Characterizing the Variability of Soil Denitrification

Abstract

It is generally recognized that quantification of the spatial variability associated with soil processes is important, and over the past 5–10 years a wide variety of fairly novel statistical techniques have been applied to the study of soil variables in an attempt to characterize variability including i) geostatistics (Burgess and Webster, 1980; Vieira et al., 1983; Vauclin et al., 1983), ii) spectral and autocorrelation analysis (Kachanoski et al., 1985a; 1985h), iii) the state space approach (Morkoc et al., 1985), and iv) bootstrapping (Dane et al., 1986). The spatial variability exhibited by soil denitrification is among the highest reported for soil processes, with coefficients of variation in the range of 100–500% (Folorunso and Rolston, 1984; 1985; Parkin et al., 1987; Myrold, 1988). While there has been much emphasis on applying novel techniques to characterize variability, less emphasis has been placed on trying to answer the question, “How does characterizing variability really help us understand the process under study?”