Variability of a Double Ring Infiltration Test

Abstract

Fifty-eight double-ring infiltrometer tests were conducted in a Yolo loam soil near the University of California, Davis campus to investigate the spatial variability of water infiltration data and to determine the minimum sample size necessary to determine infiltration parameters with a desired level of confidence. The cumulative infiltration was represented by: Q = A0(l e ct) + A1t in all cases tested, where Q is cumulative infiltration in mm; t is time in h; A0 (in mm), C (1/h), and A1(mm/h) are empirical coefficients. The parameters A0, C, and A1 were found to be normally distributed. Only the parameter A1 exhibited spatial variation. A sample size of 10 can be used to estimate the population mean of infiltration parameters A0, C, and A1 with 90% confidence and an accuracy of 9.7%, 15.5%, and 14%, respectively, provided the spatial variation in A1 is properly treated.