Hydrological Processes

Published by Wiley
Online ISSN: 1099-1085
Print ISSN: 0885-6087
Publications
The nuclear weapons testing programme of the USA has released radionuclides to the subsurface at the Nevada Test Site. One of these tests has been used to study the hydrological transport of radionuclides for over 25 years in groundwater and the deep unsaturated zone. Ten years after the weapon's test, a 16 year groundwater pumping experiment was initiated to study the mobility of radionuclides from that test in an alluvial aquifer. The continuously pumped groundwater was released into an unlined ditch where some of the water infiltrated into the 200 m deep vadose zone. The pumped groundwater had well-characterized tritium activities that were utilized to trace water migration in the shallow and deep vadose zones. Within the near-surface vadose zone, tritium levels in the soil water are modelled by a simple one-dimensional, analytical wetting front model. In the case of the near-surface soils at the Cambric Ditch experimental site, water flow and salt accumulation appear to be dominated by rooted vegetation, a mechanism not included within the wetting front model. Simulation results from a two-dimensional vadose groundwater flow model illustrate the dominance of vertical flow in the vadose zone and the recharge of the aquifer with the pumped groundwater. The long-time series of hydrological data provides opportunities to understand contaminant transport processes better in the vadose zone with an appropriate level of modelling.
 
Ground-based frequency-modulated continuous wave (FMCW) radar measurements made throughout the 2002–03 NASA Cold Lands Processes (CLPX) mission in Colorado, USA, were designed to study the major electromagnetic transitions within the alpine snowpack over a wide range of conditions, and their effect on measurements made with different active radar measurement parameters. Measurements during 2002 determined that measurements at C-, X-, and Ku-band frequencies were necessary to retrieve the most snowpack information in a wide range of conditions. Measurements at four different incidence angles indicated that the snowpack layering was still visible at 15° but that the rough surface scattering of the snow–ground interface dominated the signal above 30° incidence. Measurements during 2003 were focused on characterizing the microwave response at C-, X-, and Ku-band frequencies at four different sites with different snowpack conditions, indicating that the optimal measurement parameters vary with snowpack conditions. Measurements at different bandwidths illustrate the effect of bandwidth on vertical resolution. This ground-truth data will help interpretation of air- and space-borne active and passive microwave radar measurements that were made coincident with this study. In addition, this work may help guide future researchers when choosing FMCW radar measurement parameters, depending on the type of snowpack information their study requires. Copyright © 2004 John Wiley & Sons, Ltd.
 
Between mid January and early May 2004, during the Canadian Arctic Shelf Exchange Study 2003–04 expedition, we conducted a field project in which we deployed blowing-snow particle counters, visibility sensors, a FlowCapt device, snow bags, an electric field meter and other instruments related to the observation of drifting and blowing snow on first-year ice in Franklin Bay, Northwest Territories, Canada. Average snow depth was 7 cm at the beginning of the campaign and increased to 17 cm towards the end. Snow density in the first 8 cm of snow cover varied from 80 to 450 g m−3, with a median value of 240 g m−3. Basic results will be presented that relate particle counter profiles, wind speed and other data to visibility. We discuss threshold conditions for blowing snow, heat and water vapour fluxes and the electric field. Relative humidity with respect to ice was generally greater than 95% and sublimation rates are believed to have been low. Copyright © 2006 John Wiley & Sons, Ltd.
 
Comparison of mean speci®c catchment area [ln(A s )], mean slope [ln(S)], and mean wetness index [ln(A ds /S)] values computed from 100-m and 1000-m digital elevation models (DEMs) for 50 locations in the conterminous USA (orig100-m is the original 100-m DEM; 1000-m is the 1000-m DEM resampled from the orig100-m DEM; and new100-m is the 100-m DEM interpolated from the 1000- m DEM). The solid line indicates a one-to-one relationship. Values for the 1000-m DEMs are indicated by plus (‡) signs and values for the new100-m DEMs are indicated by circles (s)  
Topographic characteristics computed from 100- and 1000-m resolution digital elevation model (DEM) data are compared for 50 locations representing varied terrain in the conterminous USA. The topographic characteristics are three parameters used extensively in hydrological research and modelling—slope (S), specific catchment area (As) and a wetness index computed as the logarithm of the specific catchment area divided by slope [ln(As/S)]. Slope values computed from 1000-m DEMs are smaller than those computed from 100-m DEMs; specific catchment area and the wetness index are larger for the 1000-m DEMs compared with the 100-m DEMs. Most of the differences between the 100- and 1000-m resolution DEMs can be attributed to terrain-discretization effects in the computation of the topographic characteristics and are not the result of smoothing or loss of terrain detail in the coarse data. In general, the terrain-discretization effects are greatest on flat terrain with long length-scale features, and the smoothing effects are greatest on steep terrain with short length-scale features. For the most part, the differences in the average values of the topographic characteristics computed from 100- and 1000-m resolution DEMs are predictable; that is, biases in the mean values for the characteristics computed from a 1000-m DEM can be corrected with simple linear equations. Copyright © 2000 John Wiley & Sons, Ltd.
 
An analysis of scaling effects is performed to evaluate whether data aggregation is a useful regionalization tool or whether it leads to an unacceptable loss of information. One issue concerns the appropriate resolution of digital elevation models (DEMs) used to derive geomorphological input parameters for hydrological models. In particular, the scale problem of watershed division by a channel network and smaller sub-basins is addressed. The investigation involved commercially available data sets with different horizontal and vertical resolutions and systematically aggregated DEMs. A stream network and the contributing subareas were derived from a DEM with a distinct critical support area. By varying this threshold area various watershed configurations were obtained. The sensitivity of surface runoff simulations to all watershed configurations was studied with synthetic storms and by means of an infiltration excess runoff model.
 
Interception loss, I, was determined by continuous concurrent measurements of the canopy precipitation balances of a mature seed orchard tree of Pinus radiata, and a dominant tree of Eucalyptus viminalis at a mountainous high rainfall site (900 m a.s.l.) in Tallaganda State Forest of the Upper Shoalhaven Catchment. Approximate canopy storage capacity (Sc) of the pine was 54 l, and that of the eucalypt was 11·3 l. Gross pine I was 26·5 per cent and eucalypt I was 8·3 per cent of total incident rainfall over a period of 18 months, from June 1975 to December 1976. The exponential model that provided the best fit to overall data relating I to gross rainfall (Pg) was of good precision for the pine (r2 = 0·73) but rather poor precision for the eucalypt (r2 = 0·27). A consistent pattern in interception data of the two canopy types suggested that variation in I was related to change in pervasive conditions influencing rates of evaporation from wet canopies during rainfall. Multiple regression analyses confirmed that factors such as rainfall intensity and windspeed explained some of the variation in eucalypt I but little in pine I. Negative eucalypt I and corresponding low values of pine I over a wide range of Pg (up to 20 mm) suggest that capture of wind-borne precipitation (cloud, mist, or fog) had also complicated the canopy precipitation balances.
 
