During the last half century, advances in geomorphology—abetted by conceptual and technical developments in geophysics, geochemistry, remote sensing, geodesy, computing and ecology—have enhanced the potential value of fluvial history for reconstructing erosional and depositional sequences on the Earth and on Mars and for evaluating climatic and tectonic changes, the impact of fluvial processes on human settlement and health, and the problems faced in managing unstable fluvial systems.
Almost half a century ago, Luna Leopold, Reds Wolman and John Milller published their ‘Fluvial processes in geomorphology’ . It championed a trenchantly quantitative approach to the subject, with numerous dense plots showing the relationship between channel attributes and catchment area, river discharge, flow velocity and other variables, and it demonstrated that regularities lurk beneath messy landscapes. However, as the authors themselves admitted, although concerned with ‘landform development under processes associated with running water’ their emphasis was on process more than development. To be sure, there are chapters on geochronology, drainage pattern evolution, changes in channel form and the evolution of hillslopes, but the data were lacking for ‘a truly satisfactory translation … to historical interpretation’.
Recent decades have witnessed major developments in the techniques required for the translation, the most obvious being the numerical dating of deposits and topographies and the monitoring of surface change by space geodesy. However, advances in neighbouring fields have been as influential, if indirectly. With the rise of plate tectonics in the 1960s, for example, and the geodynamic reassessments it prompted, river history acquired new applications and sometimes unexpected explanations. Planetary exploration yielded striking analogues of terrestrial fluvial morphologies and also underlined the need for a global perspective in their interpretation. Alarm over current climate trends called for meticulous assessment of palaeohydrology as an environmental indicator and its possible links to solar and orbital fluctuations, as well as to man-made atmospheric changes. In addition, ballooning populations increasingly required the informed evaluation of all sources of fresh water and the role of surface and groundwater in disease transmission.
The time therefore seemed ripe for a reappraisal of river history in its own right and also as a constituent of collaborative enquiries. This is not to overlook the many fine reviews of individual rivers or of some of the themes mentioned above that have been published in recent years [2–5], or the wealth of data collected by international projects devoted to fluvial systems [6,7], but merely to focus principally on method.
Besides developments within geomorphology itself, three major sources of ideas and techniques bearing on fluvial studies can be identified: isotope-based environmental analysis and dating; plate tectonics and planetary exploration; and the integration of biological and physical brands of geoscience. The papers making up the present issue illustrate the impact of these advances on river history at different time scales and its bearing on such matters as palaeoclimatology, soil erosion, tectonics, epidemiology and land management.
2. Isotopic histories and chronologies
The foundations for the modern study of fluvial geomorphology were laid between 1890 and 1965 . In 1969, the drainage basin was declared to be the fundamental geomorphological unit ; but without the benefit of any such declaration, it was within the basin that Davis  had formulated the cycle of erosion in the 1890s, and R. E. Horton, A. E. Strahler and US Geological Survey hydrologists patiently revealed in the 1940s and 1950s, by introducing and implementing morphometric analysis , the superiority of quantification over intuition in geomorphology.
Isotopic geochemistry accelerated the process in two key ways: by supplying the wherewithal for detecting and quantifying morphological change, and by erecting a framework that risks undoing some of the chronological gains by its seductiveness; correlation with the marine isotopic sequence sometimes pre-empts age determination, just as it did with the old fourfold Alpine glacial succession.
In 1954, Wheeler  had mocked the emphasis on chronologies that prevailed in Old World archaeology: ‘we have … been preparing time-tables; let us now have some trains’. In the 1960s, there were plenty of geological trains, but their timekeeping remained weak. Ages expressed in years had been obtained by varve (rhythmic lake deposit) and tree-ring counting for a century, but evidently only at favoured locations and sites. The first radiometric ages were measured in 1907 before isotopes had been recognized; many other methods spanning different time ranges soon followed. For river history, the most influential innovation was the radiocarbon (14C) method (developed in 1949), even though it depends on the presence of organic material within the alluvium and its range is a mere 60 000 years or so. From the outset, 14C dating was something of a luxury. Leopold et al.  cited only a handful of radiocarbon ages bearing on two deposits in the American southwest.
