Elsevier

Environmental Modelling & Software

Volume 119, September 2019, Pages 433-444
Environmental Modelling & Software

Crop yield simulation optimization using precision irrigation and subsurface water retention technology

https://doi.org/10.1016/j.envsoft.2019.07.006Get rights and content

Highlights

  • Precision irrigation using subsurface water retention technology (SWRT) is optimized

  • Water and nutrient mobility are simulated using HYDRUS-2D software

  • HYDRUS-2D’s computational complexity is alleviated using a calibration procedure of DSSAT software which is fast

  • A multi-objective optimization method is employed to obtain optimal irrigation practices for minimum water usage and maximize crop growth.

  • This paper depicts how recent computational intelligence methods can be utilized to integrate two irrigation-based simulation software with weather and soil characteristics to obtain two important goals of agriculture practices.

Abstract

Maximizing crop production with minimal resources such as water and energy is the primary focus of sustainable agriculture. Subsurface water retention technology (SWRT) is a stable approach that preserves water in sandy soils using water saving membranes. An optimal use of SWRT depends on its shape, location and other factors. In order to predict crop yield for different irrigation schedule, we require at least two computational processes: (i) a crop growth modeling process and (ii) a water and nutrient permeation process through soil to the root system. Validation of software parameters to suit properties of specific field becomes increasingly hard since they involve a coordination with field data and coordination between two software. In this paper, we propose a computationally fast approach that utilizes HYDRUS-2D software for water and nutrient flow simulation and DSSAT crop simulation software with an evolutionary multi-objective optimization (EMO) procedure in a coordinated manner to minimize water utilization and maximize crop yield prediction. Our proposed method consists of training one-dimensional crop model (DSSAT) on data generated by two dimensional model calibrates and validates (HYDRUS-2D), that accounts for water accumulation in the SWRT membranes. Then we used DSSAT model to find the best irrigation schedules for maximizing crop yield with the highest plant water use efficiency (Tambussi et al., 2007; Blum, 2009) using for the EMO methodology. The optimization procedure minimizes water usage with the help of rainfall water and increases corn yield prediction as much as six times compare to a non-optimized and random irrigation schedule without any SWRT membrane. Our framework also demonstrates an integration of latest computing software and hardware technologies synergistically to facilitate better crop production with minimal water requirement.

Introduction

In today's world, optimal coordination of food, energy and water (FEW) nexus has become an international priority and necessity (Verma, 2015). To achieve such a task, sustainable technologies for crop production by using minimal water and energy has become essential (Wallace, 2000). National and international projections identify 65% more food and biomass production will be required to support global human populations approaching 9.6 billion by requiring 50% more irrigation water by 2050 (How to feed the world in, 2050). Water is vital for irrigation of crops but water is also scarce and today's food and biomass producers are obligated to use judicial consumption of water for irrigation (Greenland et al., 2018).

In order to have a sustainable crop production system, we need to use minimal water yet holding most of them in the soil for plant growth. Sandy soil has much less holding capacity and high hydraulic conductivity at least for high water contents (Rawls et al., 1991). There are some other agronomic solution to the problem of water sparing such as choice of genotype, time of sowing, soil and soil nutrient management, soil organic carbon management, mulching, etc (Neupane and Guo, 2019). Researchers have also used machine learning methods on soil-moisture data collected from remote sensors to facilitate precision irrigation (McCarthy et al., 2014; Goldstein et al., 2018). Traditional approach using asphalt barrier has proved to be efficient and is widely accepted for sandy soil (Guber et al.,2015; Tutum et al., 2015). Other technologies are also introduced in literature to hold water in sandy soil (Yang et al., 2014; Bruun et al., 2014). Being costly and labor intensive, new polyethylene membranes are also used. Recently Subsurface Water Retention Technology (SWRT) (Smucker, 2011)) has been developed and commercialized to improve soil water holding capacity in plant root zone. The proper membrane design and installation depth in specific soil and weather condition has been studied in (Guber et al.,2015) using two-dimensional modeling of water flow using HYDRUS-2D software (Simunek et al., 2012) in sandy soils. They have investigated a profile distribution of water in a lysimeter filled with pure sand with installed SWRT membranes at different depths. Based on their experiments, it is evident that HYDRUS-2D model with membrane geometry highly reliable estimating water content as in practice with a considerable accuracy. It is also established that SWRT technology can reduce deep drainage and increase water availability in time by helping in retaining water close to the root zone thereby helping to increase crop production. It has been shown that the membrane based water retention technology (Smucker, 2011) in which bowl-shaped troughs of impermeable membranes (Fig. 1) are placed at a certain depth below the soil surface in a systematic staggered manner increases crop yields between 1.4 and 3.4-fold (Smucker et al., 2014) compare to not using a membrane.

Coupling the shape and placement of SWRT membranes with prescriptive irrigation and fertilization schedules are vital parameters for achieving an optimal crop yield. In a previous study (Tutum et al., 2015), water-flow and nutrient transport simulation model using HYDRUS-2D was combined with an evolutionary multi-objective optimization (EMO) algorithm (Deb, 2001) to obtain optimal membrane geometry and placement in soil profile along with prescriptive irrigation scheduling under two conflicting objectives. The study revealed a number of insights about the worthwhile sizes of the membranes and useful supply of irrigation water for achieving certain level of water at the soil root zone.

