the heterogeneity of runoff and its significance for water quality problems
TRANSCRIPT
This article was downloaded by: [Universitaetsbibliothek Wuerzburg]On: 01 November 2014, At: 07:19Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registeredoffice: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
Hydrological Sciences JournalPublication details, including instructions for authors andsubscription information:http://www.tandfonline.com/loi/thsj20
The heterogeneity of runoff andits significance for water qualityproblemsWOLFHARD SYMADER a & REINHARD BIERL aa Hydrological Department , University of Trier , D-54228, Trier,GermanyPublished online: 25 Dec 2009.
To cite this article: WOLFHARD SYMADER & REINHARD BIERL (1998) The heterogeneity of runoffand its significance for water quality problems, Hydrological Sciences Journal, 43:1, 103-113,DOI: 10.1080/02626669809492105
To link to this article: http://dx.doi.org/10.1080/02626669809492105
PLEASE SCROLL DOWN FOR ARTICLE
Taylor & Francis makes every effort to ensure the accuracy of all the information (the“Content”) contained in the publications on our platform. However, Taylor & Francis,our agents, and our licensors make no representations or warranties whatsoever as tothe accuracy, completeness, or suitability for any purpose of the Content. Any opinionsand views expressed in this publication are the opinions and views of the authors,and are not the views of or endorsed by Taylor & Francis. The accuracy of the Contentshould not be relied upon and should be independently verified with primary sourcesof information. Taylor and Francis shall not be liable for any losses, actions, claims,proceedings, demands, costs, expenses, damages, and other liabilities whatsoeveror howsoever caused arising directly or indirectly in connection with, in relation to orarising out of the use of the Content.
This article may be used for research, teaching, and private study purposes. Anysubstantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,systematic supply, or distribution in any form to anyone is expressly forbidden. Terms& Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions
Hydrological Sciences—Journal—des Sciences Hydrologiques, 43(1) February 1998 103
The heterogeneity of runoff and its significance for water quality problems
WOLFHARD SYMADER & REINHARB BIERL Hydrological Department, University of Trier, D-54228 Trier, Germany
Abstract In a small heterogeneous limestone basin with mixed land use twelve flow components were distinguished by using chemographs of flood events and baseflow. Only a few of them are endmembers. The number of flow components depends on the tracers that are used and on the degree of heterogeneity. If the concept of flow components is restricted to endmembers only, the runoff generation process cannot be explained completely because runoff does not form by the mixing of chemically distinct water types alone. Beside water types, there are all kinds of temporal variations ranging from steep gradients to continuous changing, e.g. due to incomplete mixing, exhaustion of sources, etc. This means that the original concept of flow components should be extended to behaviour patterns in order to obtain a basic understanding of the very complex process of runoff generation. Furthermore it is an excellent approach to facilitate understanding of the temporal variation of water quality in flowing waters.
L'hétérogénéité de l'écoulement et sa signification pour les problèmes de qualité de l'eau Résumé Dans un petit bassin calcaire hétérogène dont l'occupation des sols est diversifiée, douze composantes de l'écoulement ont pu être distinguées en utilisant les chemogrammes de diverses crues et de l'écoulement de base. Peu d'entre elles correspondent à des réservoirs contributifs. Le nombre de composantes dépend des traceurs utilisés et du degré d'hétérogénéité. Si l'on réduit le concept de composante à celui de réservoir contributif, le processus de formation du débit ne peut pas être complètement expliqué car l'écoulement ne procède pas seulement du mélange de différentes eaux. A côté des différents types d'eau, il y a toutes sortes de variations allant de gradients brutaux à des modifications continues, dues à des mélanges imparfaits, au tarissement de certaines sources etc. Cela signifie que le concept de composantes de l'écoulement doit être étendu aux types de comportements afin d'atteindre une connaissance approfondie du très complexe processus de formation de l'écoulement. Au delà, il s'agit là d'un excellente approche en vue de comprendre les variations au cours du temps de la qualité des eaux courantes.
PROBLEM AND OBJECTIVES
An analysis of flood wave responses to precipitation suggested the existence of flow components. From the shape of the hydrograph using response time as a criterion a separation into quick flow, interflow and baseflow was made. Using radioactive isotopes or chemical tracers it was possible to distinguish between the two components of old and new water with old water being groundwater and new water being rain water. A third component which was labelled soil water existed somewhere between old and new water (Stichler & Herrmann, 1978; Kennedy et al., 1986). The general approach to model the behaviour of flow components was based on the
Open for discussion until 1 August 1998
Dow
nloa
ded
by [
Uni
vers
itaet
sbib
lioth
ek W
uerz
burg
] at
07:
19 0
1 N
ovem
ber
2014
104 Wolfhard Symader & Reinhard Bierl
additional assumption that flow components can be considered as endmembers, which change only slowly over time (Cfaristopherson et al, 1990; Hooper et al., 1990; Neal & Cfaristopherson, 1989; Robson et al., 1992).
