The spatial variability of the soil surface roughness was analysed by instruments of different precision degrees in a grid of data points. For measuring the roughness it was used a laser reliefmeter before and after application of different amounts of rain (0 mm, 10 mm, 55 mm, 85 mm) in the laboratory. At field conditions (pasture and ploughed surface) a pin reliefmeter was used at four different dates. The results showed a great spatial dependence of soil surface roughness in both methods and for all treatments. The great variation of semivariogram parameters obtained with pin reliefmeter data for all different dates was due to the distribution and size of soil aggregates. This indicates how difficult it is to obtain a representative tridimensional microrelief model at field conditions. All other semivariograms obtained with laser reliefmeter data (simulated surface) were similar among themselves, indicating that despite the alterations in the soil surface roughness, the spatial dependence pattern did not change with different amounts of rain.
microrelief; scaled semivariogram; geostatistics; soil aggregates