The distribution of soil 137Cs in relation to selected soil and landform properties was studied across a 16 ha hillslope hollow in the Hunter valley, New South Wales, Australia. The hillslope was used as grazing for cattle. Caesium-137 was not significantly related to the amount of sand, silt, or clay, the bulk density, the organic matter content, the slope angle or the relative distance downslope. However, 137Cs was significantly related to the thickness of the soil A horizon. Spatial variations in 137Cs were compared with topographic units and a six-element hillslope model, but there was little correspondence. It was thought that the effects of microtopography could have masked potential interrelationships between 137Cs and broader scale landform parameters.
 
The objective of this study was to examine a new resampling methodology for estimating reference levels of 137Cs in uneroded locations. Accurate and precise measurement of 137Cs is required from reference locations to estimate long-term (c. 40 years) sediment redistribution (SRD) and landscape stability. Without reliable long-term, quantitative erosion data it is extremely difficult for land managers to make optimal decisions to ensure landscape sustainability. To determine the influence of 137Cs reference site sampling, particularly under-sampling, on SRD and landscape stability, two statistical approaches were applied to a grid-based data set. Caesium-137 inventories in the reference location (n=36) indicated that data were normally distributed, with a mean inventory of 2150±130 Bq m−2 (±95% confidence band) and a coefficient of variation of 18%. The two approaches used to determine the effect of under sampling included: (1) one-time random subsampling from the total sample collected, subsamples ranged from n=3 to n=30; from these data means and parametric confidence bands were calculated; and (2) random subsamples (n=3 to n=36) were selected from the total 137Cs reference sample, and each subsample was in turn resampled 1000 times with replacement to establish a sampling distribution of means. Thus, an empirically derived mean and 95% confidence bands were established. Caesium-137 activities determined from each approach were input into equations to estimate SRD from two cultivated fields. Results indicate that the one-time random sampling approach for subsamples of size ≤12 significantly over- or under-estimated net SRD, particularly from the gently sloping agricultural field. Computer-intensive resampling produced significantly better estimates of net SRD when compared with the random one-sample approach, especially when a subsample of size three was used. Landscape stability, based on partitioning the agricultural fields into areas exhibiting erosion, stability and deposition, was better approximated for both fields by applying resampling. © 1998 John Wiley & Sons, Ltd.
 
Soil sampling design, the number of samples collected and the lateral variation of caesium-137 (137Cs) in uneroded reference locations were extracted from previously published work. The focus was on published work which used 137Cs reference inventory (Bq m−2) for qualitative or quantitative estimation of sediment redistribution (SRD) within the landscape. The objective of this study was to address one of the methodological concerns facing the 137Cs technique—that is, the lack of a rigorous statistical treatment of reference locations. The limited attention paid to the reference location is not justified as ‘true’ estimates of SRD are based on the assumption of an unbiased, independent, random probability sample estimate, commonly the arithmetic mean. Results from the literature survey indicated that only 11% of the reference locations sampled for 137Cs expressly stated that a probability sampling design was used (transect or systematic-aligned grid). The remaining locations were generally sampled using a non-probability based design, more commonly known as haphazard sampling. Of the 75 reference study areas identified only 40 provided enough information to determine the dispersion around the mean, and from this the coefficient of variation (CV) was calculated for all available data. The median CV was 19·3%, with 95% confidence limits of 13·0–23.4%, indicating that approximately 11 random, independent samples would generally be necessary to adequately quantify the reference 137Cs area activity with an allowable error of 10% at 90% confidence. Further analysis indicated that only one-third of the studies sampled a sufficient number of 137Cs reference locations. This value would actually be lower as sampling frameworks were based on non-probability sampling procedures. For 137Cs reference locations it is recommended that a probability sampling design be utilized, preferably the systematic-aligned grid method, and as a minimum first-order estimate about 11 samples should be collected for inventory estimates.
 
The estimation of erosion and sediment delivery rates in tropical mountain watersheds is difficult and most of the methods widely used for estimating soil erosion over large areas have serious limitations. The 137Cs approach has potential for quantifying soil erosion because it can provide retrospective estimates of long-term (since 1963) net sediment redistribution relatively quickly. Despite its great potential, 137Cs has not yet been used in an extensive, reconnaissance level survey of erosion in complex tropical mountain environments. The objective of this study was to examine the applicability of the 137Cs method to estimate erosion on steep tropical agricultural lands (23 to 80% slopes) in the Nizao watershed, a humid, tropical mountain area of the Dominican Republic. In this study we (i) examine the variation of 137Cs in ten reference sites—eight coffee groves and two forested sites—and (ii) estimate erosion from 14 cultivated fields.The soil pool of 137Cs ranged from to 150 to 192 mBq cm−2 on reference sites with minimal erosion. Variability among reference sites was less than expected for such complex mountain terrain. The variability within coffee and forested reference (average CV=28%) sites was similar to the variability found on grassland and forested reference sites in the temperate zone.The estimated annual soil loss from 14 sampled fields ranged from 6 to 61 t ha−1 year−1 with an overall mean of 26 t ha−1 year−1. Overall, the soil erosion estimates found using the 137Cs method were much lower than those often assumed for such steep tropical hillsides. These erosion estimates account for soil loss since 1963 only and it seems likely that soil losses may have been much higher in earlier decades immediately after initial forest clearing earlier in the 20th century. Copyright © 2000 John Wiley & Sons, Ltd.
 
The caesium-137 method of quantifying soil erosion is used to provide field data for validating the capability of the SHETRAN modelling system for predicting long-term (30-year) erosion rates and their spatial variability. Simulations were carried out for two arable farm sites (area 3–5 ha) in central England for which average annual erosion rates of 6·5 and 10·4 t ha−1 year−1 had already been determined using caesium-137 measurements. These rates were compared with a range of simulated values representing the uncertainty in model output derived from uncertainty in the evaluation of model parameters. A successful validation was achieved in that the simulation range contained the measured rate at both sites, whereas the spatial variability was reproduced excellently at one site and partially at the other. The results indicate that, as the caesium-137 technique measures the erosion caused by all the processes acting at a site, it is relevant to hydrologically based models such as SHETRAN only if erosion by wind, agricultural activities and other processes not represented in the model are insignificant. The results also indicate a need to reduce the uncertainty in model parameter evaluation. More generally, the caesium-137 technique is shown to provide field data that improve the severity of the validation procedure (accounting for internal as well as outlet conditions) and that add spatial variability to magnitude as a condition for identifying unrealistic parameter sets when seeking to reduce simulation uncertainty. Copyright © 2004 John Wiley & Sons, Ltd.
 