There is no doubt that this concept has its merits. However, there are two fundamental flaws obvious to everyone dealing with water pollution coming from diffuse sources. First of all, the concept is much too simple. Although it can be argued that a rather small number of sources with unique absolute and relative concentrations of principal and minor anions and cations can be selected (Woolhiser, et al., 1985), three flow components cannot explain a multitude of minor sources and do not explain the temporal variations of dissolved solids in flowing waters. Secondly, it is not proven at all that the total runoff is formed of endmembers. In fact there is a considerable debate on that assumption (Pilgrim et al., 1979; Kennedy et al, 1986; Mulder et al, 1990; Kendall & McDonnell 1993; Chapman et al, 1993). In any case, the number of flow components seems to depend on the identification procedure, and it may be that the number is indefinite, which means that part of the total runoff behaves like a continuum.
So the crucial questions are: (1) Is the assumption that stream water forms by the mixing of flow components a promising working hypothesis? (2) Are these components endmembers or just theoretical constructs that are used to simplify reality? (3) Will the understanding of flow components, as far as they exist, help in understanding the whole system, e.g. the basin or the river?
The answer to the first question depends mainly on the definition of a flow component. There is a general agreement that different stages of runoff can be distinguished due to changes in the properties of the water body. However, the contributing water types show all kinds of temporal patterns ranging from steep sharp gradients to continuous changing, which can hardly be noticed. From this it follows that not all flow components are defined by nature. Some are defined in terms of the characterizing properties. In this paper flow components are defined by patterns of water body characteristics, which is similar to the basic idea of the fingerprinting approach. Questions two and three above are ones to be examined by this study.
THE AREA UNDER INVESTIGATION, MATERIAL AND METHODS
The Kartelbornsbach drains a basin of 2.75 km2 in the southern Eifel mountains about 8 km northwest of Trier. The shallow soils cover Triassic bedrock consisting of clayey, silty or sandy marls or of limestone, which locally contain gypsiferous pockets. A partially collapsed field drain system is active during the whole year. Although its main contribution is soil water and perched groundwater, it can respond very quickly to rainfall events. Land use is arable land, pastures and small patches of scrub and woodland. A small village with an inadequate treatment plant, several streets and a small storm sewer system coming from a cluster of houses not connected to the treatment plant, causes both a continuous and an event-controlled
Dow
nloa
ded
by [
Uni
vers
itaet
sbib
lioth
ek W
uerz
burg
] at
07:
19 0
1 N
ovem
ber
2014
The heterogeneity of runoff and its significance for water quality problems 105
pollution of the small stream. In spite of the calciferous bedrock there are no karst phenomena. Water budget calculations indicate no significant groundwater loss. The effective rainfall is very low in summer and hardly ever exceeds 4%.
The main characteristic of this small basin, under investigation since 1988, is its heterogeneity, which is a necessary prerequisite for the identification or determination of components, sources, or flow paths of runoff, dissolved solids and suspended particles.
Since the summer of 1988, precipitation and discharge have been recorded continuously. From 1988 to 1992 flood waves have been investigated. Bulk samples between 20 and 50 1 were analysed for dissolved material and suspended particles, associated major ions and heavy metals. Suspended particles were characterized by laser grain size and shape analysis (CIS-1, Galai equipment; Aharonson et al., 1986). Analyses for chlorophyll, organic carbon and nitrogen, and occasionally organic contaminants such as PAHs and PCBs (Symader et al, 1991) were made. From 1990 to 1994 dissolved major ions and heavy metals were analysed in daily water samples. Additional projects dealing with individual sources such as field drains or macropores, suspended sediment transport and the temporal variations of particles and associated contaminants in river bottom sediments were included later. The main idea of the sampling programme has been given by Symader (1992). Field work is still in progress.
RESULTS AND DISCUSSION
Flood response
The typical flood response to showers and thunderstorms in summer is a small preflush followed by a sharp single peaked flood wave with an exponential recession limb. What looks quite simple is a very complex response, as can be shown by different chemographs (Fig. 1).
