Location: National Soil Erosion ResearchTitle: Quantification of soil surface roughness evolution under simulated rainfall) Author
|Huang, Chi Hua|
Submitted to: Transactions of the ASABE
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 2/1/2013
Publication Date: 2/1/2013
Citation: Vermang, J., Norton, L.D., Baetens, J.M., Huang, C., Cornelis, W.M., Gabriels, D. 2013. Quantification of soil surface roughness evolution under simulated rainfall. Transactions of the ASABE. 56(2):501-514. Interpretive Summary: Soil surface roughness is known to influence many processes occurring on the soil surface, including infiltration, depression storage, runoff, flow velocity and meandering, soil loss and sediment deposition. With the advancement of laser scanning technology, we can now obtain very detailed surface microtopographic data and one challenging question is how can we mathematically express soil roughness so surfaces with different roughness elements or micromorphology can be distinguished. To quantify soil roughness, we not only need to quantify the elevation change, we also need to know the change spatially. In this research, we created surfaces with different roughness with different sizes of soil clods and we exposed the surfaces to a simulated rain storm. Soil surface elevations before and after the rainstorm were measured by a laser scanner at millimeter grids. We tested several soil roughness indices to see which one is more sensitive to different roughness surfaces. These indices included random roughness, variogram sill and range, fractal dimension and fractal length using a fractional Brownian motion (fBm) model, variance and correlation length according to a Markov Gaussian model and fractal dimension using the Revised Triangular Prism surface area Method (RTPM). The RTPM procedure uses square prisms to represent the surface morphology and the logarithmic correlation between the prism size and the total prism surface area can be used as an index for soil roughness. We found random roughness to be the best estimator that significantly distinguishes soil surface roughness classes. Nevertheless, the random roughness does not describe the spatial dependency. When we examine all the indices representing the spatial elevation change, the RTPM covered a greater range of scales, and hence is a better estimator of the overall roughness. Furthermore, the RTPM roughness index reflects the self-similar fractal character of the surface microtopography and is easier to calculate than the fBm fractal dimension, making this method a valuable tool to characterize soil surface roughness
Technical Abstract: Soil surface roughness is commonly identified as one of the dominant factors governing runoff and interrill erosion. The objective of this study was to compare several existing soil surface roughness indices and to test the Revised Triangular Prism surface area Method (RTPM) as a new approach to calculate the fractal dimension as a roughness index. A silty clay loam soil was sampled, sieved to four aggregate sizes and each were packed in soil trays in order to derive soils from four soil surface roughness classes. Rainfall simulations using an oscillating nozzle simulator were conducted for 90 min at 50.2 mm.h-1 average intensity. The surface microtopography was digitized by an instantaneous profile laser scanner before and after the rainfall application. Calculated roughness indices included random roughness, variogram sill and range, fractal dimension and fractal length using a fractional Brownian motion (fBm) model, variance and correlation length according to a Markov Gaussian model and fractal dimension using the RTPM. Random roughness proved to be the best estimator to significantly distinguish soil surface roughness classes. When taking spatial dependency into account, the variogram sill proved to be the best alternative. The fractal dimension as calculated from the fBm model did not yield good results as only short range variations were incorporated. The MG variance proved to describe the large scale roughness better than the parameters of the fBm model did. The fractal dimension from the RTPM performed well, although it could not significantly discriminate between all roughness classes. It covered a greater range of scales, and hence is a good estimator of the overall roughness.