Calibration of simulated rainfall characteristics for the study of soil erosion on agricultural land

摘要

Rainfall simulation is a widely used method for soil erosion studies on agricultural land. Major problem of this experimental research method is the comparability between different simulators due to differences in simulated rainfall. Therefore the purpose of this study is to characterize the rainfall produced by a rainfall simulator which was widely used during the last decades. Four different calibration methods were used to assess the drop size distribution: (1) Indication Paper, (2) Plaster Micro Plot, (3) Joss-Waldvogel Disdrometer and (4) Laser Distrometer (Thies). Additionally, the latter one was used to measure drop fall velocity in combination with drop diameter. The spatial rainfall distribution pattern on the plot was measured with 100 rainfall gauges. The spatial rainfall distribution pattern clearly shows a heterogeneity, which is caused by the used nozzle configuration. Considerable differences in drop-size distribution can be observed depending on the used measurement technique. Laser Disdrometer and Plaster Micro Plot cover the whole produced drop size spectrum ranging from <0.5mm to >3.0mm, whereas Indication Paper as well as the Joss-Waldvogel Disdrometer primarily show drops smaller than 2.0mm. Characterisation of rainfall is therefore strongly dependent on the used method and if different methods are used, may lead to contradictory results. The volume drop size distribution reflected by the Laser Distrometer is very similar to that one produced by rain with an intensity of 40mmhp#. Nevertheless, with maximum velocities above 10msp# small drops are by far too fast and large drops with velocities dominantly below 5msp# are too slow compared to natural rainfall. As an overall result, the simulator can be characterised as suitable for runoff and infiltration measurements, but with constraints due to the low reproducibility of the spatial rain distribution. As a consequence of the produced drop spectrum and fall velocity the erosion quantities may be underestimated systematically. For this, methodological development has to be focussed on homogeneous spatial rainfall distributions and on increasing the amount of large drops with higher fall velocities.