The succession of flow components starts with rain water. The corresponding suspended particles are leaf remnants or small twigs. It is followed by surface runoff from the rural pathway system that brings a yellow material into the river. In Fig. 1 this component is characterized by the first peak of zinc and the small trough of sodium (sample no. 3). About 3-5 min later suspended material turns black indicating the preponderance of water from the small storm sewer system. This waste water component is by no means an endmember, because, as can be seen very clearly in the Figure, it starts with high concentrations of zinc, nitrate and ammonium, but changes rapidly thereafter. It is assumed that there is a first flush of rain water that travels much faster in the sewer than does the remobilized mud. After this flush nitrate drops to zero, but ammonium remains at a high level. Concentrations of oxygen become very low as well. The significance of this component decreases with the increasing proportion of the following components, but can be traced well beyond the main peak.
Dow
nloa
ded
by [
Uni
vers
itaet
sbib
lioth
ek W
uerz
burg
] at
07:
19 0
1 N
ovem
ber
2014
106 Wolfliard Symader & Reinhard Bierl
Dissolved Solids N03 NH4
[mg/l]
HCO3 Ca [mg/l]
400
300
200 +40
100
0
Na [mg/l]
20
15
10
5
• 0 19uu 20™ 21u
Fig. 1 Flood wave at Kartelbornsbach, 30 May 1992.
Saturation overland flow with brownish suspended material from top soil is responsible for a very sharp peak in calcium and magnesium during the rising limb of the flood wave. In Fig. 1 calcium shows only some kind of a shoulder, because the sample was not taken at the precise moment of the peak. Saturation overland flow is followed by the replacement of soil water from the near stream zone. It is characterized by peaks of manganese and iron. Manganese responds faster and does not require the same amount of rainfall as does iron. This corresponds to the different solubilities of manganese and iron due to redox conditions. It can be concluded that these flow components come from different soil depths.
The sharp peaks of calcium, manganese and iron suggest that the components belong to replaced water, but the proof that this water can be considered an end-member cannot be given. During the recession stage no further components can be distinguished. The probability of a flow component being an endmember is highest
Dow
nloa
ded
by [
Uni
vers
itaet
sbib
lioth
ek W
uerz
burg
] at
07:
19 0
1 N
ovem
ber
2014
The heterogeneity of runoff and its significance for water quality problems ] 07
during the recession stage, where flush effects do not occur and where no significant changes in the composition of water quality can be detected.
Assuming that the flood wave consists during the recession of only two components, that both are endmembers and that one of them does not vary with discharge, the relationship of load against discharge is a linear one as given by equation (1):
CrQr=CxQT-(C2-Cx)Q2 (1)
with C for concentration, Q for discharge, T for total, 1 and 2 for the two end-members.
This is equivalent to the bivariate linear regression equation:
Y= bX + a
with CT x QT equivalent to Y, QT equivalent to X, Cx to the regression coefficient b, and (C2 - Q) x Q2 equivalent to a. Nakamura (1971) used this concept to separate interflow from baseflow.
For the example discussed here, the two components are labelled soil water and baseflow. Baseflow includes perched groundwater and waste water and is assumed to remain constant in concentration and discharge during the recession stage of an event.
so4
[mg/s],
2000-
1500-
1000-
50 Q [l/s]
Fig. 2 Sulphate loads vs discharge; recession limb of an event on 22 September 1990.
Dow
nloa
ded
by [
Uni
vers
itaet
sbib
lioth
ek W
uerz
burg
] at
07:
19 0
1 N
ovem
ber
2014
108 Wolftiard Symader & Reinhard Bierl
[mg/s]
6000
5000'
4000 H
/ Ca
3000-
— I — 100 200 Q [l/seo]
Fig. 3 Loads of sulphate and calcium against discharge; recession limb of an event on 23 July 1989.
In Figs 2 and 3 the relationships between load and discharge are shown for the recession limbs of two flood waves of quite different magnitudes. As a more or less linear relationship exists, when the discharge remains below 120 1 s"1 the concentrations of calcium and sulphate can be derived from the regression coefficient b and range from 26 to 30 for most of the events that have been investigated. This subsurface component is considered to be an endmember. The increasing loads at the very end of the flood wave indicate that the replacement of soil water may be the controlling factor, as both calcium and sulphate concentrations can be assumed to increase with soil depth. For ions enriched at the soil surface, e.g. potassium, no linear relationships could be found. These results fit well to those of Hirata & Muraoka (1993).
