Use of caesium-137 data for validation of spatially distributed erosion models: the implications of tillage erosion

Validation of spatially distributed models using spatially distributed data represents a vital element in the development process; however, it is rarely undertaken. To a large extent, this reflects the problems associated with assembling erosion rate data, at appropriate temporal and spatial scales and with a suitable spatial resolution, for comparison with model results. The caesium-137 (¹³⁷Cs) technique would appear to offer considerable potential for meeting this need for data, at least for longer timescales. Nevertheless, initial attempts to use ¹³⁷Cs for model validation did not prove successful. This lack of success may be explained by the important role of tillage erosion in redistributing soil within agricultural fields and, therefore, contributing to the ¹³⁷Cs-derived soil redistribution rates. This paper examines the implications of tillage erosion for the use of ¹³⁷Cs in erosion model validation and presents an outline methodology for the use of ¹³⁷Cs in model validation. This methodology acknowledges and addresses the constraints imposed by the need to: (1) separate water and tillage erosion contributions to total soil redistribution as represented in ¹³⁷Cs derived rates; (2) account for lateral mixing of ¹³⁷Cs within fields as a result of tillage translocation; (3) simulate long-term water erosion rates using the model under evaluation if ¹³⁷Cs-derived water erosion rates are to be used in model validation. The methodology is dependent on accurate simulation of tillage erosion and tillage translocation. Therefore, as greater understanding of tillage erosion is obtained, the potential for the use of ¹³⁷Cs in water erosion model validation will increase. Caesium-137 measurements remain one of the few sources of spatially distributed erosion information and, therefore, their potential value should be exploited to the full.