“…As noted earlier, recent applications of hydrologic similarity methods have shown that it is possible to reduce run times by 2 to 3 orders of magnitude, without any loss in information content (Newman et al, 2014;Chaney et al, 2016a). Also, hydrologic similarity concepts can be effectively applied using multiscale methods to resolve the dominant spatial gradients that drive flow; for example, using representative hillslopes to explicitly resolve lateral flow processes (Troch et al, 2003;Berne et al, 2005;Hazenberg et al, 2015;Ajami et al, 2016). In exploring these solutions, we recognize that there is not necessarily a tradeoff between physical realism and computational efficiency -the linkage between spatial complexity and process complexity may be rather weak, as models that are run using a large number of spatial elements may still miss dominant processes (e.g., Hartmann et al, 2017).…”