Flow pattern and seasonal as well as diurnal temperature variations control ecological and biogeochemical conditions
in hyporheic sediments. In particular, hyporheic temperatures have a great impact on many microbial processes.
In this study we used 3-D coupled water flow and heat transport simulations applying the HydroGeoSphere
code in combination with high frequent observations of hydraulic heads and temperatures for quantifying reach
scale water and heat flux across the river groundwater interface and hyporheic temperature dynamics of a lowland
gravel-bed river. The magnitude and dynamics of simulated temperatures matched the observed with an average
mean absolute error of 0.7 C and an average Nash Sutcliffe Efficiency of 0.87. Our results highlight that the
average temperature in the hyporheic zone follows the temperature in the river which is characterized by distinct
seasonal and daily temperature cycles. Individual hyporheic flow path temperature substantially varies around the
average hyporheic temperature. Hyporheic flow path temperature was found to strongly depend on the flow path
residence time and the temperature gradient between river and groundwater; that is, in winter the average flow
path temperature of long flow paths is potentially higher compared to short flow paths. Based on the simulation
results we derived a general empirical relationship, estimating the influence of hyporheic flow path residence time
on hyporheic flow path temperature. Furthermore we used an empirical temperature relationship between effective
temperature and respiration rate to estimate the influence of hyporheic flow path residence time and temperature
on hyporheic oxygen consumption. This study highlights the relation between complex hyporheic temperature
patterns, hyporheic residence times and their implications on temperature sensitive biogeochemical processes.
Analysis of temperature time series to estimate direction and magnitude of water fluxes in near-surface sediments
Matthias Munz, Sascha E. Oswald, and Christian Schmidt
The application of heat as a hydrological tracer has become a standard method for quantifying water fluxes between
groundwater and surface water. Typically, time series of temperatures in the surface water and in the sediment are
observed and are subsequently evaluated by a vertical 1D representation of heat transport by advection and dispersion.
Several analytical solutions as well as their implementation into user-friendly software exist in order to
estimate water fluxes from the observed temperatures. The underlying assumption of a stationary, one-dimensional
vertical flow field is frequently violated in natural systems. Here subsurface water flow often has a significant horizontal
component.We developed a methodology for identifying the geometry of the subsurface flow field based on
the variations of diurnal temperature amplitudes with depths. For instance: Purely vertical heat transport is characterized
by an exponential decline of temperature amplitudes with increasing depth. Pure horizontal flow would be
indicated by a constant, depth independent vertical amplitude profile. The decline of temperature amplitudes with
depths could be fitted by polynomials of different order whereby the best fit was defined by the highest Akaike
Information Criterion. The stepwise model optimization and selection, evaluating the shape of vertical amplitude
ratio profiles was used to determine the predominant subsurface flow field, which could be systematically categorized
in purely vertical and horizontal (hyporheic, parafluvial) components.
Analytical solutions to estimate water fluxes from the observed temperatures are restricted to specific boundary
conditions such as a sinusoidal upper temperature boundary. In contrast numerical solutions offer higher flexibility
and can handle temperature data which is characterized by irregular variations such as storm-event induced temperature
changes and thus cannot readily be incorporated in analytical solutions. There are several numerical models
that simulate heat transport in porous media (e.g. VS2DH, HydroGeoSphere, FEFLOW) but there can be a steep
learning curve to the modelling frameworks and may therefore not readily accessible to routinely infer water fluxes
between groundwater and surface water.We developed a user-friendly, straightforeward to use software to estimate
water FLUXes Based On Temperatures- FLUX-BOT. FLUX-BOT is a numerical code written in MATLAB that
calculates time variable vertical water fluxes in saturated sediments based on the inversion of measured temperature
time series observed at multiple depths. It applies a cell-centered Crank-Nicolson implicit finite difference
scheme to solve the one-dimensional heat advection-conduction equation (FLUX-BOT can be downloaded from
the following web site: https://bitbucket.org/flux-bot/flux-bot). We provide applications of FLUX-BOT to generic
as well as to measured temperature data to demonstrate its performance.
Both, the empirical analysis of temperature amplitudes as well as the numerical inversion of measured temperature
time series to estimate the vertical magnitude of water fluxes extent the suite of current heat tracing methods and
may provide insight into temperature data from an additional perspective.
Estimation of global plastic loads delivered by rivers into the sea
Christian Schmidt, Tobias Krauth, Phillipp Klöckner, Melina-Sophie Römer, Britta Stier, Thorsten Reemtsma, and Stephan Wagner
A considerable fraction of marine plastic debris likely originates from land-based sources. Transport of plastics by
rivers is a potential mechanism that connects plastic debris generated on land with the marine environment.