The introduction of accelerator mass spectrometric (AMS) 14C analysis in the late 1970s led to greater precision and improved calibration to calendar ages, and small sample size (1/1000–1/100 000 of what is required by radiometric methods) has greatly increased the scope for quantitative time correlation and the evaluation of rates of change. An alluvial deposit that could formerly be given a limiting age if it included an undisturbed hearth can now be traced in its entirety by the analysis of dispersed specks of charcoal. However, progress remains inhibited by high costs and long delays for investigators lacking direct access to an AMS 14C laboratory or enjoying lavish funding.
The mass spectrometer has also greatly improved the precision (and again reduced the required sample size) of U-series dating, notably of calcite, and accelerator mass spectrometry is being applied to many other radionuclides with half-lives too long for decay counting that are relevant to geomorphological analysis (cf. ). In certain settings, the results can be complemented by short-term dating methods, such as those based on lead-210 (half-life of 22.3 years and applicable to the last 150 years) and caesium-137 (a radionuclide with a half-life of 30 years, introduced into the atmosphere by nuclear tests and accidents, and applicable to the last 60 years), which employ gamma spectrometers and other analytical systems that have also experienced dramatic advances in recent years . Mass spectrometric assay of the cosmogenic isotopes beryllium-10 and aluminium-26 permits the dating of land surfaces up to 11 Myr old. Radiation exposure-dating methods—by electron spin resonance, thermoluminescence (TL), optically stimulated luminescence (OSL) and fission-track (FT)—which were developed in the mid 1960s, have expanded the range of datable materials: for example, OSL dating can be applied directly to fluvial sediments laid down within the last 1 Myr , and FT dating to interstratified volcanic ash. The sherds and flints that have long served to date alluvial fills gratis are now themselves endowed with numerical ages obtained on the artefacts themselves (as with TL dating of burnt flint) or in stratified excavations.
Progress in the direct monitoring of landscape change has been equally conspicuous, although the period covered is necessarily limited to the last few decades. Repetitive remote sensing, especially by light detection and ranging (LIDAR) (figure 1), global positioning systems and interferometric synthetic aperture radars, makes it possible to detect morphological changes as small as a few millimetres per year. The devices and techniques for tracing channel evolution that were devised for the Vigil Network —a name now appropriated by an anti-terrorist organization—have proved their worth over the period under review at 82 sites in the USA, Botswana, Sweden, Puerto Rico and Israel, and need to remain modest if they are to be adopted in the many parts of the world that lack information on erosion and channel changes. There are now procedures for monitoring bedload using radioactive labelling, but pebble counts retain their value by virtue of their simplicity.
Similarly, the retrieval and analysis of archival records detailing channel changes in historical times have gained from computer-aided sorting and image processing, and the physical and numerical modelling of channel flow is both routine and innovative. However, there is still scope for investigations and interpretations, which depend on human shrewdness more than technical subtlety. In addition, none of the new techniques would much advance fluvial history without the organization provided by a robust and coherent sequence of erosional and depositional episodes .
This applies equally well to the second major strand in the isotopic upheaval that has struck river history in the last five decades: the marine isotopic time scale. Grove  has reviewed the events that led to the revolution in palaeoclimatology in about 1970 with the recognition from isotopic analysis of deep-sea cores that there had been a multiplicity of glacial advances and retreats and that they had resulted from orbitally controlled fluctuations in insolation. The revolution impinged on fluvial chronologies through sea-level control of delta growth but more generally by providing stratigraphers with a climatic time scale in which to slot major episodes of erosion and deposition.
The marine record accordingly served as a template for correlation and comparison of the numerous fluvial sequences documented by the International Geological Correlation Programme project 449, which ran for the period 2000–2004, and its follow-up project 518 in 2005, and which had as their primary aim the compilation of fluvial sedimentary records across the world. The organizers recognized the value of the sequences for extending the oceanic record into continental interiors. However, they took care to supplement it with data from sediment composition, biostratigraphy and geochronology . There would seem to be a good case for independent dating of river histories , and not just at millennial scales, before any comparison is made with the isotopic time scale, as oceanic composition integrates climatic events at many locations and consequently obscures local anomalies and gradients, which could prove informative.
Improved dating also reduces the area of doubt over the contribution of solar oscillations to the climatic changes of the Pleistocene. At present, the evidence is considered overwhelmingly to favour the Milankovitch model of orbital periodicities as the driver of glacial–interglacial cycles. However, climatic correlation between land and deep sea has long been known to break down at millennial and shorter time scales because oceanic circulation and the distribution of water masses can change, and mixing times are relatively long . Moreover, there is evidence for solar influence on stream behaviour at millennial (and centennial) scale for at least the last 5000 years in middle and low latitudes in the Old World and the Americas . The much-debated question of the solar factor in climate change becomes more tractable when approached through the tangible data of river behaviour.
