Journal of hydrology vol:263 issue:1-4 pages:217-229
Spatial variability of surface hydraulic properties is an important factor for infiltration and runoff processes in agricultural fields. At 120 locations on a hillslope, steady-state infiltration rates were measured at two applied pressure heads with a single-ring infiltrometer. The solution of two steady-state infiltration equations for each location (the simultaneous-equation approach, SEA) yielded 41 negative alpha-values, 79 positive alpha values and 120 positive K-fs-values. The sensitivity of K-fs and alpha to small measurement errors was estimated using a Monte-Carlo simulation (MC). Results of this MC simulation showed that the uncertainty on alpha is extremely high while the uncertainty on K-fs is fairly small. Hence, although the pressure infiltrometer technique as applied here is useful to estimate K-fs at each measurement location, it is not suited for the estimation of an alpha-value at each measurement location. A new procedure is proposed for the simultaneous estimation of one overall 'field alpha' and the 79 K-fs values of measurement locations having a positively calculated a using SEA. Using this field alpha, K-fs values for the other locations with a negative a are hence determined. Finally, the spatial correlation of K-fs on the hillslope is examined. Ranges of ln(K-fs) between 2.85 and 3.8 m were observed, respectively, for the omnidirectional case and the y direction along the hillslope mu. (C) 2002 Elsevier Science B.V. All rights reserved.