We analyze existing and experimental data of plastic loads in rivers and relate these to the amount of mismanaged
plastic waste (MMPW) generated in the river catchments. We find a positive relationship between the plastic load
in rivers and the amount of MMPW. Using our empirical MMPW-plastic river load-relationship we estimated the
annual plastic load for 1494 rivers, ranging from small first order streams to large rivers, which have an outlet to
We estimate that the global load of plastic debris delivered by rivers to the sea is 39000 tons per year with a
large 95% prediction interval between 247 tons per year and 16.7 million tons per year, respectively. Our best
estimate is considerably lower than the estimated total land-based inputs which range between 4.8-12.7 million
tons anually (Jambeck et al. 2015). Approximately 75% of the total load is transported by the 10 top-ranked rivers
which are predominantly located in Asia. These river catchments encompass countries with a large population and
high economic growth but an insufficient waste infrastructure. Reducing the plastic loads in these rivers by 50%
would reduce the global inputs by 37%.
Of the total MMPW generated within river catchments, only a small fraction of about 0.05 % has been found to
be mobile in rivers. Thus, either only a small fraction of MMPW enters the river systems, or a substantial fraction
of plastic debris accumulates in river systems world wide.
C, N, P export regimes in rivers from headwater to downstream catchments
Rémi Dupas, Andreas Musolff, James W. Jawitz, P. Suresh C. Rao, Jan H. Fleckenstein, Michael Rode, and Dietrich Borchardt
Excessive amounts of nutrients and dissolved organic matter in freshwater bodies affect aquatic ecosystems. In this study, the spatial and temporal variability of nitrate (NO3), dissolved organic carbon (DOC) and soluble reactive
phosphorus (SRP) was analyzed along the Selke river continuum from 1 – 3 km2 headwater catchments to 184 – 456 km2 downstream catchments, within the TERENO Harz/Central German Lowland Observatory. Three
headwater catchments were selected as archetypes of the main landscape units (land use x soil type) present in the Selke catchment. Export regimes in these catchments were interpreted in terms of NO3, DOC and SRP landto-
stream transfer processes. Differences between export regimes in headwater and downstream catchments were interpreted in terms of in-stream processes and contribution of point source emissions. The results showed that
the NO3 seasonal dynamics were opposite compared to DOC and SRP in all three headwater catchments. These dynamics were interpreted as the result of the interplay of hydrological and biogeochemical processes, for which
riparian wetlands were hypothesized to play a determining role. In the two downstream catchments, NO3 was transported almost conservatively, except during the summer period where in-stream retention could exceed 50%.
Allochtonous DOC was consumed in the upstream river section (with low light and nutrient availability) and autochthonous DOC was produced in the downstream river section (with high light and nutrients availability); the
natural export regime of SRP mimicked a point source signal, which may lead to misattribution and thus overestimation of domestic contribution to phosphorus loads in rivers. Monitoring the river continuum from headwater to
downstream rivers proved effective to investigate jointly land-to-stream and in-stream transport and transformation processes.
Using a physically-based water flow model to explore the dynamics of transit times and mixing in a small agricultural catchment
Jie Yang, Ingo Heidbüchel, Andreas Musolff, and Jan H. Fleckenstein
Catchment-scale transit time distributions (TTDs) for discharge and residence time distributions of the water in storage (RTDs) are promising tools to characterize the discharge and mixing behavior of a catchment and can help to interpret the associated solute loads to the stream in a spatially implicit way. TTDs and RTDs are dynamic in time, influenced by dynamic rainfall and evapotranspiration forcing, and changing groundwater storage in the catchment. In order to understand the links between the dynamics of TTDs and groundwater mixing in the small agricultural catchment Schäfertal, in central Germany, a 3D hydrological model was set up for the catchment using the fully coupled surface-subsurface numerical model HydroGeoSphere (HGS). The model is calibrated using discharge and groundwater level measurements, and runs transiently for a period of 10 years from 1997 to 2007. A particle tracking tool was implemented in HGS to track the movement of water parcels in the subsurface, outputting TTDs of channel discharge and RTDs of groundwater storage at daily intervals.
