干旱区生态环境知识资源中心(Arid): Monthly streamflow forecasting using artificial intelligence approach: a case study in a semi-arid region of India

Arid

DOI	10.1007/s12517-021-08778-6
	Monthly streamflow forecasting using artificial intelligence approach: a case study in a semi-arid region of India
	Sharma, Priyanka; Madane, Dnyaneshwar; Bhakar, S. R.; Sharma, Survey D.
通讯作者	Sharma, P (corresponding author), Natl Inst Hydrol, Groundwater Hydrol Div, Roorkee 247667, Uttarakhand, India.
来源期刊	ARABIAN JOURNAL OF GEOSCIENCES
ISSN	1866-7511
EISSN	1866-7538
出版年	2021
卷号	14 期号:22
英文摘要	Accurate and reliable streamflow forecasting is paramount in the field of water resource planning and management, especially in semi-arid regions. However, streamflow time series are highly complex and non-linear in nature; traditional or physical-based models may fail to capture the complexity and maintain the robustness of the datasets. Therefore, the present study aims to improve the forecasting accuracy and reduce the uncertainty in the datasets by using the data-driven approach such as artificial neural network (ANN) that can efficiently handle the non-linearity in the large and complex hydrological data. This study method includes two steps: i.e., first to develop the ANN models using different combinations of inputs such as rainfall, temperature, and streamflow lag by one or two and then to validate the developed models to forecast the streamflow by using a total of four performance evaluation indices such as correlation coefficient (R), root mean square error (RMSE), modified Nash-Sutcliff efficiency (MNSE), and modified index of agreement (MIA). The proposed method is demonstrated in Jakham reservoir located in Pratapgarh district, Rajasthan, India, for improving the accuracy of monthly streamflow forecasting over a 40-year period (1975-2015). We found that increasing the number of input parameters improves the accuracy of the model and enhance its performance. According to the results, the ANN models 5 and 6 (M-5 and M-6) showed significant variation in the performance evaluation criteria. This clearly indicates that ANN model with an input combination of lag one or two streamflow (i.e., model M-5 and M-6) is performed better when compared to a model that incorporates only monthly rainfall and monthly lag one or two rainfall as inputs. Overall, the application of ANN models M-5 and M-6 (with lag one and two streamflow as an input) can forecast monthly streamflow forecasting with better accuracy.
英文关键词	ANN Rainfall Temperature Streamflow Forecasting
类型	Article
语种	英语
收录类别	SCI-E
WOS记录号	WOS:000718011600013
WOS关键词	NEURAL-NETWORK ; MODELS ; PREDICTION ; YIELD
WOS类目	Geosciences, Multidisciplinary
WOS研究方向	Geology
资源类型	期刊论文
条目标识符	http://119.78.100.177/qdio/handle/2XILL650/373823
作者单位	[Sharma, Priyanka; Sharma, Survey D.] Natl Inst Hydrol, Groundwater Hydrol Div, Roorkee 247667, Uttarakhand, India; [Madane, Dnyaneshwar] Punjab Agr Univ, Dept Soil & Water Engn, Ludhiana 141027, Punjab, India; [Bhakar, S. R.] MPUAT, Dept Soil & Water Engn, CTAE, Udaipur 313001, Rajasthan, India
推荐引用方式 GB/T 7714	Sharma, Priyanka,Madane, Dnyaneshwar,Bhakar, S. R.,et al. Monthly streamflow forecasting using artificial intelligence approach: a case study in a semi-arid region of India[J],2021,14(22).
APA	Sharma, Priyanka,Madane, Dnyaneshwar,Bhakar, S. R.,&Sharma, Survey D..(2021).Monthly streamflow forecasting using artificial intelligence approach: a case study in a semi-arid region of India.ARABIAN JOURNAL OF GEOSCIENCES,14(22).
MLA	Sharma, Priyanka,et al."Monthly streamflow forecasting using artificial intelligence approach: a case study in a semi-arid region of India".ARABIAN JOURNAL OF GEOSCIENCES 14.22(2021).