When the discharge exceeds 120 1 s"1, the situation is less clear. There is no doubt that the three assumptions made above do not reflect reality. Either there are more than two components involved or one of the components is not an endmember. In many but not in all summer events, a small minimum of the loads could be observed at discharges about 120-150 Is 4 . The flood response of 23 July 1989, which was discussed in detail elsewhere (Strank, 1992; Symader et al., 1992) shed some light on this problem. It was the second flood wave of three events that
Dow
nloa
ded
by [
Uni
vers
itaet
sbib
lioth
ek W
uerz
burg
] at
07:
19 0
1 N
ovem
ber
2014
The heterogeneity of runoff and its significance for water quality problems 109
occurred during one day. As most of the sediment supply was removed during the first event of the day, secondary particle sources could be seen more clearly. The sample with the minimum load (Fig. 3) was a sample with a major maximum of turbidity and a secondary maximum of suspended particle concentrations. From this combination of characteristics (decreased loads and high concentrations of very fine particles) it can be concluded that during the recession stage at least one component was moving rapidly and had a short residence time. This suggests a contribution of subsurface flow from soil pipes or parts of the field drain system.
Dry weather flow
Using the concept discussed above for daily analyses of dissolved solids during dry weather flow, a linear relationship with some slight deviations can be found (Fig. 4). The scatter plot shows three groups of data. A linear relationship with some scattering can be found for discharges less than 12 Is 4 . A close relationship can be found for the range of 12-25 or 30 1 s"1 as well, but the slope is not so steep. When the discharge exceeds 30 1 s"1 any linearity disappears.
The first explanation for the difference between the first two groups is a systematic error of flow measurements, which are very difficult to perform correctly during low flow at this gauging station. However, that was not the case. The difference goes back to a changing hydrological situation.
Dry weather discharge from 12 to 25 1 s"1 is measured during the first half of the year, when the soil releases the water it has received and stored during the winter months. The average concentrations of sulphate of the soil water (derived from the regression equation) are between 55 and 60 mg l"1, which is nearly the double of what could be observed during the recession stage of the flood waves. With decreasing discharge during the summer, this component becomes less and less important. From its behaviour it can be concluded that this flow component is an endmember.
With the beginning of the second half of the year the slope of the loads vs
50
40
r 30 CO
JE
•S 20
10
t,..
"*P
0 5 10 15 20 25 30 35 40 45 50 55 60
discharge [I/s]
Fig. 4 Mineral loads vs discharge for dry weather flow for 1990 and 1991.
Dow
nloa
ded
by [
Uni
vers
itaet
sbib
lioth
ek W
uerz
burg
] at
07:
19 0
1 N
ovem
ber
2014
110 Wolfhard Symader & Reinhard Bierl
40 -
35-
30
ta
S 25 H <o
S 20
10 • 1' U r L J ^ ^ af
3000
2500
h 200C
1500 —
1000
h 500
1.1.91 1.3.91 1.5.91 1.7.91
time
1.9.91 1.11.91 1.1.92
Fig. 5 Sulphate loads and discharge in daily samples during 1991.
discharge relationship still decreases but at a lesser degree. The reason for this are increasing concentrations of several ions e.g. sulphate and nitrate, which cannot be explained by decreasing discharge, whereas the concentrations of the carbonate species do not change very much.
The temporal behaviour of sulphate for 1991 is shown in Fig. 5. The effect discussed above is most clearly seen during September and October 1991. It is suggested that the increase of sulphate and nitrate goes back to soil water or groundwater coming from parts of the basin that are not adjacent to the river and take nearly 6-8 months of travel time to reach the river. This water can move through cracks, macropores and field drains. It is labelled lateral groundwater flow with its origin in the limestone area, where most of the gypsiferous pockets are located. This interpretation is supported by the fact that two ions from different sources (sulphate from gypsic minerals, nitrate from liquid manure and fertilizers) show the same temporal behaviour.
Although not all details are understood it is evident that the baseflow consists of several waters with different properties. This supports some critical comments by Kennedy et al. (1986), who doubt that the isotopic and chemical composition of prestorm soil water can readily be predicted. This has serious consequences for the application of isotopes in simple mixing models. The situation is very complex and old water or baseflow are not homogeneous flow components.
