Bridging rigorous assessment of water availability from field to catchment scale with a parsimonious agro-hydrological model
Graphical abstract
Introduction
Crop simulation models integrate various processes in the soil-crop-atmosphere continuum that determine crop growth and production. Hence, they are useful tools to investigate management strategies to optimize crop productivity and resource use efficiency. Such investigations usually focus on one individual field because of the point-based nature of crop models. However, optimization of the use of resources, particularly water, is not a local issue. A management strategy that optimizes crop water productivity in one farmer's field, may only be successful if it does not negatively affect neighbouring farmers. On an even larger scale, agricultural water management affects a whole catchment where different stakeholders, including households, industry and ecosystems, with different goals are making use of the available water resources (Bergez et al., 2012). It is clear that management strategies that are optimized for crop water productivity by a crop model, may fail to result in sustainable water use because catchment processes are disregarded.
Hydrological models, by contrast, simulate hydrological processes in a catchment and simulate crop transpiration as a part of the catchment soil water balance. However, as these models primarily focus on the simulation of hydrological processes, they rarely consider crop growth and management practices affecting crop transpiration and production explicitly. The hydrological models that do include physically based equations to estimate crop transpiration, such as the (semi)-distributed SWAT (Arnold et al., 1998, Douglas-Mankin et al., 2010), MIKE SHE (Refsgaard and Storm, 1995) and APEX (Gassman et al., 2010) models, show relatively high computational complexity. Moreover, they require a vast amount of data and elaborate calibration, or make use of parameters that are difficult to measure in the field. Despite the trend to apply remote sensing data as input or calibration data for agro-hydrological models (Boegh et al., 2004, Moulin et al., 1998), data availability remains a widespread issue (Grayson et al., 2002). Consequently, the application of such data-demanding models renders time- and resource-consuming, or even unfeasible in data-scarce regions.
These limitations of existing crop and hydrological models urge for another approach. A coupling between both types of models, combining their advantages and functionality, can be a solution to obtain agro-hydrological models that (i) simulate crop production and water productivity at field scale, as well as upscale their effects on hydrological processes and water availability at catchment scale, (ii) consider the effect of management and environmental changes on crop transpiration, crop (water) productivity and catchment hydrological processes, (iii) are parsimonious, i.e. require a feasible amount of easily obtainable input data and parameters to be calibrated, without compromising much the accuracy of the model results, and (iv) are widely applicable to various agricultural catchments with different environmental and agronomic conditions.
Previous attempts have been made to couple crop and hydrological models to capitalize the strengths of both and enable accurate investigation of agricultural management and environmental changes within a catchment. The WOFOST crop model (Boogaard et al., 2014) has been coupled to MetaSWAP (van Walsum and Supit, 2012) and to the distributed WEP-L model (Jia, 2011) for climate change impact assessment. Also, the DAISY crop model (Abrahamsen and Hansen, 2000) has been combined with MIKE SHE for investigation of nitrogen fluxes in agricultural catchments (Styczen and Storm, 1993, Thorsen et al., 2001). DSSAT crop models (Jones et al., 2003) have been linked to hydrological models to optimize irrigation management and drainage design (McNider et al., 2014, Singh and Helmers, 2008). Also extensive modeling systems, which integrate all aspects, dimensions, scales and actors involved in agricultural management, link crop and hydrological models (Jakeman and Letcher, 2003, Letcher et al., 2006).
However, most of these model combinations fail to fit all four above mentioned criteria. Being based on the distributed physically based model MIKE SHE, high data requirements remain an issue for the DAISY-MIKE SHE model (Boegh et al., 2004, Thorsen et al., 2001). The same is true for agro-hydrological models based on the data-demanding DSSAT crop models (Jones et al., 2003) and the fully integrative modelling systems (Jakeman and Letcher, 2003). Moreover, when developed for a specific application, the existing model combinations are only applicable for a certain region or crop (McNider et al., 2014). Also, problems to accurately represent spatial heterogeneity within the catchment due to the fixed model structure or grid size of the sub-models should be mentioned (Bithell and Brasington, 2009, Thorsen et al., 2001).
