Evaluation of the Gray Model GM(1,1) Applied to Soil Particle Distribution
- Cheng-Mau Wua,
- Jet-Chau Wen *b and
- Kou-Chiang Changc
- a Research Center for Soil and Water Resources and Natural Disaster Prevention, National Yunlin Univ. of Science and Tech., 123, Section 3, University Rd., Douliou, Yunlin County 64002, Taiwan, R.O.C.
b Dep. of Safety Health and Environ. Engineering, National Yunlin Univ. of Science and Tech., 123, Section 3, University Rd., Douliou, Yunlin County 64002, Taiwan, R.O.C.
c Water Resources Agency, Ministry of Economic Affairs, 9-12F, 41-3, Section 3, Hsin-Yi Rd., Taipei 10651, Taiwan, R.O.C.
Particle size distribution (PSD) is a fundamental soil physical property. The conventional approaches for representing PSD use empirical models with two to four parameters. We developed an alternative way to predict PSD that differs from conventional approaches by using the gray model GM(1,1), which does not depend on the model shape as empirical approaches do. The performance of GM(1,1) was compared with Skaggs model by using four statistical criteria. From nine textures of soil samples in our study, the results reveal that the GM(1,1) is superior for making PSD predictions. The results show that for the overall textures, the GM(1,1) model makes better predictions than the Skaggs model except for sand. Therefore, the performance of the GM(1,1) is fairly reliable and efficient and is not affected by soil textures in general.Please view the pdf by using the Full Text (PDF) link under 'View' to the left.
Copyright © 2009.