Research papers
Towards predicting the initiation of overland flow from relatively flat agricultural fields using surface water coverage

https://doi.org/10.1016/j.jhydrol.2021.126125Get rights and content

Highlights:

  • Overland flow from relatively flat land is largely controlled by hydrological connectivity.

  • Hydrological connectivity is difficult to monitor / measure in real time.

  • We propose surface water coverage (ASW) as a proxy for hydrological connectivity.

  • A clear threshold response of overland flow to ASW was observed.

  • ASW has substantial potential for the prediction of overland flow initiation.

Abstract

On rough agricultural soils, initiation of overland flow is primarily related to the gradual filling of small depressions. As the volume of water ponding in local depressions increases, the connectivity of those depressions increases, and that connectivity permits flow across the field boundaries. Previous studies have aimed at predicting overland flow connectivity by means of depression storage, but this is exceedingly difficult to quantify during an irrigation or rainfall event. There has been little work linking overland flow connectivity through variables that are more readily-measured than depression storage, such as the proportion of the soil surface covered in water (ASW).

We propose using ASW as a proxy for hydrological connectivity, which can be measured using proximal remote sensing. A series of overland flow experiments were conducted in two contrasting plots of ~1.5 m2. Outlet discharge and changes in soil surface covered in water were continuously recorded during the experiments. The experiments demonstrated that the rougher soil surface experienced a delayed initiation of overland flow. The results also showed overland flow initiation was characterised by a distinct connectivity threshold and showed a clear response of overland flow to ASW. We further investigated the hydrological connectivity in twenty-one additional agricultural soils with contrasting micro-topographic conditions. Our results suggest that a predictor based on ASW has substantial potential to predict initiation of overland flow. The prediction is sufficiently early that it could be used in modern variable-rate irrigation systems, in combination with a sensor to measure ASW in real time, to prevent substantial flow from fields during irrigation.

Introduction

Sprinkler irrigation, a modern and commonly-used system of irrigating, is a method of applying water in a controlled manner in that is similar to rainfall (Hedley et al., 2014). However, excessive application rates or amounts of irrigation can lead to an accumulation of surface water in small depressions on the soil surface (Appels et al., 2011, Chu et al., 2015). In turn, the accumulation of surface water results in overland flow and a high risk of contaminant loss (Blaustein et al., 2015, Deasy et al., 2009, McDowell et al., 2003, McDowell and Houlbrooke, 2009). Moreover, the infiltration of ponded water in microtopographic depressions can result in increased macropore flow causing greater water losses and leaching (Close et al., 2007). While irrigation systems should be designed to apply water at rates below the infiltration rate of the soil (O’Shaughnessy et al., 2015) to minimise ponding and overland flow, soil surface conditions can change rapidly in space and time (Drewry et al., 2008, Drewry et al., 2019). As a result, even under well-designed irrigation systems, the application rate can exceed the soil’s infiltration rate and result in irrigation-induced overland flow (IOF; see Table 1 for definitions of the terms used here). Once IOF is generated, there is the potential for the rapid mobilisation of nutrients and contaminants from the soil. Although the frequency and size of IOF events can be small compared to those of rainfall-induced overland flow (Laurenson et al., 2018), IOF is similar to rainfall-induced overland flow in having a severe impact on surface water quality because of the relatively large quantity of nutrients and contaminants that can be transported rapidly (Heathwaite et al., 2005, Louchart et al., 2001, Monaghan et al., 2016, van der Salm et al., 2012). Minimising IOF is therefore key in reducing contamination and eutrophication of water bodies and improving the environmental performance of irrigation systems.

Soil surface depressions delay the initiation of overland flow (Chu et al., 2015, Darboux and Huang, 2005, Zhao et al., 2018) and enhance the retention of water that would otherwise run off a field. The soil surface is generally characterised by numerous depressions across scales. Increasing soil roughness tends to decrease overland flow volume (Chu et al., 2015), as well as reducing its velocity (Cogo et al., 1983). As a result, soil erosion and transport of nutrients and contaminants to water bodies are reduced (Govers et al., 2000, Helming et al., 1998). The development of ponding in surface depressions, and the way in which depressions merge and form flow pathways, depends on the spatial organisation of the microtopography of the field. As depressions fill, hydraulic connections are established with neighbouring depressions - the hydrologic connectivity (C) of the field increases - leading to the initiation of overland flow (Antoine et al., 2009). Consequently, overland flow may exhibit a threshold-based runoff response which can be attributed to the filling, spilling and merging processes and related threshold-based overland flow generation (Chu et al., 2015, Chu et al., 2013, Peñuela et al., 2013). A sharp increase in overland flow usually occurs when certain area of the field is connected to the outlet or field boundary (defined as connectivity threshold (CCT; Peñuela et al. (2015)).

