Dezvoltarea Func?iilor de Pedotransfer pentru Estimarea unor Propriet??i ale Solului Folosind Re?eaua Artificial? Neural? ?i Abord?rile Regresiei Multivariate

作者: M. Sadeghnejad , P. Pezeshki , F. Sarmadian , A Keshavarzi

DOI:

关键词:

摘要: Studiul propriet??ilor solului ca de exemplu capacitatea câmpului (F.C.) ?i gradul ofilire permanent (P.W.P) joac? un rol important in studiul umidit??ii a curbei reten?ie. De?i ace?ti parametrii pot fi m?sura?i direct, m?surarea lor este dificil? scump?. Func?iile pedotransfer (PTF) asigur? o alternativ? prin estimarea parametrilor din date privitoare la sol care sunt mai rapid disponibile. In aceast? investiga?ie, 70 probe au fost colectate orizonturi diferite câte 15 profile localizate regiunea Ziaran, provincia Qazvin, Iran. Datele divizate dou? subseturi calibrare (80%) testare (20%) ale modelelor normalitatea testat? metoda Kolmogorov-Smirnov. Ambele regresii multivariate tehnicile re?elei neurale artificiale (ANN) utilizate pentru dezvolta PTF-ul corespunz?tor predic?ia folosind caracteristici u?or m?surabile argil?, mâl, O.C., S.P., B.D. CaCO3. Performan?a regresiei modelele ANN evaluate set test independent. Pentru evalua modelele, eroarea mijloc radicalului valorii ridicate p?trat (RMSE) R2 folosite. Compararea RSME-ului cele modele men?ionate ar?tat c? modelul d? estim?ri bune F.C. P.W.P. decât regresie multivariat?. Valorile RMSE deriv? (2,35, 0,77) respectiv (2,83, 0,72). corespunz?toare multivariat? t (4,46, 0.68) (5,21, 0,64). Rezultatele cu cinci neuroni strat ascuns avut performan?? bun? previziunea regresia

参考文章(36)
Budiman Minasny, Alex B. McBratney, Keith L. Bristow, Comparison of different approaches to the development of pedotransfer functions for water-retention curves Geoderma. ,vol. 93, pp. 225- 253 ,(1999) , 10.1016/S0016-7061(99)00061-0
L. Cavazza, A. Patruno, E. Cirillo, Field capacity in soils with a yearly oscillating water table Biosystems Engineering. ,vol. 98, pp. 364- 370 ,(2007) , 10.1016/J.BIOSYSTEMSENG.2007.07.001
Ferhat Karaca, Bestamin Özkaya, NN-LEAP: A neural network-based model for controlling leachate flow-rate in a municipal solid waste landfill site Environmental Modelling and Software. ,vol. 21, pp. 1190- 1197 ,(2006) , 10.1016/J.ENVSOFT.2005.06.006
Bruno De Vos, Marc Van Meirvenne, Paul Quataert, Jozef Deckers, Bart Muys, Predictive Quality of Pedotransfer Functions for Estimating Bulk Density of Forest Soils Soil Science Society of America Journal. ,vol. 69, pp. 500- 510 ,(2005) , 10.2136/SSSAJ2005.0500
A. Mermoud, D. Xu, Comparative analysis of three methods to generate soil hydraulic functions Soil & Tillage Research. ,vol. 87, pp. 89- 100 ,(2006) , 10.1016/J.STILL.2005.02.034
Budiman Minasny, Alex. B. McBratney, The Method for Fitting Neural Network Parametric Pedotransfer Functions Soil Science Society of America Journal. ,vol. 66, pp. 352- 361 ,(2002) , 10.2136/SSSAJ2002.0352
JHM Wösten, A Lilly, A Nemes, Christine Le Bas, None, Development and use of a database of hydraulic properties of European soils Geoderma. ,vol. 90, pp. 169- 185 ,(1999) , 10.1016/S0016-7061(98)00132-3
C. Manyame, C.L. Morgan, J.L. Heilman, D. Fatondji, B. Gerard, W.A. Payne, Modeling hydraulic properties of sandy soils of Niger using pedotransfer functions Geoderma. ,vol. 141, pp. 407- 415 ,(2007) , 10.1016/J.GEODERMA.2007.07.006