Pedotransfer Functions for Estimating Soil Hydraulic Properties: A Review
KRISHI: Publication and Data Inventory Repository
View Archive InfoField | Value | |
Title |
Pedotransfer Functions for Estimating Soil Hydraulic Properties: A Review
Not Available |
|
Creator |
Nitin Gorakh Patil and Surendra Kumar Singh
|
|
Subject |
database, generic PTF, hydraulic conductivity, predictor properties, PTF development tools, regression, statistical indices, water retention curve
|
|
Description |
Not Available
Characterization of soil hydraulic properties is important to environment management; however, it is well recognized that it is laborious, time-consuming and expensive to directly measure soil hydraulic properties. This paper reviews the development of pedotransfer functions (PTFs) used as an alternative tool to estimate soil hydraulic properties during the last two decades. Modern soil survey techniques like satellite imagery/remote sensing has been used in developing PTFs. Compared to mechanistic approaches, empirical relationships between physical properties and hydraulic properties have received wide preference for predicting soil hydraulic properties. Many PTFs based on different parametric functions can be found in the literature. A number of researchers have pursued a universal function that can describe water retention characteristics of all types of soils, but no single function can be termed generic though van Genuchten (VG) function has been the most widely adopted. Most of the reported parametric PTFs focus on estimation of VG parameters to obtain water retention curve (WRC). A number of physical, morphological and chemical properties have been used as predictor variables in PTFs. Conventionally, regression algorithms/techniques (statistical/neural regression) have been used for calibrating PTFs. However, there are reports of utilizing data mining techniques, e.g., pattern recognition and genetic algorithm. It is inferred that it is critical to refine the data used for calibration to improve the accuracy and reliability of the PTFs. Many statistical indices, including root mean square error (RMSE), index of agreement (d), maximum absolute error (ME), mean absolute error (MAE), coefficient of determination (r2) and correlation coefficient (r), have been used by different researchers to evaluate and validate PTFs. It is argued that being location specific, research interest in PTFs will continue till generic PTFs are developed and validated. In future studies, improved methods will be required to extract information from the existing database. Not Available |
|
Date |
2024-07-01T10:49:39Z
2024-07-01T10:49:39Z 2016-03-31 |
|
Type |
Research Paper
|
|
Identifier |
Nitin Gorakh Patil and Surendra Kumar Singh
Not Available http://krishi.icar.gov.in/jspui/handle/123456789/83810 |
|
Language |
English
|
|
Relation |
Not Available;
|
|
Publisher |
Not Available
|
|