Several studies have investigated the filling, spilling and merging of water in soil surface depressions (Antoine et al., 2009, Appels et al., 2011, Chu et al., 2015, Chu et al., 2013, Darboux et al., 2002b, Peñuela et al., 2015, Shaw et al., 2013). Most of these studies have investigated the development of ponding and overland flow generation, often at plot scales, from the perspective of functional hydrological connectivity – the Relative Surface Connection function (RSCf; Antoine et al., 2009, Peñuela et al., 2015). RSCf is a simplification of the runoff hydrograph where the vertical axis represents the instantaneous overland flow rate at the field boundaries normalised by the instantaneous rainfall rate and the horizontal axis represents the depression storage depth (DS) normalised by the maximum depression storage (depression storage capacity, DSC) value yielding the relative depression storage (RDS; 0 ≤ RDS ≤ 1). Due to the assumption of instantaneous flow in the construction of the RSCf, the normalised flow rate is equivalent to the percentage of the surface connected to the lower outlet of the field. One of the main advantages of RSCf is that it provides a functional integrated metric because it explicitly integrates the flow network at the soil surface (Peñuela et al., 2016). The instantaneous flow model is also computationally much faster than a full two-dimensional hydraulic model (e.g. Wu et al. (2020)) and is therefore more suitable for analysing large numbers of elevation points or in circumstances where fast processing times are needed. Peñuela et al. (2013) further showed that, when border effects were negligible, overland flow was initiated at a CCT. In practice, CCT can be defined as the point where the slope of the RSCf equals 1 (Peñuela et al., 2015).The RDS at CCT represents the minimum depth of stored water needed to initiate overland flow.

From the perspective of overland flow prevention and management, CCT could be a decision trigger as it occurs early enough to minimise or prevent substantial IOF. For it to be successful, C needs to be known for the irrigated area, sensors need to be in place to monitor C, and the irrigation equipment needs to be equipped with controls to stop or reduce irrigation in response to sensor observations. Current developments in irrigation technology (e.g. Variable Rate Irrigation; O’Shaughnessy et al. (2015)) can enable such on-the-fly decision making. Although C can be estimated through the measurement of DS using RSCf, quantifying the changes in DS during an irrigation event is exceedingly difficult. Therefore, it would be beneficial to investigate whether DS can be replaced by a more easily quantifiable variable that could, for example, be obtained with proximal remote sensing. A recent study has demonstrated that the proportion of soil surface covered in water (ASW, m2 m−2) can be quantified in real time using proximal remote sensing during an irrigation event (Bradley et al., 2020). It is therefore of interest to investigate whether ASW can be linked to the overland flow generation process that replaces measurement of Ds. As the ASW increases, we hypothesize that overland flow connectivity will also increase and therefore increase the risk of overland flow. While previous work has suggested that such a link could be possible (Peñuela et al., 2015), our aim is to make this explicit.

The overall objective of this study was to examine whether overland flow initiation can be predicted using ASW during an irrigation event on relatively flat areas. In doing so, we first examined the effect of ASW on the overland flow initiation using field experiments We then used the Fast Areal Simulator with Topography for Runoff model (hereafter referred to as FASTR; Appels et al., 2011) to investigate the development of C as a function of ASW for soils of contrasting microtopographic conditions. The aim of the modelling was to provide a measure of C and explore how C may be affected by ASW. We further investigated the hydrological connectivity as a function of ASW in a range of agricultural soils with contrasting micro-topographic conditions. Specifically, we address the following questions: (i) is ASW sufficiently linked to the overland flow generation process that it can be used to predict the initiation of overland flow?, (ii) how variable are the microtopographies of a range of agricultural land cover types in terms of DSC and hydrological connectivity? and (iii) how are the characteristics of microtopography and ASW related to the initiation of overland flow in a range of agricultural land use types? In order to address the above research questions, this study focuses on relatively flat (<2°) agricultural field of varying surface conditions. While naturally hilly areas are more sensitive to erosion and runoff risk is higher, flat lands are relatively easy to farm and often have deep and nutrient-rich soil (Brubaker et al., 1993, Ziadat, 2005) and therefore many agricultural fields around the world are selected for being flat (FAO, 1976). To the best of our knowledge, this is the first attempt to make a quantitative link between overland flow connectivity and the proportion of soil covered in water.

Section snippets

Materials and methods

A series of overland flow experiments were carried out to investigate the relationship between surface water coverage (ASW) and overland flow connectivity (C). The overland flow experiments are described in Section 2.1. To further examine the effects of soil microtopography on C, soil-microtopography was assessed for six additional land cover types across twenty-one locations. Section 2.2 deals with the assessment of soil micro-topography for the six land cover types. FASTR was used to model C.

Results and discussion

We present the results from the overland flow experiments in Section 3.1. Section 3.2 deals with the effects of surface roughness on depression storage capacity for the various agriculture soil cover types. Finally, the effects of surface micro-topography on overland flow connectivity for the various agriculture soils and their implications for irrigation management are presented in Section 3.3.

