Analysis of soil roughness measurements using a 25 m laser profiler and a 4 m wide meshboard

The currently available backscatter inversion models often do not yield satisfactory results when applied to natural surfaces due to an inaccurate characterization of the soil roughness properties. This is caused by inadequacy of the applied soil roughness measuring techniques and by the high variation on soil roughness parameters within one agricultural field. From a theoretical study it was found that long profiles are necessary to obtain an accurate estimation of the roughness parameters. Short profiles, which are generally measured, tend to underestimate the roughness parameters. Soil roughness data were collected over three different tillages with the ESA laser profiler, which allows measurements of profiles up to 25 m, and a 4 m wide meshboard. This research compares the two measuring techniques and studies the possibility of compensating for the underestimation of short profiles. Analysis of the 25 m laser profiles confirmed the theoretical study that the rms height and the correlation length increase as profile length increases. The rms height reached a constant value for all three tillages whereas the correlation length only reached a constant value for the roughest tillages. The dependency on profile length of both rms height and correlation length could be quite accurately modeled which allowed the extrapolation from short profile data to rms height and correlation length values at higher profile lengths. Additionally, it was found that it is possible to convert the less accurate meshboard derived rms values to those obtained from the laser profiler. This conversion greatly improves when the parameter values of at least six, 4 m profiles are averaged. For the correlation length this conversion is not possible due to the high variance on this parameter