Title :
Spatial Prediction of Dissolved Organic Carbon Using GIS and ANN Modeling in River Networks
Author :
Fu, Yingchun ; Zeng, Xiantie ; Lu, Xueyu
Author_Institution :
Sch. of Geogr., South China Normal Univ., Guangzhou, China
Abstract :
That GIS-based hydrological response units (HRUs) incorporated watershed variables and their potential spatial correlation into ANN modeling was clarified in the paper. The process and final results of neural network modeling were both assessed by the deterministic or statistical methods, spatial regression kriging (RK). The relation of prediction errors and HRUs area scale can provide useful information to optimize the design of stream monitoring network. It is indicated that potential advantage of ANN for watershed and the assessment of estuarine river impacts can be done by precise spatial prediction and sensitive factors analysis.
Keywords :
correlation methods; environmental factors; geographic information systems; neural nets; regression analysis; rivers; ANN modeling; GIS-based hydrological response units; deterministic methods; dissolved organic carbon spatial prediction; estuarine river impact assessment; neural network modeling; prediction errors; river networks; sensitive factors analysis; spatial correlation; spatial regression kriging; statistical methods; stream monitoring network; watershed variables; Artificial neural networks; Biological neural networks; Carbon; Correlation; Neurons; Rivers; Soil; artificial neural network (ANN); dissolved organic carbon (DOC); hydrological response units (HRUs); regression kriging(RK);
Conference_Titel :
Computational Intelligence and Security (CIS), 2011 Seventh International Conference on
Conference_Location :
Hainan
Print_ISBN :
978-1-4577-2008-6
DOI :
10.1109/CIS.2011.96