Therefore, the aim of this study was to develop a parsimonious, physically sound and widely applicable agro-hydrological model, AquaCrop-Hydro, to simulate crop productivity and water availability in agricultural catchments without vast data requirements for model input and calibration. The new model was developed by extending the AquaCrop crop water productivity model (Hsiao et al., 2009, Raes et al., 2009, Steduto et al., 2009, Vanuytrecht et al., 2014a) with a lumped conceptual hydrological model to simulate catchment hydrology. The performance of AquaCrop-Hydro to simulate crop production as well as discharge at the catchment outlet was evaluated for an agricultural catchment in Belgium.
Section snippets
The AquaCrop-Hydro model
Fig. 1 depicts the AquaCrop-Hydro model flowchart. AquaCrop-Hydro is the combination of a crop model operating at field scale and a hydrological model working at catchment scale. The two models are integrated through an off-line, one-directional link, in which the crop model output is used as input for the hydrological model component. Model simulations are conducted on a daily time step.
AquaCrop-Hydro applies a semi-distributed approach, as it requires the catchment area to be divided into
Hydrological model parameters
Table 3 shows the seven hydrological model parameters as calibrated for the Plankbeek catchment. The calibrated deep percolation parameters resulted in a good match between the observed and simulated baseflow-interflow proportion. Comparison for all days with deep percolation resulted in a high R2 of 0.86. The recession constants were taken equal to values obtained when applying the WETSPRO filter (Table 2).
Crop production
Average simulated crop production matched well with the average crop yield reported for
Model approach
This study shows that AquaCrop-Hydro combines functionality of the parsimonious AquaCrop model and VHM conceptual hydrological model in order to upscale simulation of water availability from an individual field to a larger agricultural catchment.
AquaCrop-Hydro is able to simulate both crop (water) productivity as well as catchment hydrological processes. Due to the parsimonious nature of both sub-models, not all processes are described in a detailed physically based manner. Crop growth and
Conclusion
The AquaCrop-Hydro model, newly developed by coupling the crop simulation model AquaCrop to a conceptual hydrological model, is shown to be a parsimonious agro-hydrological model with wide potential application. Next to simulation of crop water productivity and production at field scale, AquaCrop-Hydro simulates the daily catchment soil water balance applying a semi-distributed approach. River discharge at the catchment outlet is simulated using a lumped conceptual hydrological model.
Software availability
The AquaCrop-Hydro model is a combination of the AquaCrop crop water productivity model and a lumped conceptual hydrological model. The latest version of the open access AquaCrop software (version 5.0), together with all information and training material, is freely available at: http://www.fao.org/nr/water/aquacrop.html. The standard AquaCrop software includes a graphical user-interface for data processing, running simulations and visualizing simulation results. The plug-in AquaCrop software,
Acknowledgements
This research was funded by the Research Foundation Flanders. We are grateful to Stien Keunen for helping to test and develop the AquaCrop-Hydro model. Moreover, we would like to thank Mattias Van Opstal, Toon Kerkhofs and Stefaan Dondeyne for their help with data collection. Furthermore, we are grateful for data and information that were provided by the Federal Public Service Economy (FOD economie), the Flemish Environment Agency (VMM), the Flemish Land Agency (VLM) and the Royal
References (97)
- et al.
Comparing AquaCrop and CropSyst models in simulating barley growth and yield under different water and nitrogen regimes. Does calibration year influence the performance of crop growth models?
Agric. Water Manag.
(2015) - et al.
Daisy: an open soil-crop-atmosphere system model
Environ. Model. Softw.
(2000) - et al.
Validation and testing of the AquaCrop model under full and deficit irrigated wheat production in Iran
Agric. Water Manag.
(2011) - et al.
Characterising performance of environmental models
Environ. Model. Softw.
(2013) - et al.
Computer-model tools for a better agricultural water management: tackling managers' issues at different scales – a contribution from systemic agronomists
Comput. Electron. Agric.
(2012) A manifesto for the equifinality thesis
J. Hydrol.
(2006)- et al.
To tie or not to tie ridges for water conservation in Rift Valley drylands of Ethiopia
Soil Tillage Res.
(2012) - et al.
Modelling climate change impacts on and adaptation strategies for agriculture in Sardinia and Tunisia using AquaCrop and value-at-risk
Sci. Total Environ.
(2016) - et al.
Coupling agent-based models of subsistence farming with individual-based forest models and dynamic models of water distribution
Environ. Model. Softw.