The sediment delivery processes occurring in a small Sicilian basin are modelled using the spatially distributed SEDD model recently proposed by Ferro and Minacapilli. The model is applied by using soil data (grain-size distribution, organic matter content, etc.) of 129 samples uniformly distributed over the study area and compiling the available information (topographic map, soil data, etc.) into a Geographical Information System. Finally, the predictive capability of the distributed sediment delivery approach is tested experimentally using the caesium-137 measurement technique. The comparison between calculated sediment yield and the corresponding measured caesium-137 loss is used to validate the SEDD model at the scale of both the single morphological unit and the entire basin. © 1998 John Wiley & Sons, Ltd.
 
Sediment redistribution within near-level agricultural fields in the Gray Wooded soil zone of Saskatchewan was studied using the artificial environmental tracer caesium-137 (137Cs). the objective of this study was to estimate erosion rates caused by wind erosion, and land clearing techniques (i.e. bulldozing). Net rates of erosion and deposition were quantified over the past 30 years on three fields. Wind erosion was estimated to be approximately 1.0 t ha−1 y−1 on a near-level field, with 50 per cent of the sampling sites having erosion rates in excess of soil formation. Bulldozing produced median net sediment flux values of between 40 and 90 t ha−1 y−1. in addition, between 65 and 85 per cent of the sampling sites had erosion rates in excess of the maximum tolerable limit (i.e. 11.0 t ha−1 y−1). These results indicate significant accelerated erosion results from land clearing techniques used in central Saskatchewan. Bulldozing of the fields removed the LFH-horizon and exposed the underlying mineral horizon. Bulk densitites of the 0 to 15 cm layer in the cleared fields were 27 to 55 per cent greater than the comparable depth increment within the undisturbed forest site. Organic carbon concentrations within the 1930s, 1979, and 1987 fields were decreased by 47, 42, and 37 per cent, respectively following cultivation and bulldozing. Decreases in total nitrogen for the cultivated fields ranged from 33 to 38 per cent.
 
The contributions of sediment from different geomorpholigical units within a small basin in the Loess Plateau have been determined using caesium-137 as a tracer. The mean caesium-137 content of sediment originating from the hill area, where sheet and rill erosion are predominant, was 3–37 Bq kg-1, whereas no caesium-137 was detected in the sediment originating from the gully area where gullying and gravitational erosion are predominant. The mean caesium-137 content of sediment from two flood deposits was 0–23 Bq kg-1 and 0–89 Bq kg-1. The relative contribution from the hill area in the two floods was 7 per cent and 26 per cent, whilst that from the gully area was 93 per cent and 74 per cent.
 
Sediment yields from the rolling hills area of the Loess Plateau in northern China (10000–25000 t km−2 yr−1) are amongst the highest in the world. The sediment is believed to derive from both the deep gullies that dissect the rolling plateau and the steep cultivated fields on the slopes of the mounds between the gullies. However, there are few reliable data for erosion rates on the cultivated fields and it is suspected that current estimates (10000–16000 t km−2 yr−1) based on empirical relationships (derived from erosion plot studies) exceed the true values. This study sought to address the need for more information concerning erosion of the cultivated fields through derivation of erosion rates from measurements of rill volume and caesium-137 (137Cs) inventories for typical fields near the village of Ansai, Shaanxi Province. The derived erosion rates are discussed and compared with estimates based on empirical relationships derived from erosion plot data.Where erosion rate estimates based on both rill volume data and 137Cs inventories are available, they show good agreement in the pattern of downslope variation. Both show a sharp decline in erosion rates at a slope length of c. 50 m. This is tentatively attributed to a change from transport-limited to detachment-limited conditions, where rill incision reaches the undisturbed loess at the base of the plough layer. No such decline is visible in the predictions based on empirical relationships derived from erosion plot data. Further evidence is presented that supports the suggestion that these empirical relationships overestimate erosion rates at slope lengths in excess of c. 50 m. It is tentatively suggested that the rates of soil erosion from sloping cultivated fields in the rolling hills area are more likely to lie in the range 8000–10000 t km−2 yr−1 than in the higher range suggested by the empirical relationships. © 1998 John Wiley & Sons, Ltd.
 
Growing awareness of the wider environmental significance of fine sediment transport by rivers and associated sediment problems linked to sediment–water quality interactions, nutrient and contaminant transfer, and the degradation of aquatic habitats has resulted in the need for an improved understanding of the mobilization and transfer of sediment in catchments to support the development of effective sediment management strategies. The sediment budget provides a key integrating concept for assembling information on the internal functioning of a catchment in terms of its sediment dynamics by providing information on the mobilization, transfer, storage and output of sediment. One key feature of a catchment sediment budget is the relationship between the sediment yield at the catchment outlet and rates of sediment mobilization and transfer within the catchment, which is commonly represented by the sediment delivery ratio. To date, most attempts to derive estimates of this ratio have been based on a comparison of the measured sediment yield from a catchment with an estimate of the erosion occurring within the catchment, derived from an erosion prediction procedure, such as the Universal Soil Loss Equation (USLE) or its revised version, RUSLE. There is a need to obtain more direct and spatially distributed evidence of the erosion rates occurring within a catchment and to characterize the links between sediment mobilization, transfer, storage and output more explicitly. In this context, fallout radionuclides have proved particularly useful as sediment tracers. This paper reports the results of a study aimed at exploring the use of caesium-137 (137Cs) measurements to establish sediment budgets for three catchments of different sizes and contrasting land use located in Calabria, southern Italy. Long-term measurements of sediment output were available for the catchments, and, by using the estimates of gross and net rates of soil loss within the catchments provided by 137Cs measurements, it was possible to establish the key components of the sediment budget for each catchment. By documenting the sediment budgets of three catchments of different sizes, the study provides a basis for exploring the effects of scale on catchment sediment budgets and, in particular, the increasing importance of catchment storage as the size of the catchment increases. The results of this study demonstrate a reduction in the sediment delivery ratio from 98 to 2% as catchment area increases from 1·47 ha to 31·2 km2. Copyright © 2010 John Wiley & Sons, Ltd.
 
