作者: M. Mohanty , Nishant K. Sinha , D. K. Painuli , K. K. Bandyopadhyay , K. M. Hati
DOI: 10.1007/S40009-015-0358-4
关键词:
摘要: Soil field capacity (FC) and permanent wilting point (PWP) are important input parameters in many bio- physical models. Although these can be mea- sured directly, their measurement is quite difficult expensive. Pedotransfer functions (PTFs) provide an alternative by estimating soil from more readily available data. A study has been conducted to evaluate PTFs of FC PWP created using artificial neural net- works (ANNs). total 721 different sampling locations spread all over India selected develop ANN. Results indicate that six neurons hidden layers best suited for prediction PWP. The statistical criteria (value R 2 , RMSE, MBE, ME, d) used ANN, indicated unbiased higher pre-