We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Are runoff coefficients capable to reflect the coupled effect of upslope topography and land cover in agricultural catchments?
- Authors
Schmaltz, Elmar M.; Krammer, Carmen; Strauss, Peter
- Abstract
The length of a hillslope plays a key role in the estimation of mean annual erosion rates ofagricultural catchments. Two main properties govern the erosive slope length as used in theRevised Universal Soil Loss Equation (RUSLE) and follower models: 1) The beginning of anerosive slope length which is defined as the location where runoff usually starts. 2) Theslope cutoff, which is defined by the locations where either sedimentation occursor where the hillslope is connected to a feature of linear drainage, such as riverchannels or artificial flow paths. Recent studies considered land cover specific runoffcoefficients in the computation of the slope length factor (L-Factor) in the RUSLE, bymeans of adapting the upslope contributing area with effective contributing arearatios for different land cover types, such as forest, grassland, farmland and fallowland.In this study, we assess the effect of land cover specific runoff coefficients in the computationof L-Factor, by calibrating the RUSLE to the small (approx. 66 ha) HOAL-catchment nearPetzenkirchen in Lower Austria. We compare an established grid-based approach forL-Factor computation (LG) and a modified approach (LM) that adapts the upslopecontributing area with land cover specific runoff coefficients. Further, we evaluate theperformance of the RUSLE by comparing the modelling results with measured sedimentloads at the catchment outlet for a period of 11 years (2007 to 2017) and identify particularsensitive model parameters. Lastly, we aim to verify whether the modified approach (LM) iscapable to reflect the coupled effect of upslope topography and vegetation coverreliably.Our findings indicate that the LG-approach largely overestimates annual soil erosion rates byapprox. one order of magnitude of the observed sediment load. In contrast, the LM-approachis able to reduce the absolute predicted soil erosion rate of the LG-approach by thehalf. Both approaches show similar relationships to the observed sediment loadwith R2 = 0.85 (LG) and R2 = 0.86 (LM) and similar trends throughout the yearsof observation. The direct comparison of LG and LM aggregated for the entireobservation period show greatest differences in grassland areas, but surprisingly arecognisable variability of differences among the farmland classes (e.g. LM is ∼30 %lower than LG in areas where cereals are cultivated but ∼50 % lower in areas withmaize). However, the comparison between single plots and each year of observationshows that only differences between land cover classes are observable (grassland,arable land, etc.) but not for single crops, as a direct comparison of LG and LM letassume.Considering the fact that the upslope contributing area implies the highest sensitivity to theL-Factor results, it can be suggested that the LM is not capable to reflect the topographicalvariabilities of surface runoff behaviour in differently managed cropland areas (e.g. cerealscompared to maize). In this regard, it can be questioned whether runoff coefficients areable to represent L-Factor values reliably in differently managed cropland areas.
- Subjects
LOWER Austria (Austria); LAND cover; WATERSHEDS; UNIVERSAL soil loss equation; TOPOGRAPHY; RUNOFF; FARMS; SOIL erosion
- Publication
Geophysical Research Abstracts, 2019, Vol 21, p1
- ISSN
1029-7006
- Publication type
Article