تخمين معكوس ضريب زبري مانينگ با استفاده از مدل WinSRFR و بررسي تغييرات آن در رخدادهاي مختلف آبياري

Inverse estimation of Manning roughness coefficient using WinSRFR model and investigating its variations in different irrigation events

وش ﻫﺎي ﻣﺨﺘﻠﻔﯽ ﺑﻪ ﻣﻨﻈﻮر ﺗﺨﻤﯿﻦ ﺿﺮﯾﺐ زﺑﺮي ﻣﺎﻧﯿﻨﮓ ﺗﻮﺳﻌﻪ ﯾﺎﻓﺘﻪ اﺳﺖ ﮐﻪ از اﯾﻦ ﺑﯿﻦ روش ﻣﻌﮑﻮس ﮐﺎرﺑﺮد ﻓﺮاواﻧﯽ دارد. اﯾﻦ ﺗﺤﻘﯿﻖ ﺑﺎ ﻫـﺪ ﺗﺨﻤﯿﻦ ﻣﻌﮑﻮس ﺿﺮﯾﺐ زﺑﺮي ﻣﺎﻧﯿﻨﮓ و ﺿﺮاﯾﺐ ﻧﻔﻮذ ﻣﻌﺎدﻟﻪ ﮐﻮﺳﺘﯿﺎﮐﻒ-ﻟﻮﺋﯿﺲ ﺑﺎ اﺳﺘﻔﺎده از ﻣﺪل WinSRFR در آﺑﯿﺎري ﺟﻮﯾﭽﻪ اي و ﺑﺮرﺳﯽ ﺗﻐﯿﯿﺮات آن ﻫﺎ در آﺑﯿﺎري ﻫﺎي ﻣﺨﺘﻠﻒ اﻧﺠﺎم ﺷﺪ. ﺑﺪﯾﻦ ﻣﻨﻈﻮر دو دﺑﯽ ورودي )ﺑﻪ ﻃﻮر ﻣﺘﻮﺳﻂ 0/27 و 0/53 ﻟﯿﺘﺮ ﺑﺮ ﺛﺎﻧﯿﻪ(، دو دور آﺑﯿﺎري )5 و 10 روز(، دو ﻣﺰرﻋﻪ ﺑﺎ ﺑﺎﻓﺖ ﺧﺎك ﻣﺘﻔﺎوت )E و F( و در ﺳﻪ واﻗﻌﻪ آﺑﯿﺎري )اول ﺗﺎ ﺳﻮم( و در ﺳﻪ ﺗﮑﺮار در ﻧﻈﺮ ﮔﺮﻓﺘﻪ ﺷﺪ. ﺧﻄﺎي ﻣﺪل در ﺗﺨﻤﯿﻦ زﻣﺎن ﭘﯿﺸﺮوي، زﻣﺎن ﭘﺴﺮوي و ﺣﺠﻢ آب ﻧﻔﻮذ ﯾﺎﻓﺘﻪ ﺑﺎ اﺳﺘﻔﺎده از ﺷﺎﺧﺺ ﻫﺎي آﻣﺎري ﺧﻄﺎي ﻧﺴﺒﯽ )RE(، رﯾﺸﻪ دوم ﻣﯿﺎﻧﮕﯿﻦ ﻣﺠﺬور ﺧﻄﺎ )RMSE( و رﯾﺸﻪ دوم ﻣﯿﺎﻧﮕﯿﻦ ﻣﺠﺬور ﺧﻄﺎي ﻧﺮﻣﺎل ﺷﺪه )NRMSE( ﺑﺮرﺳﯽ ﺷﺪ. ﻣﻘﺎدﯾﺮ اﯾﻦ ﺷﺎﺧﺺ ﻫﺎ در ﺗﺨﻤﯿﻦ زﻣﺎن ﭘﯿﺸﺮوي در ﻣﺰرﻋﻪ E ﺑﻪ ﺗﺮﺗﯿﺐ 0/44 درﺻﺪ، 0/21 دﻗﯿﻘﻪ و 1/11 درﺻـﺪ در در مجموع هر سه رخداد آبياري تعيين شد. همچنين مقادير اين شاخص‌ها به ترتيب 1/58 درصد، 5/25 دقيقه و 2/78درصد در زمان پسروي و 0/59 درصد،درﺻﺪ، 5/05 ﻟﯿﺘﺮ و 0/50 درﺻﺪ در ﺣﺠﻢ آب ﻧﻔﻮذ ﯾﺎﻓﺘﻪ ﺑﻪ دﺳﺖ آﻣﺪ. ﺧﻄﺎي ﺗﺨﻤﯿﻦ زﻣﺎن ﭘﯿﺸﺮوي و ﭘﺴﺮوي در ﻣﺰرﻋﻪ F ﮐﻤﺘﺮ ﺑﻮد ﮐﻪ اﯾﻦ ﻧﺘﺎﯾﺞ دﻗـﺖ ﺑﺴﯿﺎر ﺧﻮب روش ﺗﺨﻤﯿﻦ ﻣﻌﮑﻮس را در ﺗﻌﯿﯿﻦ ﺿﺮﯾﺐ زﺑﺮي ﻣﺎﻧﯿﻨﮓ ﻧﻤﺎﯾﺎن ﮐﺮد. ﻧﺘﺎﯾﺞ ﻫﻤﭽﻨﯿﻦ ﻧﺸﺎن داد ﮐﻪ در ﻣﺰرﻋﻪ E ﺣﺪاﻗﻞ و ﺣﺪاﮐﺜﺮ ﻣﻘﺪار ﺿﺮﯾﺐ زﺑﺮي ﻣﺎﻧﯿﻨﮓ ﺑﻪ ﺗﺮﺗﯿﺐ 0/017 و 0/34 و ﺑﻪ ﻃﻮر ﻣﺘﻮﺳﻂ 0/075 در ﻣﺠﻤﻮع ﻫﺮ ﺳﻪ رﺧﺪاد آﺑﯿﺎري ﺑﻮد و در ﻣﺰرﻋﻪ F ﺿﺮﯾﺐ زﺑـﺮ ي ﻣﺎﻧﯿﻨـﮓ در ﻫـﺮ ﺳـﻪ رﺧﺪاد آﺑﯿﺎري ﺣﺪاﻗﻞ 0/015، ﺣﺪاﮐﺜﺮ 0/09 و ﺑﻪ ﻃﻮر ﻣﺘﻮﺳﻂ 0/041 ﺑﺮآورد ﺷﺪ.

