بررسی تأثیر پارامترهای فیزیوگرافی و اقلیمی حوضه در شبیه ‏سازی جریان فصلی رودخانه

نوع مقاله : پژوهشی

نویسندگان

1 دانش‌آموختۀ کارشناسی ارشد مهندسی منابع آب، دانشگاه علوم کشاورزی و منابع طبیعی گرگان

2 دانشیار گروه مهندسی آب، دانشگاه علوم کشاورزی و منابع طبیعی گرگان

3 استادیار، گروه مهندسی آب، دانشگاه علوم کشاورزی و منابع طبیعی گرگان

چکیده

خصوصیات فیزیوگرافی و شرایط اقلیمی در حوضه‏های آبریز از عوامل مهم دخیل در رژیم جریان رودخانه هستند که درک روابط بین این عوامل با جریان رودخانه در یک حوضه موجب می‏شود بتوان از این روابط در زیرحوضه‏های فاقد آمار برای پیش‏بینی جریان رودخانه استفاده کرد. در این مطالعه، روابط بین پارامترهای فیزیوگرافی و اقلیمی زیرحوضه‏های آبریز استان گلستان با جریان رودخانه با کاربرد مدل‏ درختی M5، مدل نزدیک‏ترین K- همسایگی (KNN) و رگرسیون چند‌متغیرۀ خطی (MLR) بررسی شد. داده‏های روزانۀ 28 ساله (1360‌ـ 1390) بارش، دما و دبی ایستگاه‏های هیدرومتری و هواشناسی 39 زیرحوضۀ آبریز برای استخراج سری‏های فصلی به‌منظور مدل‏سازی استفاده شد. متوسط مقادیر R و RMSE در فصول مختلف برای مدل M5 به‏ترتیب برابر 768/0 و 800/0، برای مدل KNN به‌ترتیب برابر 885/0 و 501/0 و برای مدل MLR به‏ترتیب برابر693/0 و 205/1 است که نشان‏دهنده برتری مدل KNN است. همچنین بر اساس مقادیر R و RMSE دقت نتایج مدل‏سازی در فصل‏های مختلف به‏ترتیب به‏صورت زمستان، پاییز، بهار و تابستان بوده است. به‌بیان دیگر نتایج پیش‏بینی جریان رودخانه در فصول تر از فصول خشک دقت بیشتری داشته است. همچنین بررسی مقادیر MBE نشان داد مدل KNN در فصل‏های بهار و زمستان به کم‌برآوردی و در تابستان و پاییز به بیش‌برآوردی منجر می‏شود. مدل M5 صرفاً در فصل بهار به کم‌برآوردی و در سایر فصول‌ به بیش‌برآوردی و مدل MLR نیز در زمستان‌ به کم‌برآوردی و در سایر فصول‌ به بیش‌برآوردی از مقدار مشاهداتی منجر می‏شود.

کلیدواژه‌ها

موضوعات


عنوان مقاله [English]

Investigation of effect of basin’s physiographic and climatic parameters in seasonal river flow simulation

نویسندگان [English]

  • Zahra Naeimi Kalourazi 1
  • Khalil Ghorbani 2
  • Meysam Salarijazi 3
  • Amir ahmad Dehghani 2
1 M.Sc. Graduate, Dept. of Water Resources Engineering, Gorgan University of Agricultural Sciences and Natural Resource
2 Deptartment of Water Engineering, Gorgan University of Agricultural Sciences and Natural Resources
3 Deptartment of Water Engineering, Gorgan University of Agricultural Sciences and Natural Resources
چکیده [English]

Physiographic characteristics and climatic conditions are factors which contributing to river flow regime and understanding of relations between these factors and river flow in a basin result in its application for the ungauged sub-basins river flow prediction. In this research the relation between physiographic and climatic parameters of Golestan province and rivers flow were examined by application of M5 regression tree model, k-nearest neighbors (KNN) model and multiple linear model (MLR). Daily recorded data for 28 years (1984-2011) including rainfall, temperature and river flow, belonging to hydrometry and meteorological stations of 39 sub-basins were used to extract seasonal series. The average of R and RMSE criteria in different seasons were 0.768 and 0.800 for M5 model, 0.885 and 0.501 for KNN model and 0.693 and 1.205 for MLR model which revealed better results for KNN model. In addition, according to R and RMSE, the accuracy of modeling results in different seasons were respectively as winter, autumn, spring and summer. In other words, the results of predicted river flows in the wet seasons were more accurate than dry seasons. Moreover, the MBE criterion indicated that the KNN model led to underestimation for spring and winter and overestimation for summer and autumn, M5 model led to underestimation in spring and overestimation in other seasons and MLR model had underestimation in winter and overestimation in other seasons.

