No abstract

The famous French mathematician Henri Poincaré said that "Science is built up of facts, as a house is with stones. But a collection of facts is no more a science than a heap of stones is a house" (Poincaré, 1902). Our perceptual understanding of runoff processes sometimes appears as a heap of facts (or observations), with many of them dating back to the First International Hydrological Decade and the following years, when key observations were made regarding translatory flow (Hewlett & Hibbert, 1967), saturation excess overland flow (Dunne & Black, 1970), saturated wedges (Weyman, 1973), old water effusion (Martinec, 1975), the importance of topographic convergence (Anderson & Burt, 1978) and preferential flow (Mosley, 1979). Since then, field hydrologists have continued to document more and more runoff responses in more and more areas.What we have learned from these myriad studies is that (i) runoff processes are scale dependent, making the "up-scaling" of field-based process observations very difficult to nearly impossible, (ii) runoff concepts developed for humid areas (e.g., the variable source area concept) often do not apply in semi-arid and arid areas, and (iii) runoff responses are non-linear (thresholds, hysteresis, etc.) because regardless of whether the dominant flowpaths occur at the surface as overland flow or as subsurface stormflow within the soil or subsoil or saprolite or bedrock, some storage needs to be filled before runoff occurs.Our heap of facts about runoff generation obtained by process observations have continued to grow, but there has been little construction of these facts into a body of knowledge. This seems to be a crucial next step in catchment hydrology, as it has been in other areas of science where observations gain significance when one uses reason to group them in meaningful ways (Beveridge, 1950). Indeed, "grouping" is a theme in the development of many fields; notably Darwin's (1859) observation that "science consists of grouping facts so that general laws or conclusions may be drawn from them" and Hughlings-Jackson's (1882) advice that we have multitudes of facts but "we require, as they accumulate, organizations of them into higher knowledge; we require generalizations and working hypotheses."

Abstract. Beaver ponds are surface-water features that are transient through space and time. Such qualities complicate the inclusion of beaver ponds in local and regional water balances, and in hydrological models, as reliable estimates of surface-water storage are difficult to acquire without time-and labour-intensive topographic surveys. A simpler approach to overcome this challenge is needed, given the abundance of the beaver ponds in North America, Eurasia, and southern South America. We investigated whether simple morphometric characteristics derived from readily available aerial imagery or quickly measured field attributes of beaver ponds can be used to approximate surface-water storage among the range of environmental settings in which beaver ponds are found. Studied were a total of 40 beaver ponds from four different sites in North and South America. The simplified volume-area-depth (V-A-h) approach, originally developed for prairie potholes, was tested. With only two measurements of pond depth and corresponding surface area, this method estimated surface-water storage in beaver ponds within 5 % on average. Beaver pond morphometry was characterized by a median basin coefficient of 0.91, and dam length and pond surface area were strongly correlated with beaver pond storage capacity, regardless of geographic setting. These attributes provide a means for coarsely estimating surface-water storage capacity in beaver ponds. Overall, this research demonstrates that reliable estimates of surfacewater storage in beaver ponds only requires simple measurements derived from aerial imagery and/or brief visits to the field. Future research efforts should be directed at incorporating these simple methods into both broader beaver-related tools and catchment-scale hydrological models.