3. The mobile planet
The revelations and revolutions in geoscience inspired by plate tectonics in the 1960s came, as Bishop  notes, just when computing power was becoming sufficient to address some ‘hitherto unanswerable questions’. Initially, the questions and answers were largely confined to a kinematic description of plate displacement; further progress in computing in due course drove three-dimensional dynamic models, which linked surface motion to mantle processes .
The implications for geomorphology of crustal mobility at the scale of lithospheric plates were most readily grasped by geophysicists, witness the plasticene models employed to reconnoitre the collision between India and Asia, and the numerous analyses of river longitudinal profiles designed to identify crustal movements. Broad-brush landscape analysis of this kind had long been abandoned by most geomorphologists. But the novel view of Earth history impinged on fluvial studies more generally by demolishing the distinction between stable and unstable terrains and by providing a tectonic framework within which the major drainage systems could be set and their gross evolution approximated [24,25].
In zones of plate convergence, the chronology provided by magnetic dating of seafloor spreading and fault azimuths  offered some measure of the resulting uplift and tilting onshore; in strike-slip settings, the disruption of drainage systems became an integral part of their development rather than a complicating factor. In addition, the growing quality and availability of remotely sensed and digital topographic data  gave the greatest benefit to the greatest number.
The crucial half century also witnessed the early years of the space age: Sputnik was launched in 1957. The resulting vision and technological wizardry have brought tangible benefits to terrestrial geosciences. Consider, for example, space imagery and radar altimetry, geodesy from space, and improved understanding of space weather and its terrestrial implications such as the impact of cosmic rays on climate and their isotopic products. The need for a global perspective in implementing these techniques and procedures has come to colour all aspects of fluvial history, for example, the study of a minor tributary may need to consider the former seasonality of the local rainfall and consequently its place in the weather patterns that characterized the latitudes in question at some point in a meteorological sequence partly ordained by orbital or solar periodicities.
Besides missions to planet Earth (such as Aqua, launched in 2002), the National Aeronautics and Space Administration (NASA), the European Space Agency and other space agencies have enriched fluvial (and of course many other) terrestrial studies by their exploration of other planets and their moons and the comparative studies they have stimulated . In the words of NASA's Newell , space science has proved to be integrative in two ways: by drawing on many scientific disciplines, and by informing the study of Earth and its processes.
The channels on Mars have invited comparison with terrestrial networks because Schiaparelli's canali were mistranslated as canals, and there is growing acceptance that they were fashioned at least partly by running water. Carr  has assessed the evidence and compared its chronology with that of the Earth. The discussion can be broadened to consider whether river-like features on other terrestrial planets and their moons were created by water or another fluid and what the answer tells us about past millennia (figure 1). In the words of Huygens [31, p. 24], written in 1698, ‘Since ‘tis certain that Earth and Jupiter have their Water and Clouds, there is no reason why the other Planets should be without them. I can't say that they are exactly of the same nature with our Water; but that they should be liquid their use requires, as their beauty does that they be clear’.
As Pelletier [32, p. 164] notes, ‘alluvial rivers, supraglacial meltwater streams, Gulf Stream meanders, and lava channels on Earth, the Moon, and Venus … exhibit meandering with similar proportionality between meander wavelength and channel width … [which] suggests a common mechanism for meandering’ (figure 2), although, as bank vegetation has been found to play a key role in terrestrial meander formation , it remains to be seen whether a binding agent is required for meander development by fluids other than water.
Besides improvements in the detection and chronicling of fluvial change (figure 3), the latter twentieth century saw a blossoming of research into the interaction between biological, geological and atmospheric processes. An important catalyst for this fusion was the Gaia hypothesis—first formulated by Lovelock & Griffin —which captured the imagination of professionals and laymen alike by its emphasis on life as a critical component of the terrestrial complex. The integration of the three areas of science has long occupied some of their more audacious exponents: consider, for example, George Perkins Marsh and Alexander von Humboldt, any geochemist worth his salt, and all those who have investigated a non-miraculous origin of life. The novelty resides in the notion of their intersection as a scientific field of its own: Earth-system science.
