ArticlePDF Available

Heavy-Metal Pollution of the River Rhine and Meuse Floodplains in the Netherlands

December 2000
Geologie en Mijnbouw 79(4)

DOI:10.1017/S0016774600021910

Authors:

Utrecht University

The embanked floodplains of the lower Rhine river in the Netherlands contain large amounts of heavy metals, which is a result of many years of deposition of contaminated overbank sediments. The metal pollution varies greatly between the various floodplain sections as well as in vertical direction within the floodplain soil profiles. The present contribution describes the key processes producing the spatial variability of the metal pollution in floodplain soils: (1) spatial patterns of the concentrations and deposition of Cd, Cu, Pb and Zn during a single flood, which have been determined from samples collected after a high-magnitude flood event; (2) the pollution trends of the lower Rhine over the past 150 years, which were reconstructed on the basis of metal concentrations in sediments from small ponds within the floodplain area. During the flood the largest metal depositions (0.03 g/m 2 Cd, 0.7 g/m 2 Cu, 1.1 g/m 2 Pb and 5.0 g/m 2 Zn for the Rhine) occurred along the natural levees, decreasing to about one third of these values at larger distance from the river. Deposition of heavy metals occurred since the end of the nineteenth century. Periods of maximum pollution occurred in the 1930s and 1960s, when Cu, Pb and Zn concentrations were about 6–10 times as high as background values. The resulting metal distribution in the floodplain soil profiles is illustrated by means of typical examples. Maximum metal concentrations in floodplain soils vary from 30 to 130 mg/kg for Cu, from 70 to 490 mg/kg for Pb, and from 170 to 1450 mg/kg for Zn. The lowest metal pollution is found in the distal parts of floodplain sections with low flooding frequencies, where average sedimentation rates have been less than about 5 mm/a. The largest metal accumulations occur in low-lying floodplain sections where average sedimentation rates have been more than 10 mm/a.

. Estimated ranges of sedimentation rates for different groups of floodplain sites, based on the depths of onset of pollution and maximum pollution.

…

Cross-section and heavy-metal profiles of the Amerongen (AM) floodplain section.

…

Site characteristics of the locations of the investigated soil profiles.

…

Contents of clay, organic matter and heavy metals in sediments deposited on the Waal and Meuse floodplains during the flood of De- cember 1993.

…

Figures - uploaded by Hans Middelkoop

Content may be subject to copyright.

Content uploaded by Hans Middelkoop

Content may be subject to copyright.

Introduction

Due to pollution of the Rhine and Meuse rivers over

many years, considerable amounts of heavy metals

have accumulated in the overbank sediments of the

embanked floodplains of their lower river reaches. Sev-

eral studies have been carried out in order to map the

metal pollution of these floodplains (a.o., Hoogerwerf,

1992). It is often attempted in such inventory studies

to predict metal concentrations in the floodplain soils

by establishing empirical relationships with factors that

are assumed to determine differences in sedimentation

rate. In spite of some successful results obtained at

small floodplain sections (Rang et al., 1987; Leenaers,

1989; Burrough et al., 1993; Rikken & Van Rijn, 1993),

Hoogerwerf (1992) found – in a large-scale inventory

for the entire Rhine and Meuse floodplain – only weak

correlations for metal concentrations in surface sam-

ples versus flooding frequency and local elevation. In

addition, the amounts of pollutants appeared at some

places to be much higher at several decimetres depth

than at the surface. Because of the apparent low pre-

dictive value of such statistical relationships for surface

samples, a better knowledge was needed of the proces-

Geologie en Mijnbouw / Netherlands Journal of Geosciences 79(4) 2000 411

Geologie en Mijnbouw / Netherlands Journal of Geosciences 79 (4): 411-428 (2000)

Heavy-metal pollution of the river Rhine and Meuse floodplains

in the Netherlands

H. Middelkoop1

1Centre for Geo-ecological Research – ICG, Utrecht University, Department of Physical