کلیدواژه‌ها [English]

  • Keywords: River Flow
  • ungauged basins
  • M5 Decision Tree Model
  • KNN Model
  • MLR Model
 
 
[1].Govindaraju RS. Artificial neural networks in hydrology. I: Preliminary concepts. Journal of Hydrologic Engineering. 2000; 5(2): 115-123.
[2].Salajegheh A, Fathabadi A, Gholami H. Predict river discharge using the nearest neighbor. 5th national conference on science and management engineering Iran. Gorgan University of Agricultural Sciences and Natural Resources. 2010. [Persian].
[3].Lohani AK, Kumar R, Singh RD. Hydrological time series modeling: A comparison between adaptive neuro-fuzzy, neural network and autoregressive techniques. Journal of Hydrology. 2012; 442: 23-35.
[4].Kisi Ö. Daily river flow forecasting using artificial neural networks and auto-regressive models. Turkish Journal of Engineering and Environmental Sciences. 2005; 29(1): 9-20.
[5].Nabizadeh M, Mosaedi A, Hesam M, Dehghani AA. Comparing the performance of Fuzzy based models in stream flow on Lighvan River. J. of Water and Soil Conservation. 2012; 19(1): 117-134. [Persian].
[6].Parviz L, Kholghi M, Malmir M. Comparison of Methods temporal resolution and artificial neural networks in anticipation of a seasonal river flow. Journal of Iran Water Research. 2008; 2(2): 9-17. [Persian].
[7].Zarezadeh-Mehrizi M, Bozorg Haddad O. Inflow Simulation and Forecasting Optimization Using Hybrid ANN-GA Algorithm. Journal of Water and Soil. 2010; 24(5): 942-954. [Persian].
[8].Seyedian SM, Soleimani M, Kashani M. Predicting streamflow data-driven model and time series. Iranian Journal of Eco Hydrology. 2015; 1(3): 167-179. [Persian].
[9].Ahmadi F, Radmanesh F, Mirabbasi Najaf abadi R. Comparison between Genetic Programming and Support Vector Machine Methods for Daily River Flow Forecasting (Case Study: Barandoozchay River). Journal of Water and Soil. 2014; 28(6): 1162-1171. [Persian].
[10].Sanikhani H, Dinpajuh Y, Ghorbani MM. River flow modeling using K- nearest neighborhood and intelligent methods. Journal of Water and Soil Science. 2015; 25(1): 219-233. [Persian].
[11].Firat M, Gungor M. River flow estimation using adaptive neuro-fuzzy inference system. Journal of Mathematics and Computers in Simulation. 2006; 75(3-4): 87-96
[12].Negaresh H, Tavousi T, Mehdinasab M. Modeling the Production of Runoff in Kashkan River Catchment Based on the Statistical Methods. Journal of Research in urban ecology. 2014; 3(6): 81-92. [Persian].
13. Zare Abyaneh H, Bayat Varkeshi M. Evaluation of Artificial Intelligent and Empirical Models in Estimation of
 Annual Runoff. Journal of Water and Soil. 2010; 25(2): 365-379. [Persian].
[14].Eskandarinia AR, Nazarpour H, Ahmadi MZ, Teimouri M, Moshfegh MZ. Examine the effect of antecedent precipitation in the river flow estimates by artificial neural network (case study: Bakhtiari River). Journal of watershed management. 2011; 2(3): 51-62. [Persian].
[15].Khedmati H, Manshouri M, Heydarizade M, Sedghi H. Zonation and Estimation of Flood Discharge in Unguaged Sites Located in South-East Basins of Iran Using a Combination of Flood Index and Multi-Variable Regression Methods (Sistan and Baluchistan, Kerman, Yazd and Hormozgan Provinces). J. Water Soil. 2010; 24: 3: 593-609. [Persian].
[16].Akbari M, Van Overloop PJ, Afshar A. Clustered K nearest neighbor algorithm for daily inflow forecasting. Water resources management. 2011; 25(5):1341-57.
[17].Ghorbani Kh, Sohrabian E, Salarijazi M. Evaluation of hydrological and data mining models in monthly river discharge simulation and prediction (Case study: Araz-Kouseh watershed). Journal of Water and Soil Conservation. 2016; 23(1): 203-217. [Persian].
[18]. Ghorbani Kh, Meftah Halaghi M, Sohrabian E. Evaluation of hydrological and data-based models in estimation of daily runoff in Galikesh watershed. Int. J. Hydrology Science and Technology. 2016; 6(1): 27-44.
[19].Naeimi Kalourazi Z, Ghorbani Kh, Salarijazi M, Dehghani A. A. Estimation of monthly discharge using climatic and physiographic parameters of ungauged basins. Journal of Water and Soil Conservation. 2016; 23(3): 207-224. [Persian].