Results show that the mean age of the discharge water is significantly younger than that of the water in storage, indicating a poorly mixed subsurface. Discharge preferentially samples faster flowing younger water originating from the more conductive top parts of the aquifer. Spatial variations of the age of water in storage are observed, highly influenced by aquifer heterogeneity. Computed StorAge Selection (SAS) functions [Rinaldo et al. 2015] show clear shifts in the discharge sampling preferences between wet and dry states: during wet states in winter and spring, discharge has a preference for younger water because the shallow flow paths are active due to high groundwater levels and low evapotranspiration. Conversely, during dry states in summer and autumn, discharge has a preference for older water because the shallow flow paths are inactive due to low groundwater levels and stronger evapotranspiration. Measured nitrate (NO3) loads in discharge, mainly originating from fertilizer in shallow soils, decrease significantly with decreasing wetness of the catchment. This trend confirms the shifts of discharge sampling preferences between wet and dry states.
The Selke field site, Central Germany
Nico Trauth, Andreas Musolff, Christian Schmidt, Michael Vieweg, Toralf Keller, Ulrike Werban, Kay Knöller, Erik Nixdorf, and Jan H. Fleckenstein
Objectives and Research Questions:
•Detailed understanding of groundwater – river water exchange dynamics at the meander to reach scale
•Quantifying water and solute fluxes across this interface
•Delineating controls of turnover of redox-sensitive compounds (carbon, oxygen, nitrate) by aerobic and anaerobic reactions
•Spatial extent and temporal dynamics of water exchange, solute transport and reactive zones
Stream discharge events increase the reaction efficiency of the hyporheic zone of an in-stream gravel bar
Nico Trauth, Christian Schmidt, and Jan H. Fleckenstein
Streambed structures such as dunes, pool-riffles or bars enhance the exchange of stream water and solutes with the subsurface, the hyporheic zone. Prior studies have evaluated the factors which control hyporheic exchange and biogeochemical processes for steady state hydrological conditions using numerical models. However, the impact of natural discharge variability on water and solute exchange, creating hydraulically specific conditions for the reactions in the shallow streambed, has not been studied thus far.
In our study, we set up a transient flow and reactive transport model to elucidate the impact of single stream discharge events on water exchange, solute transport and reactions within the hyporheic zone of an in-stream gravel bar. The discharge events were varied by their duration and the maximum stream discharge. Temporally variable hydraulic heads were assigned as hydraulic head boundary conditions at the top of the reactive groundwater model MIN3P. A steady ambient groundwater flow field was introduced by lateral upstream and downstream hydraulic head boundaries, generating in combination with the stream water level, losing, neutral, or gaining stream conditions. Stream water borne dissolved oxygen, dissolved organic carbon and nitrate can infiltrate across the top of the modelling domain, where aerobic respiration and denitrification are simulated.
Our results show that water and solute exchange through the hyporheic zone (only stream water that infiltrates into the subsurface and exfiltrates back to the stream) is highly dependent on the interplay between event characteristics and the ambient groundwater level. In scenarios where the stream discharge shifts the hydraulic system to strong and long-lasting losing conditions, hyporheic flow paths are longer and the extent of the hyporheic zone deeper than under base flow conditions and small hydrologic events where gaining conditions prevail. Consequently, stream discharge events may temporally lead to increased reactivity efficiency (the proportion of solute consumption of solute influx) of the hyporheic zone by approximately 2.0 and 3.8 times for aerobic respiration and denitrification, respectively.
A New Optical Oxygen Sensor Reveals Spatial and Temporal Variations of Dissolved Oxygen at Ecohydrological Interfaces
Tanja Brandt, Michael Vieweg, Astrid Harjung, Jan H. Fleckenstein, and Christian Schmidt
The spatial and temporal distribution of dissolved oxygen (DO) at highly reactive aquatic interfaces, e.g. in the hyporheic zone (HZ), is a primary indicator of redox and interlinked biogeochemical zonations. However, continuous measuring of DO over time and depths is challenging due to the dynamic and potentially heterogenic nature of the HZ. We further developed a novel technology for spatially continuous in situ vertical oxygen profiling based on optical sensing (Vieweg et al, 2013). Continuous vertical measurements to a depth of 50 cm are obtained by the motor-controlled insertion of a side-firing Polymer Optical Fiber (POF) into tubular DO probes. Our technology allows minimally invasive DO measurements without DO consumption at high spatial resolution in the mm range. The reduced size of the tubular probe (diameter 5 mm) substantially minimizes disturbance of flow conditions. We tested our technology in situ in the HZ of an intermittent stream during the drying period. Repeated DO measurements were taken over a total duration of six weeks at two locations up- and downstream of a pool-cascade sequence. We were able to precisely map the spatial DO distribution which exhibited sharp gradients and rapid temporal changes as a function of changing hydrologic conditions. Our new vertical oxygen sensing technology will help to provide new insights to the coupling of transport of DO and biogeochemical reactions at aquatic interfaces.