Geography, P.O. Box 80.115, 3508 TC UTRECHT, the Netherlands;

e-mail: h.middelkoop@geog.uu.nl

Manuscript received: 11 November 1998;accepted in revised form: 9 June 2000

Abstract

The embanked floodplains of the lower Rhine river in the Netherlands contain large amounts of heavy metals, which is a re-

sult of many years of deposition of contaminated overbank sediments.The metal pollution varies greatly between the various

floodplain sections as well as in vertical direction within the floodplain soil profiles. The present contribution describes the

key processes producing the spatial variability of the metal pollution in floodplain soils: (1) spatial patterns of the concentra-

tions and deposition of Cd, Cu, Pb and Zn during a single flood, which have been determined from samples collected after a

high-magnitude flood event; (2) the pollution trends of the lower Rhine over the past 150 years, which were reconstructed on

the basis of metal concentrations in sediments from small ponds within the floodplain area. During the flood the largest metal

depositions (0.03 g/m2Cd, 0.7 g/m2Cu, 1.1 g/m2Pb and 5.0 g/m2Zn for the Rhine) occurred along the natural levees, de-

creasing to about one third of these values at larger distance from the river. Deposition of heavy metals occurred since the end

of the nineteenth century. Periods of maximum pollution occurred in the 1930s and 1960s, when Cu, Pb and Zn concentra-

tions were about 6-10 times as high as background values.

The resulting metal distribution in the floodplain soil profiles is illustrated by means of typical examples. Maximum metal

concentrations in floodplain soils vary from 30 to 130 mg/kg for Cu, from 70 to 490 mg/kg for Pb, and from 170 to 1450

mg/kg for Zn. The lowest metal pollution is found in the distal parts of floodplain sections with low flooding frequencies,

where average sedimentation rates have been less than about 5 mm/a. The largest metal accumulations occur in low-lying

floodplain sections where average sedimentation rates have been more than 10 mm/a.

Keywords: floodplain sedimentation, heavy-metal contamination

ses that have caused the present-day distribution of the

metals in the floodplain soils.

The present contribution deals with the spatial

variability of metal pollution of floodplain soils, which

is controlled primarily by deposition of contaminated

overbank sediments during flood events. Spatial vari-

ations in floodplain sedimentation rates and metal de-

position during flood events were determined using

measurements of the deposition of contaminated

overbank sediments along the lower Rhine and

Meuse reaches during a high-magnitude flood. This

process alone is, however, insufficient to explain the

variability in pollution depth. The metal concentra-

tions in the river sediment have greatly varied during

the twentieth century, resulting in corresponding vari-

ations in the metal concentrations in vertical direction

within the floodplain soils.These temporal changes in

metal contamination of the Rhine were reconstructed

from dated sediment cores from small ponds within

the floodplain. The resulting spatial variability of the

metal pollution of the floodplains is discussed by

means of a series of metal profiles in floodplain soils

from various sites along the distributaries of the lower

Rhine reaches.

Sample collection

Deposition of sediment and heavy metals

Deposition of Cd, Cu, Pb and Zn on the floodplains

was measured during the high-magnitude flood of De-

cember 1993. Sediments deposited during the flood

were collected using about 800 sediment traps, consist-

ing of pliable mats of artificial grass of 50 cm x 50 cm

size, placed on seven floodplain sections along the Waal

river and four sections along the Meuse river (Fig. 1).

412 Geologie en Mijnbouw / Netherlands Journal of Geosciences 79(4) 2000

Rhine

Germany

Belgium

The NetherlandsThe Netherlands

Germany

Belgium

Rhine

Waal:

= Klompenwaard

= Winssen

= Slijk-Ewijk

= Willemspolder

= Variksche Plaat

= Brakelsche

Benedenwaarden

Nederrijn - Lek:

IJssel:

= Kersbergen

= Amerongen

= Huissen

= Bronkhorst

= Vorchten

Keent

Klompenwaard

Bemmelsche Waard

Slijk-Ewijk

Nijmegen

Willemspolder

Stiftsche Uiterwaard

Variksche Plaat

Alem

Hoenzadriel

Bern

Tie

Brakelsche Benedenwaarden

0 10 km

LobithLobith

KB WP

SE HU

❋

Waal

Lek

❋

Rhine

❋

Arnhem

Nijmegen

Zwolle

Nederrijn

Meuse

IJssel

Lobith

010 km

embanked floodplain

transect of floodplain soil profiles

dike-breach pond; ZG = Zwarte Gat; WU = Wamel; TU = Tuil

❋

floodplain section where sediment-trap measurements were carried out during the December 1993 flood

United

Kingdom

Fig. 1. Location of the sample areas.

a: floodplain soil profiles.

b: sediment traps during the December 1993 flood, and dike-breach ponds.

