Modelling available water capacity of topsoil in a Bayesian paradigm

https://doi.org/10.1016/j.envsoft.2019.104500Get rights and content

Highlights

  • •

    New pedotransfer models are developed in a Bayesian framework for topsoil AWC.

  • •

    Silt has the greatest influence on AWC – both inside and outside Bayesian framework.

  • •

    The relative importance of sand and organic matter on AWC PTFs are region-dependent.

Abstract

Available water capacity (AWC) is a fundamental factor in energy-water nexus. Particularly, AWC influences climate dynamics, hydrological processes, and water management. Measuring AWC is often laborious, time consuming, and sometimes impossible. Pedotransfer functions (PTFs) are therefore used as alternatives to model AWC. Classical statistical-based pedotransfer functions are widely used to model AWC. However, these classical statistical models do not incorporate prior knowledge of the existing AWC data. Therefore, the objective of this study is two-fold: Firstly, to develop AWC PTFs for topsoil using five physical soil properties; and secondly, to evaluate relative importance of the physical soil properties in the Bayesian-based AWC model. Results show that Bayesian PTFs plausibly simulate AWC with percent bias (PBIAS) of <1%; root mean square error (RMSE) of 0.021 to 0.027 cm/cm; and degree of agreement (d1) of 0.6–0.79. Silt had the greatest influence on the AWC, both inside and outside Bayesian framework.

Section snippets

Software and/or data availability

The following is a summary of the source code for developing pedotransfer functions (PTFs) of soil available water capacity (AWC) in a Bayesian framework. The code can be freely downloaded from https://github.com/kachiengz/BPTFs.

  • Name of CodeBayesian-based pedotransfer functions (BPTFs)

  • DeveloperKevin O. Achieng

  • Contact addressUniversity of Wyoming, Department of Civil and Architectural Engineering, 1000 E. University Ave., Laramie, WY 82071, United States

  • Telephone number+1 3O7.761.3OO7

  • E-mail [email protected]

Study area and data

This study focuses on topsoils in the Ogallala Aquifer. This aquifer transverses eight states of the US, and it is the largest aquifer in North America – covering an area of 451000 km2. The aquifer supports about 30% of US irrigated agriculture; 97% of irrigation water supply within the aquifer (Brown and Pervez, 2014); and fresh drinking water to about 80% of people living within its vicinity (Dennehy, 2000). Since planning of irrigation across this area requires soil AWC, reliable measurement

AWC and its explanatory variables

The study area from which the soil data was obtained, and for which the Bayesian-based PTFs was developed, is shown in Fig. 1. A summary of the physical soil properties can be found in Fig. 2. The violin-plots shown in Fig. 2 portray both distribution and the statistics (with respect to percentiles i.e. 25th, 50th, 75th, minimum, and maximum) of the physical soil properties across the six regions within the study area. Accordingly, the AWC seems to be decreasing southwards from region1 to

Conclusions

Bayesian-based AWC pedotransfer functions were developed for the Ogallala Aquifer topsoil, which supports most irrigated agriculture, especially in the western USA. This study is an improvement over the commonly used classical statistical regression analysis to develop PTFs. This is because of the fact that the Bayesian framework allows for incorporation of prior knowledge of the distribution and statistics of the physical soil parameters into the PTFs. Since the study area lies within

Acknowledgements

Special thank you goes to the Paul A. Rechard Fellowship of The University of Wyoming, for financial support. The anonymous reviewers and the editor have provided valuable comments throughout the review process of the original manuscript.

References (63)

  • M. Román Dobarco et al.

    Pedotransfer functions for predicting available water capacity in French soils, their applicability domain and associated uncertainty

    Geoderma

    (2019)
  • K. Sreelash et al.

    Estimation of available water capacity components of two-layered soils using crop model inversion: effect of crop type and water regime

    J. Hydrol.

    (2017)
  • J.H.M. Wösten et al.

    Pedotransfer functions: bridging the gap between available basic soil data and missing soil hydraulic characteristics

    J. Hydrol.

    (2001)
  • C.J. Willmott

    Some comments on the evaluation of model performance

    Bull. Am. Meteorol. Soc.

    (1982)
  • P. Aina et al.

    Estimating available water-holding capacity of western Nigerian soils from soil texture and bulk density, using core and sieved samples

    Soil Sci.

    (1985)
  • R.G. Allen et al.

    Crop Evapotranspiration -Guidelines for Computing Crop Water Requirements -FAO Irrigation and Drainage Paper 56

    (1998)
  • S. Assouline et al.

    The concept of field capacity revisited: defining intrinsic static and dynamic criteria for soil internal drainage dynamics

    Water Resour. Res.

    (2014)
  • J.F. Brown et al.

    Merging remote sensing data and national agricultural statistics to model change in irrigated agriculture Merging remote sensing data and national agricultural statistics to model change in irrigated agriculture

  • C.F. Chiu et al.

    Estimation of water retention curve of granular soils from particle-size distribution — a Bayesian probabilistic approach

    Can. Geotech. J.

    (2012)
  • W. Cui et al.

    Empirical Bayes vs. Fully Bayes Variable Selection

    (2004)
  • J.L. De Paepe et al.

    Simple estimation of available water capacity in soils of semiarid and subhumid environments

    Arid Land Res. Manag.

    (2018)
  • J.L. De Paepe et al.

    Arid Land Research and Management Simple estimation of available water capacity in soils of semiarid and subhumid environments

  • K.F. Dennehy

    High Plains Regional Ground-Water Study: U.S. Geological Survey Fact Sheet FS-091-00

    (2000)
  • T.S. Eicher et al.

    Determining Growth Determinants: Default Priors and Predictive Performance in Bayesian Model Averaging (No. 76)

    (2007)
  • C. Fernández et al.

    Model uncertainty in cross-country growth regressions

    J. Appl. Econom.

    (2001)
  • C. Fernández et al.

    Bayesian modelling of catch in a north-west Atlantic fishery

    J. R. Stat. Soc.: Ser. C Appl. Stat.

    (2002)
  • E.I. George et al.

    Calibration and empirical Bayes variable selection

    Biometrika

    (2000)
  • A.K. Guber et al.

    Multimodel simulation of water flow in a field soil using pedotransfer functions

    Vadose Zone J.

    (2009)
  • S.C. Gupta et al.

    Estimating soil water retention characteristics from particle size distribution, organic matter percent, and bulk density

    Water Resour. Res.

    (1979)
  • H.V. Gupta et al.

    Toward improved calibration of hydrologic models: multiple and noncommensurable measures of information

    Water Resour. Res.

    (1998)
  • R.B. Jana et al.

    Multiscale pedotransfer functions for soil water retention

    Vadose Zone J.

    (2007)

    • Cited by (3)

    • Pedotransfer functions and machine learning: Advancements and challenges in tropical soils

      2023, Geoderma Regional

    • Accounting for the spatial range of soil properties in pedotransfer functions

      2023, Geoderma

    • Shallow aquifer monitoring using handpump vibration data

      2020, Journal of Hydrology X
      Citation Excerpt :

      This framework is further detailed in Section 2. Machine learning approaches have been explored to understand hydrological processes and facilitate water resource management (Herrera et al., 2010; Yaseen et al., 2019; Achieng, 2019). These approaches have also been used to model variations in aquifer level based on hydro-climatic data (e.g., rainfall, temperature) (Nayak et al., 2006; Behzad et al., 2010; Yoon et al., 2011; Taormina et al., 2012; Tapoglou et al., 2014; Suryanarayana et al., 2014; Sahoo et al., 2017).

    View full text
    Published by Elsevier Ltd.