DRAINMOD simulation of macropore flow: model modification and field testing

Authors: ASKAR, MANAL - Michigan State University; YOUSSEF, MOHAMMED - North Carolina State University; CHESCHEIR, GEORGE - North Carolina State University; NEGM, LAMYAA - North Carolina State University; King, Kevin; HESTERBERG, DEAN - North Carolina State University; AMOOZEGAR, AZIZ - North Carolina State University; SKAGGS, RICHARD - North Carolina State University

Submitted to: Agricultural Water Management
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 7/20/2020
Publication Date: 7/29/2020
Citation: Askar, M.H., Youssef, M.A., Chescheir, G.M., Negm, L.M., King, K.W., Hesterberg, D.L., Amoozegar, A., Skaggs, R.W. 2020. DRAINMOD simulation of macropore flow: model modification and field testing. Agricultural Water Management. 242. Article 106401. https://doi.org/10.1016/j.agwat.2020.106401.
DOI: https://doi.org/10.1016/j.agwat.2020.106401

Interpretive Summary: Poorly formed soils or soils low in quality tend to form preferential flow paths that result from natural wetting or drying cycles, old root channels, or biological activity such as worms. In artificial, tile-drained landscapes, these preferential flow paths serve as conduits for both water and pollutants to rapidly move through the soil profile and into tile drains and surface waters. Representing preferential flow paths and water movement has been a major gap in hydrology and water quality simulation technologies. A simple and robust approach for simulating water movement through macropores was developed, incorporated and tested within the DRAINMOD simulation technology. The approach and methodology was generally able to represent measured events from an Ohio field with considerable confidence. Further testing is needed; however, this development shows promise and will permit a more realistic application of the model in tile drained landscapes with preferential flow paths. Furthermore, assessment of water and pollutant movement through macropores will be more accurately represented when assessing different farming practices, benefiting academia, extension, environmental stakeholders, and state and federal agencies.

Technical Abstract: Macropores are critical pathways through which water and pollutants can bypass the soil matrix and be rapidly transported to subsurface drains and freshwater bodies. We modified the hydrology component of the DRAINMOD model to simulate macropore flow using a simple approach as part of evolving the model to simulate phosphorus losses and dynamics in artificially drained agricultural lands. In our approach, the Hagen-Poiseuille's law is used to estimate the preferential flow capacity of macropores. That is, when ponding depths on the soil surface are greater than Kirkham's depth, water is assumed to flow through macropores directly to tile drains without any interaction with the soil matrix. Within the modified DRAINMOD model, macropore size is adjusted based on wet or dry conditions while connectivity is altered by tillage. The model was tested using a data set of 5 years (2013-2017) from a subsurface drained field located in northwest Ohio. The field is very poorly drained and the soils are prone to desiccation cracking. Results from the modified model showed an average monthly Nash-Sutcliffe efficiency (NSE) of 0.57 and average annual percent error (PE) of -6.5 %. Also, the new macropore component was able to capture 85% of 80 peak drainage flow events. However, in our simulations, surface runoff was over predicted for the entire study period. Annual water budgets using measured data (precipitation, subsurface drainage, and surface runoff) and model predictions (evapotranspiration, deep seepage, and change in storage) had an average annual imbalanced of 11.2 cm. A substantial lack of closure (25%) in the water balance for 2017 suggests that errors may have occurred in the field measurements. Overall, incorporating preferential flow into DRAINMOD is critical for modeling phosphorus and other strongly bound soil chemicals transport through the soil profile and additional field-scale simulations would further validate the modeling approach and test the underlying assumptions.