The 137Cs radioactivity of soils was used as a tracer of soil erosion in a catchment in the Netherlands: 143 samples were analysed to map the 137Cs redistribution using geostatistical interpolation methods. Caesium-137 activities on grassland are significantly higher than on arable land. Also, 137Cs activities on waning slopes are higher and activities on steep slopes are lower. The soil erosion estimates, derived from the 137Cs data, are used to validate the USLE erosion model. The recent Chernobyl nuclear accident also contributed to the 137Cs activity. However, the Chernobyl input of 137Cs, with a constant ratio of 1.765:1 to 134Cs, cannot be used as a tracer of soil erosion. Because of the rapid decay of 134Cs, we will not be possible to separate the sources of 137Cs in the near future in areas significantly influenced by Chernobyl fallout and in these areas 137Cs can no longer be used as a soil erosion tracer.
 
Spatially distributed models of soil erosion and sediment delivery have been increasingly used in studies of catchment sediment dynamics in recent years. Distributed models permit both the spatial heterogeneity of catchment land use, soil properties and topography, and the spatial interaction of soil erosion and sediment transport processes to be represented and can, therefore, provide spatially distributed predictions of soil redistribution rates for complex three-dimensional terrain. However, validation of such distributed models has been severely hampered by the lack of information on the spatial distribution of sediment mobilization and deposition within catchments, and has commonly been restricted to comparison of predicted and measured sediment outputs from catchments. The use of the fallout radionuclide 137Cs for documenting soil redistribution affords a means of assembling distributed information on medium-term rates and patterns of soil redistribution within a catchment which can be used for model testing. The present study uses the 137Cs technique to investigate erosion and sediment delivery processes within a cultivated field in Devon, UK, and the results have been used to test four spatially distributed soil erosion and sediment delivery models, namely the widely used AGNPS and ANSWERS models, a topography-based sediment delivery model, and a topography-driven soil erosion model. Copyright © 2003 John Wiley & Sons, Ltd.
 
To analyse suspended sediment sources in unmanaged Japanese cypress plantation watersheds, field measurements and fingerprinting of the suspended sediment was conducted in the Shimanto River basin in southern Japan. For sediment fingerprinting, 137Cs and 210Pbex were detected by means of gamma-ray spectrometry in the surface soil of the forest floor, stream bank and truck trail and mobilized sediment by interrill erosion. The 137Cs and 210Pbex activities associated with the forest floor materials were considerably higher than those of the stream bank and truck trail. The 137Cs and 210Pbex activities associated with the suspended sediment were found to vary with the sampling period. Evidently, the suspended sediment can comprise materials generated from the forest floor by interrill erosion and those from the truck trail and/or stream bank. The multivariate sediment-mixing model using 137Cs and 210Pbex showed that the contribution of the forest floor varied periodically, ranging from 23–56% in the Hinoki 156 subwatershed and from 18–85% in the Hinoki 155 subwatershed. The difference in the average contribution of the forest floor between Hinoki 156 (46%) and Hinoki 155 (69%) may relate to the presence of truck trail networks in the watershed. The truck trail network can play roles of sediment source and pathway for sediment from forest floor to stream channel due to the concentrated overland flow on the truck trail during heavy rainfall events. These results indicate that the forest floor should be recognized as a major source of suspended sediment in unmanaged Japanese cypress plantation watersheds. Copyright © 2008 John Wiley & Sons, Ltd.
 
We analysed carbon, nitrogen and δ13C values in sediment sampled in the Amazon floodplains. The surface sediment samples were collected along inland transects divided according to their vegetation cover: (a) forest, where the dominant vegetation was inundation forests; (b) grass, where the dominant vegetation was grasses, mainly from Echnochloa polystachya; (c) mixed, where the vegetation changed from grasses near the river bank to forest inland. The average sediment carbon and nitrogen concentrations were significantly higher in forest samples (C = 0·99%, N = 0·12%) than in grass samples (C = 0·84%, N = 0·10%). The average sediment δ13C was significantly heavier in grass (−26·7‰ ) than in forest samples (−28·6‰ ), reflecting the isotopically heavier C4 grasses carbon. In the forest and mixed transects inland, increases in the carbon and nitrogen concentrations were observed. At approximately 50 to 60 m from the riverbank, the concentration of carbon and nitrogen became similar to the river particulate organic carbon (POC) concentration, which is the primary floodplain sediment source. We hypothesized that this trend is caused by the deposition of carbon- and nitrogen-depleted sand-size particles in the first few metres of the floodplain, whereas the more enriched silt–clay particles are preferentially deposited far inland. In contrast to inland trends, no significant downriver trend was observed in the carbon and nitrogen concentrations. However, a downriver δ13C increase was observed for sediment in grass-covered areas. Based on δ13C values in different várzea environments, we estimated the relative contribution of riverine POC and vegetation material present in the várzea sediments under different vegetation covers. For forest sediment samples we estimated that 60% of the carbon was derived from the riverine POC and the forests provided the remaining 40%. For open areas covered with grasses we estimated that only 10% was derived from these plants, and 90% was derived from riverine POC. Copyright © 2003 John Wiley & Sons, Ltd.
 
