Hydrogeology, Chemical Weathering, and Soil Formation. Allen Hunt

were central to his thesis: first, that there are three basic landscape types relevant to chemical migration and second, that each chemical has a characteristic mobility in the landscape (Polynov, 1935, 1937). His basic landscape types were eluvial, superaqual, and aqual. In eluvial landscapes, the water table is always, or nearly always, below the ground surface; in superaqual landscapes, the water table and the ground surface coincide; in aqual landscapes, free water rests on the land surface as in lakes. From Polynov’s pioneering studies have evolved several conceptual schemes of geochemical landscapes. Notable contributions have come from Mariya A. Glazovskaya (1963, 1968).

      During the 1960s, researchers started to take up Milne’s and Morison’s seminal ideas and investigate soil evolution in the context of hillslopes. Indeed, Vance T. Holliday (2006) argued that Milne’s unique contribution was actually in his linking of soil–catenary patterns to specific slope‐related processes: wetness, solute transport, and erosion and deposition. A consensus grew that soils on lower slopes are potential sumps for the drainage of soils upslope of them (Hallsworth, 1965); that, on hilly terrain, water movement connects soils with one another and differentiates their properties (Blume, 1968); and that adjacent soils at different elevations are linked by a lateral migration of chemical elements to form a single geochemical landscape (Glazovskaya, 1968). A key point in this work is that solution and water transport act selectively so that, as vertical movement in soil profiles produces A and B horizons, the lateral concatenation of soils leads to a differentiation of soil materials along a slope: hill‐top soils are analogues of A horizons, and the valley‐bottom soils are analogues of B horizons (Blume & Schlichting, 1965; Sommer & Schlichting, 1997). Subsequently, the burgeoning sophistication of hillslope hydrological investigations has prompted increasingly detailed and revealing examinations of slope soils using statistical models (e.g. Brown et al., 2004; Brillante et al., 2017) and deterministic models (e.g. Heimsath et al., 1997; Minasny & McBratney, 1999; Yoo et al., 2007; Wackett et al., 2018), although researchers now tend to extend their analysis to the third dimension and consider soil landscapes.

      1.4.3. Soil Landscapes

      Soil catenas are two‐dimensional transects along hillslopes. They form part of a geomorphic system, the flows of material and energy within which are characteristically three‐dimensional. In moving down slopes, weathering products tend to move at right angles to land‐surface contours. Flowlines of material converge and diverge according to the curvature of contours. The pattern of vergency influences the amounts of water, solutes, colloids, and clastic sediments held in store at different landscape positions. Of course, the movement of weathering products alters the topography, which in turn influences the movement of the weathering products; there is feedback between the two systems.

      As soil evolution takes place within a three‐dimensional mantle of material, the spatial pattern of many soil properties will reflect the three‐dimensional topography of the land surface. Indeed, according to the concept of soil‐landscape systems, the dispersion of all the debris of weathering, solids, colloids, and solutes, is, in a general and basic way, influenced hugely by land surface (and phreatic surface) form and organized in three dimensions within the framework imposed by the drainage network (Huggett, 1975).

      Investigating the effect of landscape setting on pedogenesis requires a characterization of topography in three dimensions. Early attempts to describe the three‐dimensional character of topography was made by Andrew R. Aandahl (1948) and Frederick Troeh (1964). Later, geographers and geomorphologists explored methods of terrain description (e.g. Moore et al., 1991). Topographic attributes that appear to be important are those that apply to a two‐dimensional catena (elevation, slope, gradient, slope curvature, and slope length), plus those pertaining to three‐dimensional landform (slope direction, contour curvature, and specific catchment area).

      In distilling previous work on digital soil models, Alex McBratney and his colleagues (2003) proposed that soil is a function of seven factors (the so‐called SCORPAN factors), as follows:

      where S is soil (Sc for soil classes or Sa for soil attributes or properties) at a point; s is existing soil information; c represents climate; o represents organisms or biological activity; r represents topographic or landscape attributes; p represents parent material; a represents age; n represents spatial position. Nathan Odgers et al. (2008) emphasized the strong influence of topography on soil characteristics and soil formation processes, noting that such topographic variables (terrain parameters, topographic attributes) as slope, landscape curvature, and flow direction are readily derived from a digital elevation model. They demonstrated how toposequences can be generated from such models.

      Three‐dimensional topographic influences on soil properties were considered in small drainage basins by the present author (Huggett, 1973, 1975) and Willem J. Vreeken (1973), while André G. Roy and his colleagues (1980) considered soil–slope relationships within a drainage basin. Later work has confirmed that a three‐dimensional topographic influence does exist, and that some soil properties are very sensitive to minor variations in the topographic field (e.g. Moore et al., 1991, 1993; Fissore et al., 2017; Li et al., 2018; Iticha & Takele, 2018).

      1.4.4. Soil‐Landscape Modeling

      Soil‐landscape models seek to integrate soils, parent material, topography, land use and land cover, and human activities with the aim of understanding the spatial distribution of soil attributes, characteristics of soils, and their behavior through time (Grunwald, 2006, 6). They build on earlier soil models, starting with the factorial models instigated by Dokuchaev, Zacharov, and Jenny, and the later soil system models (Table 1.1).

      Soil landscapes are complex, involving geomorphological, biological, and hydrological processes acting over hundreds, thousands, or even millions of years. For this reason, understanding how soil landscapes function and evolve demands an interdisciplinary holistic approach and has benefitted hugely from the appearance of new and powerful technologies over the last few decades: satellite remote sensing, geographic information systems (GISs), global positioning systems (GPSs), digital elevation models (DEMs), and landscape evolution models (LEMs; cf. Brown, 2006). Pedologists have combined these technologies to tackle questions about soil landscapes. The value of digital terrain modeling to hydrology, geomorphology, and ecology was recognized by the early 1990s (e.g. Moore et al., 1991). More recently, pedometrics has attempted to integrate knowledge from numerous disciplines, including soil science, statistics, and GIS.

      Early numerical models of landscapes in geomorphology included soil depth as a state variable (e.g. Ahnert, 1967; Armstrong, 1980). Huggett (1975) simulated the movement of solutes on slopes within a drainage basin. The model of the soil–landscape continuum built by Kevin McSweeney and his colleagues (1994) benefitted from technical advances made through linking hydrology and land‐surface form for terrain‐based modeling of hydrological processes. In their model of soil‐landscape systems, various sources of spatial data (e.g. vegetation and geological substrate) and attribute data (e.g. soil organic‐matter content and particle‐size distribution) are integrated through GIS technology. Four interrelated and iterative stages are applied in the model: physiographic domain characterization; geomorphometric characterization; soil horizon characterization; and soil property characterization. Many subsequent models have appeared, each adding refinements (e.g. Heimsath et al., 1997; Sommer, 2006; Vanwalleghem et al., 2013; Willgoose, 2018).

1.5. SOIL AS AN INTERDEPENDENT SYSTEM

      Ideas about the interdependence of environmental factors were mooted in the late eighteenth century, principally by Johann Reinhold Forster and Alexander von Humboldt. Dokuchaev (1899) stressed the importance of interactions between the zones of the Earth, arguing a that a special discipline was needed to study the integrity of the natural sciences (geology, orography, climatology, botany, zoology) and offering genetic soil science (pedology) as a candidate for such a