“…The general applicability of these relationships has been challenged by several studies (Adams & Gelhar, ; Cortis & Berkowitz, ; Gómez‐Hernández & Wen, ; Lee et al, ; Liu et al, ; Molinari et al, ; Sánchez‐Vila et al, ; Tompson et al, ; Wen & Gómez‐Hernández, ; Zheng et al, ; Zheng & Gorelick, ; Zinn & Harvey, ), showing that the stochastic approach may be especially questionable in highly heterogeneous aquifers in which the K field structure is characterized by connected features of extreme values. This type of structure cannot be adequately represented by a MVN field (Journel & Deutsch, ), as connected features exert a strong control on groundwater flow and transport behavior in alluvial systems (Anderson, ; Bianchi et al, ; Bianchi & Zheng, ; Davies & Gibling, ; de Marsily et al, ; Dell'Arciprete et al, ; Fogg, ; Fogg et al, ; Fogg & Zhang, ; Weissmann et al, ).…”