Sediment cores (collected from nine headwater lakes and reservoirs in their catchments and adjacent areas and from ten floodplain sites in the middle and lower reaches of the two river systems) have been used to reconstruct changes in suspended sediment yields, rates of sedimentation and suspended sediment sources for the Ouse and Tweed catchments in the UK over the last ca 100–150 years in response to recent environmental change. For the small lake/reservoir catchments in or adjacent to the Ouse basin, average sediment yields for the recent period (1963–95) were generally found to exceed those for the period prior to 1963, by up to 94%, suggesting that sediment yields have increased over time. For the equivalent lake/reservoir catchments in and adjacent to the Tweed basin, sediment yields were generally lower in the period since 1963. In both basins, sediment yields over the past 100–150 years were characterized by considerable temporal variability, which was seen to reflect, at least in part, the significant changes in land use and land management that occurred within their catchments. No clear links between the variability of sediment yields and changes in climate were apparent. Changes in sediment source reconstructed for one of the lakes appeared primarily to reflect catchment disturbance during afforestation in the 1940s. Average rates of overbank sedimentation for the ten floodplain sites varied considerably between sites. For those cores collected from the Ouse basin there was no consistent temporal trend, and average sedimentation rates for the period extending from ca 1895 to 1963 were broadly similar to those for the period since 1963. For the Tweed basin, sedimentation rates for all three cores were higher prior to 1963, reflecting the increased soil erosion and sediment yields associated with the conversion of pasture to arable land in the immediate post-World War II period and the ditching and planting operations associated with afforestation in the 1940s. Substantial downcore variability in the source of the overbank sediment, considered in terms of both source type and spatial location, was found in both river basins. This was seen to reflect land-use change, including land drainage, the conversion of pasture to arable land, afforestation, and increased cereal production. The evidence provided by the floodplain cores again showed little indication of any influence linked to climate change. Although the detailed patterns of temporal variation of sediment yield, overbank sedimentation rate and sediment source reconstructed from the sediment cores collected from the lake/reservoirs and floodplain locations are site specific, there is a general consistency between the changes in sediment dynamics reconstructed for the headwater catchments (based on the cores collected from the lakes and reservoirs) and the equivalent evidence for the middle and lowland reaches of the catchments (provided by overbank floodplain deposits). Despite significant changes in climate, and, more particularly, land use over this time period, the general lack of evidence of substantial changes in downstream sediment yield and rates of overbank sedimentation over the last ca 100–150 years in the Ouse and Tweed basins suggests that their sediment dynamics have been generally insensitive to recent environmental change. This finding has important implications when considering the likely response of these, and similar, UK river basins to future environmental change. Copyright
 
Hydrological and biogeochemical relationships were evaluated for the Arbutus Watershed in the Adirondack Mountains of New York State. Wet-only precipitation inputs (weekly) and discharge losses were evaluated from 1983 to 1998. Precipitation from 1983–98 had significant (p<0·05) decreases in concentrations of Ca²⁺(−0·10µmolcl⁻¹year⁻¹), Mg²⁺(−0·07µmolcl⁻¹year⁻¹), K⁺(−0·01µmolcl⁻¹year⁻¹), Na⁺(−0·04µmolcl⁻¹year⁻¹), Cl⁻(−0·07µmolc l⁻¹year⁻¹), SO4²⁻(−0·82µmolcl⁻¹year⁻¹) and H⁺(−0·55µmolcl⁻¹year⁻¹), whereas NH4⁺ and NO3⁻ concentrations did not change.
 
We report a quantitative analysis of regional differences in the the oxygen isotope composition of river water and precipitation across the USA because data are now available to undertake a more geographically and temporally extensive analysis than was formerly possible. Maps of modern, mean annual δ18O values for both precipitation (δ18OPPT) and river water (δ18ORIV) across the 48 contiguous states of the USA have been generated using latitude and elevation as the primary predictors of stable isotope composition while also incorporating regional and local deviations based on available isotopic data. The difference between these two maps was calculated to determine regions where δ18ORIV is significantly offset from local δ18OPPT. Additional maps depicting seasonal and extreme values for δ18ORIV and δ18OPPT were also constructed. This exercise confirms the presence of regions characterized by differences in δ18ORIV and δ18OPPT and specifically identifies the magnitude and regional extent of these offsets. In particular, the Great Plains has δ18ORIV values that are more positive than precipitation, while much of the western USA is characterized by significantly lower δ18ORIV values in comparison with local δ18OPPT. The most salient feature that emerged from this comparison is the ‘catchment effect’ for the rivers. Because river water is largely derived from precipitation that fell upstream of the sample locality (i.e. at higher elevations) δ18ORIV values are often lower than local δ18OPPT values, particularly in catchments with high-elevation gradients. Seasonal patterns in the isotopic data substantiate the generally accepted notion that amplitudes of δ18O variation are greatly dampened in river water relative to those of local precipitation. Copyright
 
Reddy et al. (2006, Hydrological Processes20: 1753–1772) disregard germane published data and successful isotopic models, selectively depict data from their data sets, misplot many of their data, and use poorly correlated sinusoidal fits to their data to derive erroneous hydrologic time constants for features in the Shingobee River Headwaters Area. A damped running average model is more realistic. Copyright © 2006 John Wiley & Sons, Ltd.
 
To clarify the processes of atmospheric water vapour transport to the northern part of the North China Plain (NCP), isotopic variations in precipitation are investigated. Pattern of seasonal variation in the isotopic composition of precipitation over China can be classified into three types: winter depression, summer depression and intermediate. The northern NCP is located at the northern margin of the region where isotopic composition decreases in summertime, associated closely with the Asian summer monsoon. Inter-storm variability of isotopic composition over the northern NCP during the monsoon season has an inverse relationship with atmospheric moisture content, although it cannot be accounted for by the amount effect. Thus, the isotopic composition is inferred to be primarily controlled by large-scale monsoon activity, which supplies water vapour into the northern NCP, rather than by local precipitation processes. From an analysis based on a relationship between the deuterium excess and humidity conditions at possible oceanic source areas, it is revealed that precipitating water in the early monsoon season originates in the South China Sea, and its source area tends to shift to the East China Sea and/or Yellow Sea in the later period of the season. Copyright © 2004 John Wiley & Sons, Ltd.
 
Reconstruction of continental palaeoclimate and palaeohydrology is currently hampered by limited information about isotopic patterns in the modern hydrologic cycle. To remedy this situation and to provide baseline data for other isotope hydrology studies, more than 4800, depth- and width-integrated, stream samples from 391 selected sites within the USGS National Stream Quality Accounting Network (NASQAN) and Hydrologic Benchmark Network (HBN) were analysed for δ¹⁸O and δ²H (http://water.usgs.gov/pubs/ofr/ofr00-160/pdf/ofr00-160.pdf). Each site was sampled bimonthly or quarterly for 2·5 to 3 years between 1984 and 1987. The ability of this dataset to serve as a proxy for the isotopic composition of modern precipitation in the USA is supported by the excellent agreement between the river dataset and the isotopic compositions of adjacent precipitation monitoring sites, the strong spatial coherence of the distributions of δ¹⁸O and δ²H, the good correlations of the isotopic compositions with climatic parameters, and the good agreement between the ‘national’ meteoric water line (MWL) generated from unweighted analyses of samples from the 48 contiguous states of δ²H=8·11δ¹⁸O+8·99 (r²=0·98) and the unweighted global MWL of sites from the Global Network for Isotopes in Precipitation (GNIP) of the International Atomic Energy Agency and the World Meteorological Organization (WMO) of δ²H=8·17δ¹⁸O+10·35.
 
