When the predictions of physically-based soil erosion models are validated against measured data, a recurring theme is 'the right answer for the wrong reason'. Such spurious precision can occur in both temporal and spatial senses. For example, an acceptable estimate of long-term average erosion rate may mask unrealistic underlying distributions of annual, monthly or event values (Favis-Mortlock, 1994).
Similarly, a reasonable estimate of sediment yield at a catchment outlet may result from a simulated field-by-field pattern of erosion which is simply wrong (Jetten et al., 1999 in press). It would clearly be unwise to put much faith in other predictions by models which show this kind of behaviour.

The way to guard against such problems is to validate more than one model output against measured data. The relative abundance of time series of measured data for plot- and field-sized areas means that unrealistic temporal distributions of modelled erosion rates are not difficult to identify. It is less easy, however, to validate the spatial patterns of erosion which are estimated by current models, since suitable field measurements are rarer. But we must undertake such validations at a catchment scale, if we are to improve the ability of current models to estimate the soil loss which results from the development and growth of ephemeral gullies (Poesen et al., 1998).

This paper examines the ability of the watershed version of the WEPP model (Flanagan and Nearing, 1995) to estimate the spatial pattern of erosion on a 189 ha agricultural catchment at Breaky Bottom, in the UK South Downs. Fields on this catchment eroded severely in October 1987 during the most serious erosion event to have been recorded in Britain (Boardman, 1988).

References
Boardman, J. (1988). Severe erosion on agricultural land in East Sussex, UK, October 1987. Soil Technology 1, 333-348.
Favis-Mortlock, D.T. (1994). Use and Abuse of Soil Erosion Models in Southern England, Unpublished PhD Thesis, University of Brighton, UK. 310 pp.
Flanagan, D.C. and Nearing, M.A. (1995). USDA Water Erosion Prediction Project: Hillslope Profile and Watershed Model Documentation, NSERL Report No. 10, USDA-ARS National Soil Erosion Research Laboratory, West Lafayette, Indiana, USA.
Jetten, V., de Roo, A. and Favis-Mortlock, D.T. (1999 in press). Evaluation of field-scale and catchment-scale soil erosion models. Catena
Poesen, J., Vandaele, K. and van Wesemael, B. (1998). Gully erosion: importance and model implications. In, Boardman, J. and Favis-Mortlock, D.T. (eds), Modelling Soil Erosion by Water, Springer-Verlag NATO-ASI Series I-55, Berlin. pp. 285-312.