(2009) - et al.
Incorporating remote sensing data in physically based distributed agro-hydrological modelling
J. Hydrol.
(2004)
Assessing the impact of land use change on hydrology by ensemble modeling (LUCHEM). I: model intercomparison with current land use
Adv. Water Resour.
Selecting crop models for decision making in wheat insurance
Eur. J. Agron.
AquaCrop-OS: an open source version of FAO's crop water productivity model
Agric. Water Manag.
Using AquaCrop to derive deficit irrigation schedules
Agric. Water Manag.
Advances in the use of observed spatial patterns of catchment hydrological response
Adv. Water Resour.
Integrated assessment and modelling: features, principles and examples for catchment management
Environ. Model. Softw.
The DSSAT cropping system model
Eur. J. Agron.
Validating FAO AquaCrop using Landsat images and regional crop information
Agric. Water Manag.
An integrated modelling toolbox for water resources assessment and management in highland catchments: model description
Agric. Syst.
Statistical and graphical methods for evaluating solute transport models: overview and application
J. Contam. Hydrol.
AquaData and AquaGIS: two computer utilities for temporal and spatial simulations of water-limited yield with AquaCrop
Comput. Electron. Agric.
Improving agricultural water productivity: between optimism and caution
Agric. Water Manag.
River flow forecasting through conceptual models part I — a discussion of principles
J. Hydrol.
Evaluation and inter-comparison of global climate models' performance over katonga and ruizi catchments in lake victoria basin
Phys. Chem. Earth Parts ABC
Assessing the performance of the FAO AquaCrop model to estimate maize yields and water use under full and deficit irrigation with focus on model parameterization
Agric. Water Manag.
Revising the potential of large-scale Jatropha oil production in Tanzania: an economic land evaluation assessment
Energy Policy
Cereal yield stabilization in Terai (Nepal) by water and soil fertility management modeling
Agric. Water Manag.
A model-independent open-source geospatial tool for managing point-based environmental model simulations at multiple spatial locations
Environ. Model. Softw.
Assessment of uncertainty in simulation of nitrate leaching to aquifers at catchment scale
J. Hydrol.
Sowing and irrigation strategies for improving rainfed tef (Eragrostis tef (Zucc.) Trotter) production in the water scarce Tigray region, Ethiopia
Agric. Water Manag.
Intercomparison of five lumped and distributed models for catchment runoff and extreme flow simulation
J. Hydrol.
AquaCrop: FAO’S crop water productivity and yield response model
Environ. Model. Softw.
Global sensitivity analysis of yield output from the water productivity model
Environ. Model. Softw.
Considering sink strength to model crop production under elevated atmospheric CO2
Agric. For. Meteorol
Comparing climate change impacts on cereals based on CMIP3 and EU-ENSEMBLES climate scenarios
Agric. For. Meteorol
Performance assessment of the FAO AquaCrop model for irrigated cabbage on farmer plots in a semi-arid environment
Agric. Water Manag.
Assessing crop water productivity from field to regional scale
Agric. Water Manag.
Parsimonious rainfall–runoff model construction supported by time series processing and validation of hydrological extremes – Part 1: step-wise model-structure identification and calibration approach
J. Hydrol.
A time series tool to support the multi-criteria performance evaluation of rainfall-runoff models
Environ. Model. Softw.
Parsimonious rainfall-runoff model construction supported by time series processing and validation of hydrological extremes – Part 2: intercomparison of models and calibration approaches
J. Hydrol.
Sowing strategies for barley (Hordeum vulgare L.) based on modelled yield response to water with AquaCrop
Exp. Agric.
Grootschalig Referentiebestand
Digitaal Hoogte Model Vlaanderen (5 M)
Bodemgebruiksbestand, Opname 2001
Crop Evapotranspiration: guidelines for Computing Crop Water Requirements
All that glitters is not gold: the case of calibrating hydrological models: invited Commentary
