Artificial neural network(ANN) model of landscape ecological planning in urban fringes
Sun Huiguo1, Xu Jianhua2
1. Institute of Geology andGeophysics, CAS, Beijing 100029; 2. Dept. of Geography, East China Normal University; the open Laboratory of Urban & Environment Geocomputation, PRC's Ministry of Education, Shanghai 200062
Landscape ecological planning is an important field of landscape ecological application. In this paper, the Back-Propagation Network model was first introduced to landscape ecological planning in urban fringes. The author applied elevation, elevation deviation, slope, slope deviation, relief zonation, distance to the Y.R, the number of residential area in every sample as the model's input variables, and chose PD, MPFD, SHDI, CONTAG as the model's output variables. Twenty elaborately selected samples were used to train the model. The result showed that the model had a good performance, hence it gave a new simulation way to analyze landscape ecological planning.
[22] Forman R. 1995. Land mosaics, the ecology of landscapes and regions[M]. Cambridge:Cambridge University Press. 30~126.
[23] Goodchild M. F., Steyaert L. T. et al. 1996. GIS and Environ mental Modeling[J]. Progress and Research Issues, GIS Word, Inc.
[24] Kerkstra K, Vrijlandr P. 1990. Landscape Planning for industrial agriculture:a proposed framework for rural areas[J]. Landscape and Urban Planning, 18:275~287.
[25] Li H, Reynolds J F. 1994. A simulation experiment to quantify spatial patterns of landscapes[J]. landscape Ecology, 8:155~162.
[26] Silverman, R.H. and Noetzel, A.S. 2002. Image Processing and Pattern Recognition in Ultrasonograms by Backpropagation[J]. Neural Networds, 1:63~68.
[27] Wu J, Gao W, Tueller P T. 1997. Effects of changing spatial scale on the results of statistical analysis with landscape data:A case study[J]. Geographic Information Sciences, 3:30~41.