Increasing concentrations of organic matter (OM) in surface waters have been noted over large parts of the boreal/nemoral zone in Europe and North America. This has raised questions about the causes and the likelihood of further increases. A number of drivers have been proposed, including temperature, hydrology, as well as SO 2À 4 -and Cl À deposition. The data reported so far, however, have been insufficient to define the relative importance of different drivers in landscapes where they interact. Thirty-five years of monthly measurements of absorbance and chemical oxygen demand (COD), two common proxies for OM, from 28 large Scandinavian catchments provide an unprecedented opportunity to resolve the importance of hypothesized drivers. For 21 of the catchments, there are 18 years of total organic carbon (TOC) measurements as well. Despite the heterogeneity of the catchments with regards to climate, size and land use, there is a high degree of synchronicity in OM across the entire region. Rivers go from widespread trends of decreasing OM to increasing trends and back again three times in the 35-year record. This synchronicity in decadal scale oscillations and long-term trends suggest a common set of dominant OM drivers in these landscapes. Here, we use regression models to test the importance of different potential drivers. We show that flow and SO 2À 4 together can predict most of the interannual variability in OM proxies, up to 88% for absorbance, up to 78% for COD. Two other candidate drivers, air temperature and Cl À , add little explanatory value. Declines in anthropogenic SO 2À 4 since the mid1970s are thus related to the observed OM increases in Scandinavia, but, in contrast to many recent studies, flow emerges as an even more important driver of OM variability. Stabilizing SO 2À 4 levels also mean that hydrology is likely to be the major driver of future variability and trends in OM.

Running water comprises just over one millionth of the world's water. The importance of those streams and rivers as a resource for human welfare and biodiversity, however, is far out of proportion to that minuscule fraction. This explains why protecting running waters (the flow regimes, water quality and biota) is such a vital concern for society. Yet for all the focus and concern, how much do we actually know about these running waters, and the lotic habitat they comprise?Consider what would happen if one asked any national environmental authority to assess the basic chemical and ecological status of running waters. At the river mouths, there would be enough information to make a reasonable assessment of the status. But somewhere on the way upstream, available data would run dry, long before most stream channels did (in non-arid regions).In Sweden, with an ambitious programme for monitoring and assessing surface waters, it came as a surprise several years ago to realize that the length of all perennial streams on the country's maps was not known. When that was modelled in the form of a 'virtual network' from a 50 m × 50 m digital elevation model, the total length turned out to be 530 000 km (ca 1 km/km 2 ), which was double the previous estimates. The length was independently confirmed by another group using remote sensing data (Esseen et al., 2004). Further analysis of the virtual network revealed that over 90% of the stream length had catchment areas under 15 km 2 . Although this might seem merely of academic interest, 15 km 2 is the lower limit for what has been surveyed on a national scale in Sweden. Does this mean that we have missed something important in our assessment of water resources?When the chemistry of all flowing headwaters in a single 78 km 2 catchment was compared to the 2000 Swedish national survey of running waters, there was as much variability within the headwaters of that forested catchment as could be found in a statistically representative sample of over 260 000 km 2 of Sweden's boreal forest waters (Temnerud and Bishop, 2005). Other studies on biota have not just found such headwaters to be teeming with biodiversity, but also found species that are endemic to headwaters (Meyer et al., 2007). Discrete inquiries were made to see if national agencies in other countries of North America and Europe had come further in the documentation and assessment of headwaters. The answer was 'no'. There are, however, some significant efforts (e. g. Hutchins et al., 1999; Smart et al., (2006). But since many first and second order streams are not on the US maps, and the assessment went up to fifth order streams, headwaters are likely to be seriously underrepresented even in that landmark survey. 2001; Likens and Buso, 2006). The most notable is the US Environmental Protection Agency's (EPA's) 'Wadeable Stream AssessmentIn most regions, the overwhelming majority of stream length lies beyond the frontiers of any systematic documentation and would have to be represented as a blank space on the assessment m...

Abstract. Runoff generation processes and pathways vary widely between catchments. Credible simulations of solute and pollutant transport in surface waters are dependent on models which facilitate appropriate, catchment-specific representations of perceptual models of the runoff generation process. Here, we present a flexible, semi-distributed landscape-scale rainfall-runoff modelling toolkit suitable for simulating a broad range of user-specified perceptual models of runoff generation and stream flow occurring in different climatic regions and landscape types. PERSiST (the Precipitation, Evapotranspiration and Runoff Simulator for Solute Transport) is designed for simulating present-day hydrology; projecting possible future effects of climate or land use change on runoff and catchment water storage; and generating hydrologic inputs for the Integrated Catchments (INCA) family of models. PERSiST has limited data requirements and is calibrated using observed time series of precipitation, air temperature and runoff at one or more points in a river network. Here, we apply PERSiST to the river Thames in the UK and describe a Monte Carlo tool for model calibration, sensitivity and uncertainty analysis.

The permeability architecture of the critical zone exerts a major influence on the hydrogeochemistry of the critical zone. Water flow path dynamics drive the spatiotemporal pattern of geochemical evolution and resulting streamflow concentration‐discharge (C‐Q) relation, but these flow paths are complex and difficult to map quantitatively. Here we couple a new integrated flow and particle tracking transport model with a general reversible Transition State Theory style dissolution rate law to explore theoretically how C‐Q relations and concentration in the critical zone respond to decline in saturated hydraulic conductivity (Ks) with soil depth. We do this for a range of flow rates and mineral reaction kinetics. Our results show that for minerals with a high ratio of equilibrium concentration ( Ceq) to intrinsic weathering rate ( Rmax), vertical heterogeneity in Ks enhances the gradient of weathering‐derived solute concentration in the critical zone and strengthens the inverse stream C‐Q relation. As CeqRmax decreases, the spatial distribution of concentration in the critical zone becomes more uniform for a wide range of flow rates, and stream C‐Q relation approaches chemostatic behavior, regardless of the degree of vertical heterogeneity in Ks. These findings suggest that the transport‐controlled mechanisms in the hillslope can lead to chemostatic C‐Q relations in the stream while the hillslope surface reaction‐controlled mechanisms are associated with an inverse stream C‐Q relation. In addition, as CeqRmax decreases, the concentration in the critical zone and stream become less dependent on groundwater age (or transit time).

Boreal headwater streams have been identified as hot spots for evasion of greenhouse gases (GHGs). This study was the first to systematically determine the concentrations of CO 2 and CH 4 in hemiboreal headwater streams. The use of a headspace sampling method focusing on GHGs in combination with a statistically representative selection of more than 200 streams across two regions in Sweden was the basis for defining the base flow concentrations of CO 2 and CH 4 . All streams were supersaturated relative to the atmosphere in CO 2 and the majority in CH 4 for the 82% of streams in which CH 4 was detected. The spatial variability in both CO 2 and CH 4 was high but positively related to total organic carbon, mean annual temperature, and proportion of peatland in the catchment. There were, however, regional differences in the spatial controls, which are something that predictive models need to consider. The data set allowed for comparison between a headspace and an alkalinity-based method for determining CO 2 . More than 50% of the streams contained no alkalinity which made the alkalinity-based determination of CO 2 impossible. In addition, half of the streams with alkalinity had alkalinities low enough (<0.07 meq L À1 ) to make the CO 2 determination very uncertain. The streams with low pH and no alkalinity contained median CO 2 concentrations that were 45% higher than the streams containing alkalinity. Therefore, large-scale generalizations about CO 2 in such headwaters will be significantly underestimated if (1) headwaters are underrepresented and (2) the headwaters are sampled but CO 2 is calculated from their alkalinity.