A physically-based numerical model of catchment water flow to evaluate dominant controls of residence time distributions
Markus Neubauer, Andreas Musolff, and Jan H. Fleckenstein
The residence time distribution (RTD) of water within catchments is a tool to describe catchment-scale transport and mixing. To better understand dominant controls of RTDs and their time-invariant behavior we use a fully coupled three-dimensional numerical model of surface and subsurface flow. We carefully calibrate the model to the data-rich Schäfertal catchment in Central Germany and use this model as a base case for numerical experiments. Here we present first results of systematical variations of the model domain's bottom geometry and groundwater recharge rates. RTDs are derived on basis of post processing particle tracking algorithm. Results reveal a shift between unimodal and bimodal RTDs depending on the model volume and substantial drifts of water ages with changing groundwater recharge. Further numerical experiments will evaluate RTDs in the interplay between catchment's structural properties and dynamic climatic forcing.
Deriving variable travel times and aerobic respiration in the hyporheic zone using electrical conductivity as natural tracer
Michael Vieweg, Jan H. Fleckenstein, Andreas Musolff, and Christian Schmidt
Determining oxygen consumption (respiration) rates is important for characterizing the ecological functioning of a stream. It is known, that respiration is strongly temperature dependent, but the variability over time and the effects of changing hydrologic conditions are still scarce. Existing respiration measuring methods mostly utilize ex situ respiration chambers, which do not necessarily represent the actual conditions in a riverbed.
We present an approach of transient in situ measurements, which utilize changes in the natural stream-EC signal as tracer for the advective transport in the streambed and combine these with precise oxygen measurements.
LTC Logger and optode based oxygen logger were installed in the stream and at 45cm depth beside an in-stream gravel bar. Streambed adapted probe rods with a screened section of 2 cm ensuring a minimized flow-through volume hold the loggers which were programmed to 5min interval measuring interval. Diurnal changes in the EC signal are considered to be quasi-conservative and were tracked in the subsurface. A windowed cross correlation approach was utilized to derive a time-resolved advective travel-time. Assuming a one dimensional flow-path from the stream into the sediment, the time-shift in the EC signal is interpreted as the peak travel time of a tracer breakthrough curve. Additionally a moving average filter of variable length was applied to the stream EC signal, to account for dispersion and further maximize the correlation. For obtaining an experimental respiration rate, the physical transport conditions are then applied to the oxygen data, assuming a first order decay.
The results show that the natural EC signal is applicable as tracer, as long as the measurements show distinctive fluctuations. The cross correlation revealed transient travel times with a range between 1-7h (mean 4h) at the upstream and 8-18h (mean 11h) at the downstream location of the gravel bar. There are strong indications, that the stream level is affecting the travel-time, albeit these effects differ depending on the morphology and strength of the streamflow events. The derived travel times allowed for estimating a transient respiration rate between 3 and 12 mg/l/day. Temperature was found to control over 70% of the variation of the respiration rate. The oxygen concentration in the streambed is more influenced by the variability of the respiration rate than of the travel time.
Investigating DOC export dynamics using high-frequency instream concentration measurements
Marieke Oosterwoud, Toralf Keller, Andreas Musolff, Sven Frei, Ji-Hyung Park, and Jan H. Fleckenstein
Being able to monitor DOC concentrations using in-situ high frequency measurements makes it possible to better understand concentration-discharge behavior under different hydrological conditions. We developed a UV-Vis probe setup for modified/adapted use under field conditions. The quasi mobile probe setup allows a more flexible probe deployment.
New or existing monitoring sites can easily be equipped for quasi-continuous monitoring or measurements can be performed at changing locations, without the need for additional infrastructure. We were able to gather high frequency data on DOC dynamics for one year in two streams in the Harz mountains in Germany. It proved that obtaining accurate DOC concentrations from the UV-Vis probes required frequent maintenance and probe calibration. The advantage of the setup over standard monitoring protocols becomes evident when comparing net exports over a year. In addition to mass improved balance calculations the high-frequency measurements can reveal intricate hysteretic relationships between discharge and concentrations that can provide valuable insights into the hydrologic dynamics and mechanisms that govern the delivery of DOC to the receiving waters. Measurements with similar probes from two additional catchments in Southern Germany and South Korea will be used to illustrate different discharge-concentration relationships and what can be learned from them about the hydrologic mechanisms that control the dynamics of DOC export.