Rivers enter explicitly in some of the Gaian literature  and in the Earth-system analysis ; yet, their study has gained more from selective collaboration between biological and physical fields that remain staunchly reductionist than from holistic yearnings. An evergreen topic is soil erosion by water because both of its practical import (according to the Food and Agriculture Organization (FAO) , by 2005, it had affected 11 million km2) and its elusive interpretation. Long marred by entrenched positions, especially over the relative importance of human and climatic factors, the debate promises to mature now that a refined armoury is being brought to bear on the issues of chronology and process. For instance, per cent loss of caesium-137 has been found to correlate closely with soil loss , and, using 137Cs, Walling & Quine  were able to obtain information on soil loss and deposition in the UK over a period of 30 years within individual fields on contrasting soil type. Objective assessment of the potential impact of plantation agriculture on soil stability was at the core of a study by Trimble  on the southern US Piedmont during 1700–1970, which took into consideration such matters as tenancy, slavery and land abandonment, as well as soil conservation techniques. That work illustrated the need to combine theory with empiricism in the analysis of erosion in historical times . The suggestion that pre-industrial deforestation and other burning of fossil fuel began to modify atmospheric composition significantly through CO2 emission 7000 years ago, as did pre-industrial farming through methane emission 2000 years later , is a useful corrective to any assumption about primitive ecological innocence.
Any such enquiry feeds into the history of human economy and other strands of human evolution that heed the landscape. Schumm  compared the fluviatile setting of the Sumerian, Egyptian and Harappan early civilizations. The stability and continuity of Egyptian civilization, he suggests, echo its river. Sumerian civilization was speculative, pessimistic in its world view and insecure, reflecting the constant flux and instability of the multiple (anastomosing) channels of the Euphrates. Abrupt changes in the course of the Indus by tens or hundreds of kilometres could deprive a major site of its water or subject it to disastrous flooding; the conservatism of the Harappans may have resulted from the struggle against a powerful and dynamic river prone to such unpredictable and far-reaching avulsions.
Perhaps to legitimize his foray into fluvial determinism, Schumm  cited, among others, the ruminations of the historian Macaulay on the influence of the Garonne and the Seine on the personality of Gascons and Normans. Schumm's source was Schama  cogitating on the role of nature in western civilization. There are early antecedents: in the seventh century BC, for example, Guan Zhong compared the greedy, uncouth and warlike people of Qi, with its swift and twisting waters, with the light-hearted and self-confident people of Chu, endowed as it is with gentle and pure streams .
There are more material aspects of historical analysis that can build on river history. Capture of the headwaters of the Hakra by headwaters of the Indus led to the abandonment of numerous sites along the old channels [47,48]. In Mesopotamia, palaeochannel reconstruction at first had perforce to be based on site excavation and textual sources , but it now seeks to integrate textual and archaeological information with geological, geomorphological and chronometric data, and in so doing reveals the human contribution to channel shifts . Indeed, a subset of geoarchaeology devoted to alluvial settings has emerged that identifies what is distinctive (and what is not) about human occupation, however short-lived, of riverine sites, and investigates the interaction between the climatic signal and the cultural signal in alluvial stratigraphies . Many issues can profit from a fluvial focus: resource exploitation, for example, including irrigation farming strategies, or flood history, or, for earlier times, the reconstruction of hominin landscapes [51,52].
The results, together with a wide range of monitoring and modelling procedures , help to inform those engaged in devising river management programmes . Sediment control is perhaps the most urgent issue, as it is linked to accelerated soil erosion, reservoir sedimentation and the wider impact of sediment on aquatic ecology, river morphology and water resource exploitation .
The World Health Organization estimates that four-fifths of illnesses in developing countries are caused by water-borne diseases. Dams are often, but not invariably, to blame. The Volta River Hydro Development project has seen a marked increase in schistosomiasis ; on the other hand, the change in flow regime produced by the Owen Falls dam on the Victoria Nile in Uganda unexpectedly led to the disappearance of the blackfly (Simulium damnosum) responsible for river blindness and previously affecting 65 per cent of the local population . As Whitcombe  shows, the links between fluvial hydrology and the incidence of disease are often complex, but they can be sorted out by giving due attention to the various interlinked chronologies.
I thank Martin Williams for his shrewd comments on a draft of this paper.
One contribution of 10 to a Theme Issue ‘River history’.
- This journal is © 2012 The Royal Society