Although HYDRUS-2D can predict the water and nutrient accumulation at the root zone of a plant, it cannot simulate the crop growth, which is a direct measurable outcome of the irrigation process, that we are interested in maximizing. Besides a continuous supply of water at the root level either through an optimal irrigation pattern or through rainfall, the growth of crop and eventual crop yield depend on many other factors, such as incoming solar energy, plant transpiration rate, temperature, type of crop, etc. Thus, to have a better estimate of crop growth, it is necessary to take help of another computational simulator that can explicitly provide an estimate of crop yield for given soil-water mix, nutrient content, and other parameters in a time-series manner. For this purpose, we use DSSAT (Decision Support System for Agrotechnology Transfer) software, which is a widely accepted tool for agronomists (Hoogenboom et al., 2015). However, DSSAT uses a one-dimensional approach to water flow modeling, and thus cannot simulate SWRT membranes, due to their two (cylindrical) or three-dimensional (spherical) shapes.

We need at least two stand-alone software, namely HYDRUS-2D and DSSAT in order to simulate SWRT membrane in soil and analyze the predict for crop production. Specifically HYDRUS-2D should be used for two-dimensional water flow simulation, while DSSAT for modeling plant growth and assessment of crop yield. HYDRUS-2D is computationally expensive to run, whereas DSSAT is many times faster due to their one-dimensional modeling. Since an optimization process usually requires many iterations of different solutions to be evaluated during a run, we need to devise an efficient optimization strategy which will use a few calls to the HYDRUS-2D software. All these issues pose a great challenge to computing, plant science, and agricultural engineering researchers to devise an efficient methodology which uses each of the two software effectively and produce reliable and accurate results revealing the underlying optimal irrigation pattern and SWRT shapes for known variation of weather, soil, and other environmental conditions.

The aim of this paper is to introduce a mathematical modeling that performs crop yield simulation under different SWRT membranes by using two different software. The paper provides a methodology to optimize parameters of SWRT, namely, shape and location to achieve better crop yield prediction. It can also be used to find an irrigation pattern that minimizes predicted water usage. With the help of computational resources, the proposed method can be easily extended to different soil, crops, weather conditions that will provide useful information before any cultivation is carried out. The rest of the paper is organized as follows. In Section 2, we make a brief introduction about the background of this problem and discuss two software systems mentioned above. In Section 3, we discuss our calibration and validation process of both software in details. In Section 4, optimization methodology for maximizing crop yield and water usage is presented. Section 5 discusses our experimental results. Section 6 concludes the findings of this paper with a direction for future studies.

Section snippets

Background

In this section, we introduce two software systems used in this study: HYDRUS and DSSAT. HYDRUS-2D/3D software was developed to simulate two and three-dimensional movement of water, heat, and multiple solutes by solving Richards Equation for saturated-unsaturated water flow, and the Fickian-based convection-dispersion equation for heat and solute transport (Simunek et al., 2012). Fig. 1 shows a 2D mesh design of a SWRT membrane with water content values (after irrigation) on the right. We

Material and method for calibration

In this section, we describe the DSSAT training on HYDRUS-generated SWC dynamics. HYDRUS-2D is computationally expensive to run and its use to evaluate every solution processed within an optimization run would be an expensive proposition. Although we first attempt to negotiate the computational time by using a parallel computing platform, the crux of this paper lies in ways of simulating the results of HYDRUS-2D using a less expensive DSSAT simulation process with derived DSSAT parameter

Material and method for optimization

Our first objective is to predict crop yield given by DSSAT measured in kilogram per hectare. The second objective is the water use efficiency (WUE) (Tambussi et al., 2007; Blum, 2009) (Equation 3) – the ratio of utilized water by plants and total amount of supplied irrigation. The upper limit of WUE is 1.0, when the water is solely used by the plants, which is an idealistic situation. The upper limit of corn production is hard to predict. Thus, to measure the quality of optimized solutions, we

Simulation results

In this section, we summarize our results. First, we present computationally fast procedure of prior study. Then we show training-validation results. After that the results from the overall optimization procedure are demonstrated.

Summary and future work

In this paper, we have proposed a computational approach to find an optimum irrigation schedule that not only minimizes water usage but also maximizes crop production. In order to simulate soil water movement under embedded SWRT membranes, we have used HYDRUS-2D simulation software in a two-dimensional setting. Since HYDRUS-2D cannot simulate crop growth directly, another software DSSAT is used to predict the crop yield. Between the two software, HYDRUS-2D simulation is exceedingly more time

Acknowledgments

This material is based in part upon work supported by the National Science Foundation under Cooperative Agreement No. DBI-0939454. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.

References (36)

  • E.W. Bruun et al.

    Biochar amendment to coarse sandy subsoil improves root growth and increases water retention

    Soil Use Manag.

    (2014)
  • M. Cisty et al.

    A two-stage evolutionary optimization approach for an irrigation system design

    J. Hydroinf.

    (2017)
  • M. de Paly et al.

    Optimal irrigation scheduling with evolutionary algorithms

  • K. Deb

    Multi-objective Optimization Using Evolutionary Algorithms

    (2001)
  • K. Deb et al.

    Simulated binary crossover for continuous search space

    Complex Syst.

    (1995)
  • K. Deb et al.

    Analysing mutation schemes for real-parameter genetic algorithms

    Int. J. Artif. Intell. Soft Comput.

    (2014)
  • K. Deb et al.

    A fast and elitist multiobjective genetic algorithm: nsga-II, Evolutionary Computation

    IEEE Trans.

    (2002)
  • K. Deb et al.

    A taxonomy for metamodeling frameworks for evolutionary multi-objective optimization

    IEEE Trans. Evol. Comput.

    (2017)
  • Cited by (0)

    View full text