CONCLUSIONS
The analysis of chemographs during events and dry weather flow conditions revealed at least twelve different flow components. They are:
Dow
nloa
ded
by [
Uni
vers
itaet
sbib
lioth
ek W
uerz
burg
] at
07:
19 0
1 N
ovem
ber
2014
The heterogeneity of runoff and its significance for water quality problems \ \ \
1. Continuous waste water effluents, which are part of the baseflow and have not been discussed in this paper. They are responsible for nearly all phosphate during dry weather flow with more or less constant concentrations for the summer half of the year.
2. Perched groundwater. During this analysis there was no need to separate this component from waste water. Both were considered as the constant part of the baseflow. If a quantitative distinction is needed it can be easily done by using the flow rates of the waste water plant and a simple mixing model with phosphate as tracer. Both components are endmembers.
3. Lateral groundwater flow. From samples of deep soil water near the gypsiferous pockets it seems that this groundwater component is no endmember, but increases its concentration continuously during late summer and autumn. Beside high concentrations of sulphate it contains increased concentrations of nitrate. The proportion of this flow component is so small that it can only be detected under very low flow conditions.
4. Deep soil water, or the upper part of the perched groundwater. This water has been stored during the winter months and is thoroughly mixed. It is released during the first half of the year. The situation during dry weather flow can be modelled quite well using a simple mixing model with two components up to a discharge of 30 1 s'1. At higher discharge rates there are additional components involved, but the situation during the winter is not yet well understood.
5. Soil water during the recession stage of summer floods. Again this component is an endmember. It seems to replace at least part of the groundwater.
6. Soil water from the collapsed field drain system or from soil pipes. The existence of this component seems to be established, but for the time being its properties cannot be quantified.
7. Deep soil water or groundwater from the valley bottom. This flow component is characterized by high iron concentrations. Its short-term behaviour seems to indicate that it is replaced by soil water from adjacent parts of the basin.
8. Soil water from the valley bottom. The high concentrations of manganese leading those of iron indicate that this component moves on top of component 6.
9. Saturation overland flow. The indicator is calcium. Its occurrence in succession with the components 7 and 8 forms a basic pattern of flood wave responses In summer. For the time being it is assumed that the height of the calcium peak is supply controlled, therefore it cannot be an endmember, but the existing data are not sufficient to understand inter-storm variations. This component is involved during most of the event, but can be clearly recognized only at the beginning.
10. Waste water from the storm sewer system. This flow component belonging to the small preflush is a very heterogeneous component and shows by no means a constant temporal behaviour, although it is easily recognized.
11. Surface water from the rural pathway system. A short-lived component with chemical characteristics that are supply controlled.
12. Rain water. As long as there are no significant changes during the precipitation event, this very small component is an endmember.
Dow
nloa
ded
by [
Uni
vers
itaet
sbib
lioth
ek W
uerz
burg
] at
07:
19 0
1 N
ovem
ber
2014
112 Wolfhard Symader & Reinhard Bierl
By discriminating these twelve flow components the basin is treated as a lumped system. With additional chemical tracers such as phosphate, sodium, chloride and organic contaminants it is possible to determine the contribution of small areas to total runoff, which considerably increases the number of components that can be identified. However, from comparisons with the results of other basins it can be seen that the number of flow components that can be distinguished decreases with increasing basin size.
The answers to the two questions which were the objectives of this paper can now be given. If the concept of flow components is not restricted to endmembers, but extended to behaviour patterns, it is a very powerful tool in order to understand the runoff generation process or the variations in water quality. Some of the components are indeed endmembers and can be defined in physical terms, but others are not and are more or less characterized by the sampling programme or the methods applied. In some cases, e.g. saturation overland flow, it is only a label that disguises considerable spatial and temporal variations. By using the concept of flow components, the very complex process of runoff generation can be disaggregrated into subpatterns or simple structures, which is the first step towards understanding the whole system. Other working groups are to be encouraged to give more attention to heterogeneous basins.
REFERENCES
Aharonson, E. F., Karasikov, N., Roitberg, M. & Shamir, J. (1986) Galai-CIS-1 a novel approach to aerosol particle size analysis. / . Aerosol Sci. 17, 530-536.
Chapman, P. J., Wheater, H. S. & Reynolds, B. (1993) The effect of geochemical reactions along flowpaths on storm water chemistry in headwater catchments. In: Tracers in Hydrology (ed. by N. E. Peters, E. Hoehn, Ch. Leibundgut, N. Tase & D. E. Walling) (Proc. Yokohama Symp., July 1993), 23-30. IAHS Publ. no. 215.
