ارزیابی عملکرد مدل ihacres با روش‌های خطی armax و expuh (مطالعه موردی: حوضه رودخانه شور در قائن)

مدل‌سازی بارش -رواناب یکی از ابزارهای کلیدی در هیدرولوژی برای دستیابی به خصوصیات سیلاب، مانند میزان دبی اوج و زمان رسیدن به دبی اوج است. در این تحقیق مدل ihacres در قالب بسته نرم‌افزاری hydromad و در بستر r، برای شبیه‌سازی جریان حوضه رودخانه شور قائن مورداستفاده قرار گرفت و عملکرد آن در شبیه‌سازی رواناب با روش‌های armax و expuh در بخش خطی مدل، بررسی شد. برای انجام شبیه‌سازی با مدل مذکور، از ‌داده‌های بارش، دبی جریان و دما در مقیاس ماهانه طی سال های 1377 تا 1398 استفاده شد. پارامترهای مدل ihacres با استفاده از تابع هدف kge، واسنجی شدند.  در گام های واسنجی و صحت سنجی مدل در شبیه‌سازی مقدار ‌دبی اوج و زمان رسیدن به ‌دبی اوج ، از معیارهای ارزیابی ns، kge و  استفاده گردید. مقدار این معیارها در روش expuh به ترتیب برابر 0.86، 0.93، 0.86 و در روش armax برابر 0.7، 0.85 و 0.73 محاسبه‌شد که نشان دهنده­ی عملکرد بهتر روش expuh نسبت به روش armax می­باشد. درگام مربوط به صحت‌سنجی نیز، معیارهای ارزیابی در روش‌های expuh ‌و armax به‌ترتیب برابر 0.51، 0.63، 0.54 و 0.55، 0.73 و 0.65 برآورد شد که نشان‌دهنده عملکرد بهتر مدل با روش armax نسبت به روش expuh است و همچنین مدل با روش armax در شناسایی دبی‌های اوج از نظر مقدار و از نظر زمان وقوع عملکرد دقیق‌تری از خود نشان داد. درمجموع نتایج تحقیق، نشان‌دهنده این مطلب است که مدل ihacres در منطقه موردمطالعه، با استفاده از روش armax نسبت به روش expuh عملکرد دقیق‌تری داشته است.

Assessment the Performance of IHACRES Model Using ARMAX and EXPUH Linear Methods (Case Study: Shoor River Basin in Ghaen)

IntroductionThe watershed acts as a hydrological unit regulating the quantity and quality of the water cycle, and human beings have incurred high costs due to ignorance of this complex cycle and lack of planning of projects in terms of the relationship between water management and community development.Knowledge of features such as maximum flood discharge is essential for the design of hydraulic structures, such as dams, spillways, bridges, and culverts, in order to reduce potential damages and predict when peak discharges will be reached in the downstream areas when discussing flood warning. Rainfallrunoff modeling is one of the key tools in hydrology to achieve flood characteristics, such as peak rate and peak time. In current research, the performance of IHACRES model using &Hydromad& R package has been implemented to simulate flow in the Shoor river basin in Ghaen on a monthly scale. The model simulation was done to investigate the effect of selecting &ARMAX& and &EXPUH& methods in the linear part of the target function. Also, the modeling process and the optimized values of the model parameters were investigated.Materials and MethodsThe Shoor river basin with an area of 2412.92 square kilometers located in Ghaen between 59 degrees and 12 minutes to 59 degrees and 14 minutes east longitude and 33 degrees and 42 minutes to 33 degrees and 45 minutes north latitude. The study catchment with an average altitude of 1420 m above sea level and an average longterm annual rainfall of 173 mm has a dry climate. This river is the largest river in Ghaenat city which flows into Khaf Salt field. In this research, the IHACRES model was implemented using the Hydromad R package. To perform the flow simulation, precipitation, flow rate and temperature data on a monthly scale during the years 1998 to 2017 were used. The IHACRES model has two parts: the first part, which converts precipitation into effective precipitation at each time stage and the second part, which converts effective precipitation into modeled flow. These sections are called nonlinear and linear modules, respectively. To implement each of the sections of nonlinear modules and linear modules according to the data and conditions in the study area, methods with different parameters can be used. In this research, in the nonlinear module section, the &CWI& method and in the linear module section, &ARMAX& and &EXPUH& methods have been used for proper routing in the &reverse& calibration section. In the validation section of the &ls& method, the performance criteria of KGE, NS and R2 were used to evaluate the performance of the model in both calibration and validation process. Result and DiscussionComparison of obtained results in this study with previous studies showed that in terms of examining the performance of the model with the EXPUH linear method, the obtained results are consistent with the results of Sadeghi et al. (2015) and Lotfi Rad et al. (2015) and the model with the EXPUH linear method. The NS criteria has shown acceptable performance. According to the results of the model in the calibration section, in terms of evaluation criteria NS, KGE and , and in terms of simulation of peak flow values and the time to peak using EXPUH method in the linear part showed  better performance than ARMAX method. The value of these criteria in EXPUH method is equal to 0.86, 0.93, and 0.86 and in ARMAX method are equal to 0.7, 0.85 and 0.73, respectively. In the validation section, the evaluation criteria in EXPUH method were equal to 0.51, 0.63, and 0.54 and in ARMAX method were equal to 0.55, 0.73 and 0.65, respectively, indicating better performance of the model by ARMAX method. Comparison of the EXPUH method, and also the model with ARMAX method showed more accurate performance in terms of peak discharges, quantity and time of occurrence. The values of NS, KGE and evaluation criteria in this section were 0.51, 0.63, and 0.54 using EXPUH method and 0.55, 0.73 and 0.65 with ARMAX method, respectively.ConclusionAccording to the results, the IHACRES model using ARMAX method in the linear section resulted in more accurate performance than EXPUH method in simulation of peak flow values and time to peak.