Research, 8(15), 102-111. [In Persian]
  • Babazadeh, H., Shamsnia, A., Boostani, F., Norouzi Aghdam, E., & Khodadadi Dehkordi, D. (2012). Investigation of drought, wet season, and prediction of climatic parameters of rainfall and temperature in Shiraz region using stochastic methods. Journal of Geography and Planning, 16(41), 23-47. [In Persian]
  • Cheraghi, A. M., Najafi, B., Shajari, Sh., & Javan, M. (2020). The trend of changes in groundwater quantity and quality in the Sarvestan plain of Fars province. Journal of Watershed Management Research, 33(2), 82-96. DOI: 10.22092/wmej.2019.128069.1283 [In Persian]
  • Gibrilla, A., Anornu, G., & Adomako, D. (2017). Trend analysis and ARIMA modeling of recent groundwater levels in the White Volta River basin of Ghana. Journal of Geographical Sciences, doi.org/10.1016/j.gsd.2017.12.006
  • Jamalizadeh, M. A., Bazrafshan, O., Mahdavi Najafabadi, R., Azare, A., & Rafee Sardooni, E. (2020). Predicting groundwater level fluctuations using time series models and GMS: Case study of Rafsanjan plain. Ecohydrology, 7(1), 97-109. [In Persian]
  • Karimipoor, A., Rakhshanderoo, Gh., & Banitalebi Dehkordi, G. (2011). Evaluation of the drainage system to lower the groundwater aquifer in Shiraz plain using Pmwin model. Journal of Water and Wastewater, 91(2), 30-41. [In Persian]
  • Mirzavand, M., Sadatinejad, S. A., Ghasemieh, H., Imani, R., & SoleymaniMotlagh, M. (2014). Prediction of groundwater level in arid environment using a non-deterministic model. Journal of Water Resource and Protection, 6(2014), 669-676. doi.org/10.4236/jwarp.2014.67064.
  • Mirzavand, M., & Ghazavi, R. (2014). A stochastic modeling technique for groundwater level forecasting in an arid environment using time series methods. DOI 10.1007/s11269-014-0875-9
  • Nasseri, N. (2019). Comparison of fourteen methods of time series to analyze and predict groundwater changes in Marand Plain-North of Urmia Lake. Iranian Journal of Irrigation and Drainage, 13(1), 58-68. [In Persian]
  • Patle, G. T., Singh, D. K., Sarangi, A., Rai, A., Khanna, A., & Sahoo, R. N. (2015). Time series analysis of groundwater level projection of future trend. Geological Society of India, V85, 232-242.
  • Poormohammadi, S., Malekinezhad, H., & Poorshareyati, R. (2013). Comparison of ANN and time series appropriately in the prediction of groundwater table: Case Study Bakhtegan basin. Journal of Water and Soil Conservation, 20(4) [In Persian]
  • Rahmani, A., & Sedehi, M. (2004). Prediction of groundwater level changes in the plain of Hamedan-Bahar using the time series model. [In Persian]
  • Retike, I., Bikse, J., Kalvans, A., Delina, A., Avotniece, Z., Zaadnoordjik, W., Jemeljanova, M., Popovs, K., Babre, A., Zelenkevics, A., & Baikovs, A. (2022). Rescue of groundwater level time series: How to visually identify and treat errors. Journal of Hydrology, 605, 127294.
  • Salehi, M., Radmanesh, F., Zaree, H., Mansoori, B., & Solgi, A. (2018). Predicting groundwater level by using time series-wavelet combination model. Journal of Irrigation Science and Engineering, 41(4), 1-16. [In Persian]
  • Shahraki, N., Yunesi, M., & Taheri Tizro, A. (2019). Comparison of artificial neural network, ARIMA time series, and multivariate linear regression models in predicting groundwater level changes. Journal of Hydrogeology, 4(1), 126-139. [In Persian]
  • Takafuji, E., Rocha, M., & Manzione, R. (2018). Groundwater level prediction/forecasting and assessment of uncertainty using SGS and ARIMA models: A case study in the Bauru aquifer system (Brazil). Natural Resources Research, doi.org/10.1007/s11053-018-9403-6
  •