Christophersen, N., Neal, C , Hooper, R. P., Vogt, R. D. & Andersen, S. (1990) Modelling streamwater chemistry as a mixture of soilwater end-member—a step towards second-generation acidification models. J. Hydrol. 116, 307-320.
Hirata, T. & Muraoka, K. (1993) The relation between water migration and chemical processes in a forest ecosystem. In: Tracers in Hydrology (ed. by N. E. Peters, E. Hoehn, Ch. Leibundgut, N. Tase & D. E. Walling) (Proc. Yokohama Symp., July 1993), 31-40. IAHS Publ. no. 215.
Hooper, R. P., Peters, N. E. & Christophersen, N. (1990) Modelling stream water chemistry as a mixture of soil water end-members—an application to the Panola Mountain catchment, Georgia, USA. J. Hydrol. 116, 321-343.
Kendall, C. & McDonnell, J. J. (1993) Effect of intrastorm isotopic heterogeneities of rainfall, soil water, and groundwater on runoff modeling. In: Tracers in Hydrology (ed. by N. E. Peters, E. Hoehn, Ch. Leibundgut, N. Tase & D. E. Walling) (Proc. Yokohama Symp., July 1993), 41-48. IAHS Publ. no. 215.
Kennedy, V. C , Kendal], C , Zellweger, G. W., Wyerman, T. A. & Avanzino, R. J. (1986) Determination of the components of stormflow using water chemistry and environmental isotopes, Mattole River Basin, California. X Hydrol. 84, 107-140.
Mulder, J., Christophersen, N., Hauhs, M., Vogt, R. D., Andersen, S. & Andersen, D. (1990) Water flow paths an inferred from a rainstorm high in seasalts. Wat. Resour. Res. 26, 611-622.
Nakamura, R. (1971) Runoff analysis by electrical conductance of water. J. Hydrol. 14, 197-212. Neal, C. & Christophersen, N. (1989) Inorganic aluminium-hydrogen ion relationships for acidified streams; the role of
mixing processes. Sci. Total Environ. 80, 195-203. Pilgrim, D. H., Huff, D. D. & Steele, T. D. (1979) Use of specific conductance and contact time relations for
separating flow components in storm runoff. Wat. Resour. Res. 15, 329-339. Robson, A., Beven, K. & Neal, C. (1992) Towards identifying sources of subsurface flow: A comparison of components
identified by a physically based runoff model and those determined by chemical mixing techniques. Hydrol. Processes 6, 199-214.
Stichler, W. & Herrmann, A. (1978) Verwendung von Ol8-Messungen fur hydrologische Bilanzierungen. Deutsche Gewasserkundliche Mitteilungen 22, 9-13.
Strunk, N. (1992) Case studies of variations in suspended matter transport in small catchments. In: Sediment /Water
Dow
nloa
ded
by [
Uni
vers
itaet
sbib
lioth
ek W
uerz
burg
] at
07:
19 0
1 N
ovem
ber
2014
The heterogeneity of runoff and its significance for water quality problems \ \ 3
Interactions V, Developments in Hydrology (ed. by B. T. Hart & P. G. Sly), 247-255. Kluwer, Dordrecht, The Netherlands.
Symader, W. (1992) A sampling strategy for an investigation on particle associated contaminants. In: Erosion and Sediment Transport Monitoring Programmes in River Basins (ed. by J. Bogen, D. E. Walling & T. J. Day) (Proc. Oslo Symp., August 1992), 263-268. IAHS Publ. no. 210.
Symader, W., Bierl, R. & Strunk, N. (1991) Die zeitliche Dynamik des Schwebstofftransportes und seine Bedeutung fur die Gewâsserbeschaffenheit. Vom Wasser 77, 159-169.
Symader, W., Bierl, R., Strunk, N. & Hampe, K. (1992) Erosionsprozesse, Schwebstofftransport und partikelgebundene Umwetschadstoffe in kleinen Einzugsgebieten des Raumes Trier. In: Beitràge zur Physischen Géographie und Landeskunde (ed. by M. Millier), 131-170. Flensburger Régionale Studien, Sonderheft 2,
Woolhiser, D. A., Emmerich, W. E. & Shirley, E. D. (1985) Identification of water sources using normalized chemical ion balances: a laboratory test. /. Hydrol. 76, 205-231.
Received 17 August 1995; accepted 20 June 1997
Dow
nloa
ded
by [
Uni
vers
itaet
sbib
lioth
ek W
uerz
burg
] at
07:
19 0
1 N
ovem
ber
2014