Hydrol. Process
Large area hydrologic modeling and assessment part I: model development
J. Am. Water Resour. Assoc.
WOFOST Control Centre 2.1; User's Guide for the WOFOST Control Centre 2.1 and the Crop Growth Simulation Model WOFOST 7.1.7
Cited by (11)
A monthly distributed water and salt balance model in irrigated and non-irrigated lands of arid irrigation district with shallow groundwater table
2023, Journal of HydrologyCitation Excerpt :To account for the impact of anthropogenic activities on hydrological processes in arid irrigation districts, modules describing irrigation, drainage, and/or crop growth have also been added to traditional hydrological models to study hydrological processes in irrigation districts (e.g., Christiansen et al., 2004; Ahmadzadeh et al., 2016; Malek et al., 2017; Xiong et al., 2019). Moreover, some agro-hydrological models for the cropland have also been extended to the irrigation district or catchment scale for distributed simulation of water flow and/or salt transport processes, such as Hydrus (Van Gaelen et al., 2017), SWAP (Xue and Ren, 2016; Xu et al., 2019), RZWQM2 (Xi et al., 2017), SaltMod (Singh, 2012), and SahysMod (Singh and Panda, 2012). The above distributed agro-hydrological models have been widely used, but they may have some problems when applied to arid irrigation districts with shallow groundwater table.
A monthly distributed agro-hydrological model for irrigation district in arid region with shallow groundwater table
2022, Journal of HydrologyCitation Excerpt :However, there are several limitations in these available agro-hydrological models (Siad et al., 2019; Singh, 2019b), especially when applied to irrigation district with shallow groundwater table depth. First, lateral groundwater flow and salt transport between different landscapes, especially those between irrigated and non-irrigated lands, are generally ignored in most of the models not coupled with a regional groundwater model (Van Gaelen et al., 2017; Xu et al., 2019). These models are applicable to irrigation districts with deep groundwater table, but not reliable in irrigation districts with shallow groundwater table where both vertical and horizontal groundwater flow plays an important role in the regional hydrological cycle.
The integration of land use and climate change risks in the Programmes of Measures of River Basin Plans – assessing the influence of the Water Framework Directive in Portugal
2019, Environmental Science and PolicyCitation Excerpt :The Water Framework Directive (WFD) (2000/60/EC of 23.10) aimed to promote sustainable use of water, by requiring the Member States to implement River Basin Management Plans (RBMPs). These have motivated a wide variety of research works and approaches including comparative studies (Kallioras et al., 2006; Kochskaemper et al., 2016; Zandvoort et al., 2017), country-focused studies (Gonzalez del Tánago et al., 2012; Richter et al., 2013; Junier and Mostert, 2014; Feuillette et al., 2016; Van Gaelen et al., 2017; Pappalardo et al., 2017), national or transboundary river basins focused studies (Demetropoulou et al., 2010; Grindlay et al., 2011; Kaspersen et al., 2016) or even smaller scale case studies (Franzén et al., 2015; Newig et al., 2016; Lopes et al., 2017). Many of these have been publicized in the news alert’ EU Science for Environment Policy service.
Exploring parsimonious daily rainfall-runoff model structure using the hyperbolic tangent function and Tank model
2019, Journal of HydrologyCitation Excerpt :Parsimonious mechanistic, as opposed to statistical, models employ only the parameters needed to predict hydrological variables of interest at the required level of accuracy (Atkinson et al., 2002; Barrett and Charbeneau, 1997; Basu et al., 2010; Jackson-Blake et al., 2017; Pande et al., 2012; Perrin et al., 2003; Valentin et al., 2018). Their structures and simulation strategies are often simple enough to ignore spatial variability in landscape and hydrological processes happening in a watershed (Clark et al., 2008; Fenicia et al., 2010; Kavetski and Fenicia, 2011; van Esse et al., 2013; Vansteenkiste et al., 2014; Van Gaelen et al., 2017). Thus, the parameters of parsimonious models tend to be conceptual or empirical, and it is not possible to estimate their spatially averaged values by either measurement or prior estimation, which makes parameter calibration a necessary and critical part of simulations using a parsimonious model (Hrachowitz et al., 2014; Perrin et al., 2001).
Spatially distributed crop model based on remote sensing
2019, Agricultural Water ManagementCitation Excerpt :FAO released AquaCrop-GIS by merging the AquaCrop model with GIS software to promote the model for regional applications (Lorite et al., 2013). The traditional AquaCrop-GIS model for regional applications considers the spatial variation in soil, climate, crop, and management measures (Li et al., 2016; Van Gaelen et al., 2017). The parameters of the model tend to vary with the change of natural environment in which crops grow.