The stable isotopes of oxygen and hydrogen incorporated in the water molecule (18O and 2H) have become an important tool not only in Isotope Hydrology, routinely applied to study the origin and dynamics of surface and groundwaters, but also in studies related to atmospheric circulation and palaeoclimatic investigations. A proper understanding of the behaviour of these tracers in the water cycle is required for a meaningful use of these tools in any of these disciplines. Our knowledge of the vertical distribution and the factors controlling the stable isotope ratios of oxygen and hydrogen in atmospheric moisture derives from a limited number of observations and vertical profiles in the atmosphere. An international programme jointly operated by the International Atomic Energy Agency (IAEA) and the World Meteorological Organization (WMO), and operational since 1961, has resulted in the development of a dedicated database to monitor isotope ratios in precipitation in more than 500 meteorological stations world-wide. The main features of the spatial and temporal variations of stable isotope ratios of oxygen and hydrogen in precipitation and atmospheric moisture at the global scale are presented based on the analysis of limited data on water vapour, data obtained by the Global Network for Isotopes in Precipitation (GNIP) and the few observations at high latitudes. Copyright © 2000 John Wiley & Sons, Ltd.
 
Pukemanga is a small (3 ha) steep headwater catchment at the Whatawhata Research Station near Hamilton, New Zealand. The water balance (1996–2002) shows average annual rainfall of 1640 mm producing annual runoff of 440 mm (baseflow 326 mm, stormflow 114 mm) and ‘deep seepage’ loss of 450 mm (i.e. 450 mm of water not appearing in the stream). Oxygen-18 (18O) concentrations were measured at weekly intervals for 8–15 months at six sites, ranging from Pukemanga Stream baseflow through wetland seepage to ephemeral streams and surface runoff. The first two showed no significant 18O variations. Inferred mean residence times within the catchment ranged from at least 4 years (for the stream baseflow and seepage) to a few weeks (for the ephemeral flows and surface runoff). Silica concentrations could also be used to distinguish deep flowpath water from near-surface flowpath water. Tritium concentrations gave an estimated mean residence time of 9 years for Pukemanga Stream baseflow. Sulphur hexafluoride tended to give younger ages, while the chlorofluorocarbon ages were older, but are not considered as reliable for dating streamflow in this time range. These results show that deep pathways predominate with over 74% of runoff deriving from deep hillslope flowpaths via the wetland, and 87% of total drainage (baseflow and deep seepage) travelling via deep hillslope flowpaths. Our conception of the deep drainage process is that there is a large volume of slowly moving water in the system (above and below the water table), which reaches the wetland and stream via an unconfined groundwater system. Subsurface water equivalents are estimated to be 2·9 m for drainage at the weir and 4·1 m for drainage bypassing the weir, giving a total of 7 m depth over the catchment. The unsaturated zone plays an important role in storing water for long periods (about 4 years), while linking the surface with the groundwater water table to contribute to the fast streamflow response to rainfall. A schematic model of the various pathways with indicative residence times is given. Copyright © 2006 John Wiley & Sons, Ltd.
 
A combination of hydrometric data and observations of natural isotope (oxygen-18) variations in saturation overland flow and stream discharge were used to investigate the sources of storm run-off in a headwater swamp located in a permanent groundwater discharge zone near Toronto, Canada. The results of a two-component hydrograph separation indicated that pre-event water formed 80–90% of the stream hydrograph volume for six of the seven storms analyzed in June–November 1990. However, the instantaneous event water contribution showed considerable variability, ranging from maximum values of 20–25% for four moderate intensity storms to 63% for a high intensity thunderstorm with a return period of two years. The relative contribution of event and pre-event water to storm run-off from saturated areas and small streamlets within the swamp was similar to the main outlet stream. The dominance of pre-event water during storms could be accounted for by the mixing of a small volume of event water with a large pool of pre-event water on the surface of permanently saturated areas within the swamp. Occasional storms of high intensity or long duration produced a greater shift towards an event water signature in the saturated areas and a larger event water contribution to the outlet stream hydrograph.
 
The differences between δ18O in throughfall and open rainfall were studied for 16 selected spring and summer storms in deciduous, pine and spruce forests in central Pennsylvania, USA. Throughfall δ18O averaged 0.17, 0.32 and 0.24%o greater than δ18O of open rainfall for all storms at the deciduous, pine and spruce sites, respectively. Throughfall 18O enrichment was greater in low intensity spring rainfall events than higher intensity growing season storms and greater in the coniferous stands than the deciduous hardwood stand. Maximum throughfall 18O enrichment of l.61%o occurred in the spruce forest during one spring event. The differences between rainfall and throughfall 18O observed in this study for individual storm events may have important implications for isotope flow separation studies.
 
Quantifying of direct recharge derived from precipitation is crucial for assessing sustainability of well-irrigated agriculture. In the North China Plain, the land use is dominated by groundwater-irrigated farmland where the direct recharge derived from precipitation and irrigation. To characterize the mean rate and historical variance of direct recharge derived from precipitation, unsaturated zone profiles of chloride and δ18O in the dry river bed of the Beiyishui River were employed. The results show that archival time scale of the profile covers the duration from 1980 to 2002 (corresponding to depths from 5 to 2 m) which is indicated by matching the δ18O peaks in the isotope profile with the aridity indexes gained by instrumental records of annual precipitation and annual potential evaporation. Using the chloride mass balance method, the mean rate of the direct recharge corresponding to the archival time scale is estimated to be 3·8 ± 0·8 mm year−1, which accounts for about 0·7% of the long-term average annual precipitation. Further, the direct recharge rates vary from 2·1 to 6·8 mm year−1 since 1980. Despite the subhumid climate, the estimate of recharge rates is in line with other findings in semiarid regions. The low rate of direct recharge is considered as a result of the relative dry climate in recent decades. In dry river bed, unsaturated zone profiles of chloride and δ18O combined with instrumental records could offer valuable information about the direct recharge derived from precipitation during droughts. Copyright © 2011 John Wiley & Sons, Ltd.
 
There are both practical and theoretical issues to consider when doing recession analyses of the type introduced by Brutsaert and Nieber (1977). Tung et al. (2004) point out a practical issue, which is the error introduced by using time-averaged values of discharge. The issue, however, is probably not serious for many practical applications and the error can be made negligible by using time intervals that are appropriate to the time rate of change in discharge. Copyright © 2007 John Wiley & Sons, Ltd.
 
