Deriving tillage-controlled runoff patterns for agricultural fields

On agricultural fields, management (especially tillage) operations with a distinct orientation often lead to a corresponding preferred orientation of surface runoff and associated sediment transport. When deriving surface properties like flow directions and slope for runoff modelling from digital elevation model (DEM) data with grid sizes larger than 1m, these features of the surface will usually remain undetected and by default predict runoff and sediment transport patterns based on the topographic slope and flow directions alone.

A methodology proposed by (Takken et al., 2001) involves calculating 1) topographic slope and flow directions and 2) slope and flow directions assuming that surface runoff takes place exclusively along the tillage orientation. A decision algorithm then decides for each grid cell, whether 1) or 2) is to be used, based on cell slope, oriented roughness and the angle between topographic and tillage-controlled flow directions. An exception is made for distinct thalweg situations, where 1) is always used.

For larger areas, where the manual assignment of the management direction of individual fields (e.g. based on orthophotos) is not feasible, automatic estimation of a field’s tillage orientation is done using field geometry parameters and assuming tillage taking place in the direction of the longest field extent.

The output of the methodology is to be used subsequently in grid-based soil erosion modelling and is expected to provide more realistic results of surface runoff and soil loss patterns. Initial tests using the output flow directions and slope of the method as input for an MMF (Morgan-Morgan-Finney) based soil erosion model in a small experimental catchment (0.7 km²) show surface runoff and soil loss concentrating on the field borders (headlands) for some fields, potentially leading to a shift of priority for protection of either whole individual fields or particularly affected portions of fields.

The improved modelling results can in some situations be significant for decisions on the placement of best management practices (BMP) that intend to limit either soil loss from the field or sediment input into adjacent surface water bodies (e.g. vegetated filter strips, grassed waterways or the feasibility of contouring), since these measures might be rendered useless, when their placement is based on topographic flow directions alone, as is the default practice.