|Title||Support vector machine for the prediction of future trend of Athabasca River (Alberta) flow rate|
|Source||Geological Survey of Canada, Open File 6864, 2017, 31 pages, https://doi.org/10.4095/299739 (Open Access)|
|Publisher||Natural Resources Canada|
|Related||This publication is related to Chen, G; Chen, Z; (2017).
Extreme river flow prediction for river water supply to oil sands mining sites, Athabasca River near Fort McMurray, Alberta,, Geological Survey of Canada, Open File 6863|
|NTS||73L; 73M; 74D; 82N; 82O; 83B; 83C; 83D; 83E; 83F; 83G; 83H; 83I; 83J; 83K; 83N; 83O; 83P; 84A; 84B; 84C|
|Area||Athabasca River; Fort McMurray; Clear Water River; Edson Creek; Slave Lake; Whitecourt|
|Lat/Long WENS||-120.0000 -110.0000 57.0000 51.5000|
|Subjects||hydrogeology; mathematical and computational geology; Nature and Environment; modelling; climate; climatic fluctuations; temperature; precipitation; hydrologic properties; hydrologic environment;
surface waters; rivers; climate change; hydrology; river flow rates; support vector machine (SVM); water supply|
|Illustrations||tables; location maps; time series|
|Program||Geoscience for New Energy Supply (GNES), Program Coordination|
|Released||2017 03 07|
River flow process usually was affected by a wide variety of factors and it is very difficult to make the trend prediction. Various mathematical techniques have been developed to
tackle the prediction, but these techniques are less accurate compared with physically-based models. In this paper I presented the recurrent support vector machine methods to study a series of climate variables, such as temperature and precipitation,
and then predict the future trend of flow river rate based on these climate variables. The Athabasca River data have been tested and the flow river rates in 100 years have been predicted.
|Summary||(Plain Language Summary, not published)|
Many factors contribute to the fluctuation of river flow rate through time. Complicated mathematical models based on physical principles are difficult to
apply in many cases due to limitation of data and constraints of resources. This open file reports a method that is based on the support vector machine to predict river flow rate using series of relevant climate variables, such as temperature and
precipitation only, as model inputs. The historical climate data from Athabasca River were used to demonstrate the application of the proposed method, as an example, for future river flow prediction.