Influence of varying hydraulic conditions on hyporheic exchange and reactions in an in-stream gravel bar
Nico Trauth, Christian Schmidt, Uli Maier, and Jan H. Fleckenstein
Hyporheic exchange transports solutes into the subsurface where they can undergo biogeochemical transformations, affecting fluvial water quality and ecology. A three-dimensional numerical model of a natural in-stream gravel bar (20 m 3 6 m) is presented. Multiple steady state streamflow is simulated with a computational fluid dynamics code that is sequentially coupled to a reactive transport groundwater model via the hydraulic head distribution at the streambed.
Ambient groundwater flow is considered by scenarios of neutral, gaining, and losing conditions. The transformation of oxygen, nitrate, and dissolved organic carbon by aerobic respiration and denitrification in the hyporheic zone are modeled, as is the denitrification of groundwater-borne nitrate when mixed with stream-sourced carbon. In contrast to fully submerged structures, hyporheic exchange flux decreases with increasing stream discharge, due to decreasing hydraulic head gradients across the partially submerged structure. Hyporheic residence time distributions are skewed in the log-space with medians of up to 8 h and shift to symmetric distributions with increasing level of submergence. Solute turnover is mainly controlled by residence times and the extent of the hyporheic exchange flow, which defines the potential reaction area. Although streamflow is the primary driver of hyporheic exchange, its impact on hyporheic exchange flux, residence times, and solute turnover is small, as these quantities exponentially decrease under losing and gaining conditions. Hence, highest reaction potential exists under neutral conditions, when the capacity for denitrification in the partially submerged structure can be orders of magnitude higher than in fully submerged structures.
Impact of model geometry and recharge rates in catchment's residence time distributions - numerical experiments
Markus Neubauer, Andreas Musolff, and Jan H. Fleckenstein
Residence time distributions (RTD) of water in catchments are promising tools to characterize and model solute transport on a larger scale. In the last decade, much research has been conducted on the estimation and the application of RTD's. However, there are still some major issues to be addressed to complex derivation, parameterization and transient behavior. Through improved remote sensing data, the surface elevation can mostly be resolved in detail, while subsurface volumes and boundaries remain highly undetermined.
Our objectives are to systematically evaluate the impact of different depths and geometries of the domain bottom and groundwater recharge rates on RTD's. The study site is a small (1.6 km2) headwater catchment located within the Harz Mountains, Germany. For this catchment long time series of climate, discharge and hydrochemistry are available while groundwater flow field and subsurface structure are less known. The site is intensively influenced by agricultural land use and exhibits strong seasonal dynamics of water flow and hydrochemistry due to the snowmelt. The modeling was performed using HydroGeoSphere, a coupled surface and subsurface model, which solves the Richards Equation for variable saturated soils. The Open Source software Paraview and R was chosen as postprocessors to perform and analyze forward particle tracking algorithms under steady state conditions. Ten depth and geometry scenarios of the domain bottom were created (5 horizontal bottom geometries - constant base and 5 variable bottom geometries - parallel to surface topography; both minimum depths ranging from 2 m to 50 m). The model's internal structure was discretized by two homogenous layers (averaged catchment representation) parallel to the input digital elevation model (2x2 m). The geometry scenarios were combined with fifteen steady state simulations for different groundwater recharge rate scenarios (0.1 mm up to 15 mm per day). Model results indicate a strong influence of the chosen model geometry on the RTD's. In cases of horizontal lower boundaries with large model volumes, RTD's are characterized by bimodal behavior, indicating that two distinct main pathways (short and long) dominate within this catchment. The shallower the horizontal bottom geometry the shorter are the particle paths without losing the bimodal behavior. Variable bottom geometry setups with small model volumes result in unimodal RTD's with a significant shorter mean residence time (RT). Higher recharge rates results in a shift to the younger ages. At the same time the variance of the observed particle ages increase. Our result highlights that too shallow model domains truncate longer flow paths and lead to unimodal instead of bimodal RTD's. Although, the subsurface is represented by a simple 2-layer model, highly resolved topography results in complex bimodal RTD's. RT changes in recharge scenarios hint to a significant transient behavior of RTD's under seasonal recharge changes. Further work will concentrate on transient behavior, transient particle tracking, relation to hydrochemistry and the impact of soil heterogeneity.