Conclusions

This study investigated the overland flow connectivity as a function of surface water coverage, and how this connectivity is linked to overland flow generation and dynamics in low slope agricultural fields. In doing so, the FASTR overland flow model was tested using the field data, to examine its capacity to simulate soil depression filling, spilling and merging processes and to quantify the hydrologic connectivity. Our results showed the significant effects of surface-microtopography on the

CRediT authorship contribution statement

Chandra Prasad Ghimire: Methodology, Formal analysis, Writing - original draft. Willemijn M. Appels: Methodology, Validation, Writing - review & editing. Laura Grundy: Investigation. Willis Ritchie: Investigation, Writing - review & editing. Stuart Bradley: Validation, Writing - review & editing. Val Snow: Supervision, Validation, Writing - review & editing.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

This research was completed with funding from the New Zealand Ministry of Business, Innovation and Employment under the programme “Surface- Water Assessment and Mitigation for Irrigation” (contract C10X1708). We are grateful to Reuben Carter, Anna Taylor, Florjan Camlek and Kendal Buchanan for their indispensable help with the overland flow experiment. We thank Martin Bates, Jono Satterthwaite and Matt McEvedy for their help at acquiring 3D soil images from various agricultural fields. We would

References (67)

  • A.L. Heathwaite et al.

    Modelling and managing critical source areas of diffuse pollution from agricultural land using flow connectivity simulation

    J. Hydrol.

    (2005)
  • J. Lampurlanés et al.

    Hydraulic conductivity, residue cover and soil surface roughness under different tillage systems in semiarid conditions

    Soil Tillage Res.

    (2006)
  • R.M. Monaghan et al.

    Pathways of contaminant transfers to water from an artificially-drained soil under intensive grazing by dairy cows

    Agric. Ecosyst. Environ.

    (2016)
  • R.G. Moreno et al.

    Tillage and soil type effects on soil surface roughness at semiarid climatic conditions

    Soil Tillage Res.

    (2008)
  • S.A. O’Shaughnessy et al.

    Dynamic prescription maps for site-specific variable rate irrigation of cotton

    Agric. Water Manag.

    (2015)
  • G. Pardini et al.

    Laser elevation measurements of a smectite-rich mudrock following freeze-thawing and wet-drying cycles

    Soil Technol.

    (1995)
  • A. Peñuela et al.

    How do slope and surface roughness affect plot-scale overland flow connectivity?

    J. Hydrol.

    (2015)
  • Songbai Wu et al.

    Two-dimensional rainfall-runoff and soil erosion model on an irregularly rilled hillslope

    J. Hydrol.

    (2020)
  • Z. Xingming et al.

    Improvements in farmland surface roughness measurement by employing a new laser scanner

    Soil Tillage Res.

    (2014)
  • L. Zhao et al.

    Effect of soil surface roughness on infiltration water, ponding and runoff on tilled soils under rainfall simulation experiments

    Soil Tillage Res.

    (2018)
  • T.M. Zobeck et al.

    Tillage and rainfall effects on random roughness: A review

    Soil Tillage Res.

    (1987)
  • I. Bertol et al.

    Relações da rugosidade superficial do solo com o volume de chuva e com a estabilidade de agregados em água

    Revista Brasileira de Ciência do Solo

    (2006)
  • Bilotta, G.S., Brazier, R.E., Haygarth, P.M. 2007. The Impacts of Grazing Animals on the Quality of Soils, Vegetation,...
  • Bissonnais, Y.l. 1990. Experimental study and modelling of soil surface crusting processes. Catena, Supplement (No....
  • R.A. Blaustein et al.

    Release and removal of microorganisms from land-deposited animal waste and animal manures: A review of data and models

    J. Environ. Qual.

    (2015)
  • Stuart Bradley et al.

    Irrigation control through acoustic proximal sensing of the onset of surface water

    Rem. Sens.

    (2020)
  • S.C. Brubaker et al.

    Soil properties associated with landscape position

    Soil Sci. Soc. Am. J.

    (1993)
  • Xuefeng Chu et al.

    Dynamic puddle delineation and modeling of puddle-to-puddle filling-spilling-merging-splitting overland flow processes.

    (2013)
  • M. Close

    Microbial groundwater quality and its health implications for a border-strip irrigated dairy farm catchment, South Island, New Zealand

    J. Water Health

    (2007)
  • N.P. Cogo et al.

    Effect of crop residue, tillage-induced roughness, and runoff velocity on size distribution of eroded soil aggregates

    Soil Sci. Soc. Am. J.

    (1983)
  • Darboux, F., Davy, P., Gascuel-Odoux, C. 2002a. Effect of depression storage capacity on overland-flow generation for...
  • F. Darboux et al.

    Effects of surface water storage by soil roughness on overland-flow generation

    Earth Surf. Proc. Land.

    (2002)
  • F. Darboux et al.

    Does soil surface roughness increase or decrease water and particle transfers?

    Soil Sci. Soc. Am. J.

    (2005)
  • Cited by (0)

    1

    Presently at: Precision Farming Ltd., Christchurch, New Zealand.

    View full text