We have investigated riverflow variability in England and Wales by examining the reconstructed monthly discharge time series from fifteen catchments in these regions for the period 1865–2002. The riverflow fluctuations exhibit a strong annual cycle. The flow in the annual cycle is found to be intermittent, with the degree of intermittency varying from one catchment to another. An intermittent flow is characterized by bursts of high discharge separated by intervals with low or no discharge. By applying a continuous wavelet transform to the time series, we have identified the occurrence of intermittency in the annual cycle. The riverflow activity is also found to exhibit variations at interannual and quasi-decadal time scales. These variations may be linked to large-scale climatic processes such as the North Atlantic Oscillation (NAO). We have used the kurtosis of the probability density functions of the various time series as a measure of the degree of intermittency. An intermittent flow is characterized by a peaked (super-Gaussian) probability density function with kurtosis in excess of 3. A higher value of kurtosis signifies a higher degree of intermittency. Intermittent fluctuations are more difficult to predict accurately than persistent oscillations, i.e., those lasting continuously over a long time interval. Copyright © 2009 John Wiley & Sons, Ltd.
 
The Columbia River is a major source of and conduit for Pacific Northwest economic activity, and is one of the more heavily modified rivers in North America. Understanding human and climate-induced changes in its hydrologic properties is, therefore, vital. Long streamflow records are essential to determining how runoff has changed over time, and Columbia River daily streamflow record at The Dalles began in 1878. To understand and separate anthropogenic and climate effects, however, it is also necessary to have a basin-scale estimate of virgin or naturalized flow. The United States Geological Survey has calculated a monthly averaged adjusted river flow at The Dalles for 1879–1999 that accounts for the effects of flow regulation. The Bonneville Power Administration has estimated the monthly averaged virgin flow at The Dalles, i.e. the flow in the absence of both flow regulation and irrigation depletion for 1929–89. We have estimated the monthly virgin flow of the Columbia River at The Dalles from records of irrigated area for the missing early years, i.e. for the period 1879–1928. In addition, to allow hindcasting of a virgin flow sediment transport for the system, a daily virgin flow index with realistic higher moments and spectral properties has been calculated. Examination of the virgin flow record shows that climate change since the late 19th century has decreased annual average flow volume by > 7%; irrigation depletion has reduced the flow by another ∼7%. Copyright © 2005 John Wiley & Sons, Ltd.
 
[ The d 07 O and deuterium excess "d dD−7×d 07 O# values for the IAEA\ Ottawa\ Canada site and for the National Atmospheric Deposition Program "NADP#\ North Platte\ NE site "0878\ 0889 and 0880# 
A portion of the precipitation samples collected and stored by the National Atmospheric Deposition Program (NADP) are shown to be useful for analysis of isotopes in precipitation. The potential problems with evaporation are small based on deuterium excess analyses and comparisons with the Global Meteroic Water Line. Presented here are the δ18O values of precipitation collected from nine NADP sites during 1989, 1990 and 1991. The trends in the isotopic (δ18O) characteristics of recent precipitation are in agreement with findings from previous International Atomic Energy Agency (IAEA) sites in the USA. The findings are also in agreement with several major isotope–environment relationships, further supporting the use of these samples for a modern global data base on the isotopes in precipitation being developed by IAEA, called GNIP (Global Network for Isotopes in Precipitation) and for use by research groups in the hydrological modelling, palaeoclimate and ecological communities.As expected, the average δ18O values of precipitation that is derived from the Gulf of Mexico (−3‰) and from the Pacific North-west are isotopically distinct (−7‰). In addition, using the NADP network, isotopic depletion in the δ18O values of precipitation in the range of 8‰ was observed from coastal to inland locations either in the Pacific North-west or along the east side of the Rocky Mountains, from Texas to Eastern Montana. In central USA, especially at high elevation, there is a strong seasonal variation in the δ18O values of precipitation, differing by almost 25‰ between January and August, whereas at coastal locations the seasonal variation in the δ18O values of precipitation was minimal. Comparisons between the average δ18O values of precipitation from several NADP sites and those of the IAEA collected in the 1960s and 1970s are exceptionally comparable, differing in some cases by only 0\1‰, further substantiating the quality and utility of precipitation from this network. The δ18O–temperature relationships for two of our sites had slopes of 0\55 (Colorado) and 0\46 (Nebraska) that are within the range of those reported earlier (0\29–0\7), although they are slightly lower, possibly owing to the large temperature variation between winter and summer in central USA. Further, the seasonal variation in the δ18O values of precipitation are used in a pair of ecological studies to articulate water sources of different plant life forms in riparian and prairie ecosystems, with grasses typically using summer rain to a much greater extent than woody species, such as willow or saltbush. Copyright © 2000 John Wiley & Sons, Ltd.
 
18O is an ideal tracer for characterizing hydrological processes because it can be reliably measured in several watershed hydrological compartments. Here, we present multiyear isotopic data, i.e. 18O variations (δ18O), for precipitation inputs, surface water and groundwater in the Shingobee River Headwaters Area (SRHA), a well-instrumented research catchment in north-central Minnesota. SRHA surface waters exhibit δ18O seasonal variations similar to those of groundwaters, and seasonal δ18O variations plotted versus time fit seasonal sine functions. These seasonal δ18O variations were interpreted to estimate surface water and groundwater mean residence times (MRTs) at sampling locations near topographically closed-basin lakes. MRT variations of about 1 to 16 years have been estimated over an area covering about 9 km2 from the basin boundary to the most downgradient well. Estimated MRT error (±0·3 to ±0·7 years) is small for short MRTs and is much larger (±10 years) for a well with an MRT (16 years) near the limit of the method. Groundwater transit time estimates based on Darcy's law, tritium content, and the seasonal δ18O amplitude approach appear to be consistent within the limits of each method. The results from this study suggest that use of the δ18O seasonal variation method to determine MRTs can help assess groundwater recharge areas in small headwaters catchments. Copyright © 2005 John Wiley & Sons, Ltd.
 
The first results of a regional circulation model REMOiso fitted with water isotope diagnostics are compared with various isotope series from central Europe. A 2 year case study is conducted from March 1997 to February 1999 centred over Europe, analysing daily and monthly measurements. Isotope signals over Europe are dominated by the typical isotopic effects such as temperature, continental and altitude effects, both on annual and seasonal scales. These well-known isotopic effects are successfully reproduced by REMOiso, using two different boundary data sets. In a first simulation, the European Centre for Medium-range Weather Forecasts (ECMWF) analyses serve as boundary conditions, where water isotopes were parameterized by a simple temperature dependence. In a second simulation, boundary conditions both for climatic and isotopic variables are taken from the ECHAMiso general circulation model output. The comparison of both simulations shows a very high sensitivity of the simulated δ18O signal to boundary conditions. The ECMWF-nested simulation shows an average offset of −4·5‰ in mean δ18O values and exaggerated seasonal amplitude. The ECHAM-nested simulation represents correctly the observed mean δ18O values, although with a dampened seasonality. REMOiso's isotope module is further validated against daily δ18O measurements at selected stations (Nordeney, Arkona and Hohenpeissenberg) situated in Germany. Copyright © 2005 John Wiley & Sons, Ltd.
 