Various methods have been developed to estimate the Manning roughness coefficient, of which the inverse method is widely used. This study aimed to estimate the inverse of Manning roughness coefficient using the WinSRFR model and estimating infiltration parameters of Kostiakov-Lewis equation in furrow irrigation and to investigate their variations in different irrigation events. For this purpose, two inflow discharges (average 0.27 and 0.53 liters per second), two irrigation cycles (5 and 10 days), two fields with different soil texture (E and F) in three irrigation events (first to third) were considered in three replications. Model error in estimating advance and recession time, and volume of infiltrated water was investigated using statistical indices including RE, RMSE and NRMSE. The values of these indices in estimating the advance time in the field E were 0.44%, 0.21 min and 1.11% in the total three irrigation events, respectively. Also, the values of these indices were 1.58%, 5.25 min and 2.78% in the recession time and 0.59%, 5.5 liters and 0.50% in the volume of infiltrated water, respectively. There was less error in estimating the advance and recession time in the field F which showed the excellent accuracy of the inverse estimation method using the WinSRFR model in determining the Manning roughness coefficient. The results also showed that in the field E, the minimum and maximum values of Manning roughness coefficient were estimated as 0.017 and 0.34 and on average 0.075 in the total three irrigation events, respectively. In the field F, the minimum, maximum and the average values of Manning roughness coefficient in three irrigation events were 0.015 and 0.09 and 0.041, respectively.