Rainfall and river indices for both the northern and southern Amazon were used to identify and explore long-term climate variability on the region. From a statistical analysis of the hydrometeorological series, it is concluded that no systematic unidirectional long-term trends towards drier or wetter conditions have been identified since the 1920s. The rainfall and river series showing variability at inter-annual scales linked to El Niño Southern Oscillation was detected in rainfall in the northern Amazon. It has a low-frequency variability with a peak at − 30 years identified in both rainfall and river series in the Amazon. The presence of cycles rather than a trend is characteristic of rainfall in the Amazon. These cycles are real indicators of decadal and multi-decadal variations in hydrology for both sides of the basin. Sea-level pressure (SLP) gradients between tropics and sub topics were explored in order to explain variability in the hydrometeorology of the basin. Sea surface temperature (SST) gradients inside the tropical Atlantic and between the tropical Atlantic and the sub-tropical Atlantic have been assessed in the context of changes in rainfall in the Amazon, as compared to northern Argentina. Trends in SSTs in the subtropical Atlantic are linked to changes in rainfall and circulation in northern Argentina, and they seem to be related to multi-decadal variations of rainfall in the Amazon. Copyright © 2009 John Wiley & Sons, Ltd.
 
Twenty-three snow-course sites in and near Maine, USA, with records spanning at least 50 years through to 2004 were tested for changes over time in snowpack depth, water equivalent, and density in March and April. Of the 23 sites, 18 had a significant decrease (Mann-Kendall test, p < 0·1) in snowpack depth or a significant increase in snowpack density over time. Data from four sites in the mountains of western Maine–northern New Hampshire with mostly complete records from 1926 to 2004 indicate that average snowpack depths have decreased by about 16% and densities have increased by about 11%. Average snowpack depths and water equivalents in western Maine–northern New Hampshire peaked in the 1950s and 1960s, and densities peaked in the most recent decade. Previous studies in western North America also found a water-equivalent peak in the third quarter of the 20th century. Published in 2006 by John Wiley & Sons, Ltd.
 
Since ancient times meltwater from glaciers in mountain zones has been an important water resource in the arid region of northwest China. Meteorological elements for heat balance calculations at the glacier surface were simplified with the aim of simulating discharge fluctuations from the July 1st Glacier over a span of decades. By estimating relative humidity and downward solar radiation from precipitation, heat balance calculations at the glacier can be made using only daily temperature and precipitation, which are easily obtainable. Calculations of daily discharge using the above method during the melting season in 2002 produced a better simulation than those using the conventional positive degree-day (PDD) factor method. (The ablation of glacier during any particular period is assumed to be proportional to the sum of daily mean temperatures above the melting point during that period.)We calculated the glacier runoff and mass balance at the July 1st Glacier since 1935 from monthly meteorological data taken near the glacier using the above method. We used three patterns of daily temperature and precipitation to estimate daily meteorological data from monthly or annual data. Calculated mass balance using the patterns of air temperature and precipitation in years when the pressure distributions were similar to its long-term average corresponded well with observed data. Our simulation of the glacier runoff since 1935 also showed good results. Copyright © 2008 John Wiley & Sons, Ltd.
 
An extensive dataset (230 precipitation gauges and 79 stream gauges) was used to analyse rainfall–runoff relationships in 10 subregions of a 482000 km2 area in the south-eastern USA (Maryland, Virginia, North Carolina, South Carolina and Georgia). The average annual rainfall and runoff for this study area between 1938 and 2005 were 1201 and 439 mm, respectively. Average runoff/rainfall ratios during this period varied between 0·24 in the southernmost Coastal Plain subregion to 0·64 in the Blue Ridge Province. Watershed elevation and relief are the principal determinants governing the conversion of rainfall to runoff. Temporal rainfall variation throughout the south-eastern USA ranges from ∼40% above and below normal while the variation for runoff is higher, from − 75% to + 100%. In any given year there can exist a ± 25–50% error in predicted runoff deviation using the annual rainfall–runoff regression. Fast Fourier Transform and autoregressive spectral analysis revealed dominant cyclicities for rainfall and runoff between 14 and 17 years. Secondary periodicities were typically between 6–7 and 10–12 years. The inferred cyclicity may be related to ENSO and/or Central North Pacific atmospheric phenomena. Mann–Kendall analyses indicate that there were no consistent statistically significant temporal trends with respect to south-eastern US rainfall and runoff during the study period. The results of U-tests similarly indicated that rainfall between 1996 and 2005 was not statistically higher or lower than during earlier in the study period. Copyright © 2008 John Wiley & Sons, Ltd.
 
Before 1900, the Missouri–Mississippi River system transported an estimated 400 million metric tons per year of sediment from the interior of the United States to coastal Louisiana. During the last two decades (1987–2006), this transport has averaged 145 million metric tons per year. The cause for this substantial decrease in sediment has been attributed to the trapping characteristics of dams constructed on the muddy part of the Missouri River during the 1950s. However, reexamination of more than 60 years of water- and sediment-discharge data indicates that the dams alone are not the sole cause. These dams trap about 100–150 million metric tons per year, which represent about half the decrease in sediment discharge near the mouth of the Mississippi. Changes in relations between water discharge and suspended-sediment concentration suggest that the Missouri–Mississippi has been transformed from a transport-limited to a supply-limited system. Thus, other engineering activities such as meander cutoffs, river-training structures, and bank revetments as well as soil erosion controls have trapped sediment, eliminated sediment sources, or protected sediment that was once available for transport episodically throughout the year. Removing major engineering structures such as dams probably would not restore sediment discharges to pre-1900 state, mainly because of the numerous smaller engineering structures and other soil-retention works throughout the Missouri–Mississippi system. Published in 2009 by John Wiley & Sons, Ltd.
 
Top-cited authors
Carol Kendall
  • United States Geological Survey
James W. Kirchner
Des E Walling
  • University of Exeter
G. Blöschl
Jeff Arnold
  • United States Department of Agriculture