The sample areas comprised floodplain sections

with natural topography (Klompenwaard, Varikse

Plaat, Brakel, Hoenzadriel, parts of Slijk-Ewijk and

Stiftsche Uiterwaard), levelled sections (Bemmelsche

Waard, Willemspolder, Keent, Alem, Bern), flood-

plain sections bordered by a natural levee (Klompen-

waard, Slijk-Ewijk, Keent, Bern) and sections bor-

dered by a minor dike that protects the floodplain

from inundation during low-magnitude floods (Bem-

melsche Waard, Willemspolder, Stiftsche Uiterwaard).

The sampling was performed along transects perpen-

dicular to the main channel.

Lake sediment from ponds

Sediment cores were obtained from three small ponds

located within the embanked floodplain area: Zwarte

Gat (ZG), Wamel (WU) and Tuil (TU) (Fig. 1).

These ponds are the scars of historic dike breaches

and for more than two centuries they have been func-

tioning as natural sediment traps in which 3 to 6 m

thick bodies of dark-grey overbank sediments have

accumulated (Middelkoop, 1997).

The cores were taken from a raft using a modified

Livingstone piston corer. In the laboratory, the sedi-

ment cores were cut along the longer axis into two

halves. At about 10-cm intervals, samples of a con-

stant volume were taken to determine the compaction

of the sediment. At intervals of about 10-20 cm, sam-

ples were taken for the analysis of the clay and organ-

ic-matter contents and metal concentrations. To

achieve an age/depth framework, 5-11 samples were

taken from each core for 210Pb-dating.

Floodplain soils

Samples of floodplain soils were collected from 29

vertical profiles from 11 floodplain sections along the

lower Rhine distributaries (Nederrijn/Lek, IJssel and

Waal) (Fig. 1a). These sites comprise different flood-

ing frequencies, local elevations and distance to the

main river channel (Table 1). The samples from the

upper 50 cm were taken using a gouge; samples at

greater depth were collected using a hand auger.The

Geologie en Mijnbouw / Netherlands Journal of Geosciences 79(4) 2000 413

Table 1. Site characteristics of the locations of the investigated soil profiles.

average flood average flood distance from site description

freqency (n/100 a) time (days/a) levee (m)

Nederrijn – Lek

1 AM1 35 1.5 150 natural levee

2 AM2 35 1.5 550 central part of floodplain

3 AM3 35 1.5 790 central part of floodplain

4 KB2 25 1.0 60 central part of floodplain, behind minor dike

5 KB3 25 1.0 170 central part of floodplain

6 HU1 55 3.8 170 natural levee

7 HU2 45 2.5 400 behind minor dike

8 HU3 45 2.5 430 behind minor dike, depression

IJssel

9 BH1 90 6 330 central part of floodplain

10 BH2 90 6 20 natural levee

11 VO1 110 8 50 natural levee

12 VO2 110 8 190 central part of floodplain

Waal

13 KL1 120 8 260 central part of low floodplain

14 KL2 120 8 50 part of low floodplain

– KL3 120 8 120 central part of floodplain, depression

15 WI4 40 1.8 220 behind minor dike, elevated area

16 WI5 45 2.5 20 natural levee

17 SE1 100 7 50 behind natural levee, elevated area

18 SE2 100 7 180 behind natural levee, depression

19 WP1 44 2.2 670 central part of floodplain, behind minor dike

20 WP2 44 2.2 250 central part of floodplain, behind minor dike

21 WP3 44 2.2 360 depression, behind minor dike

22 WP4 44 2.2 610 depression, behind minor dike

23 VP1 80 5.4 380 central part of low floodplain

24 BR1 62 4.4 10 depression behind natural levee

25 BR2 11 <1 480 central part of floodplain, behind minor dike

26 BR3 40 1.8 210 central part of floodplain, behind minor dike

27 BR4 11 <1 360 central part of floodplain, behind minor dike

28 BR5 11 <1 480 central part of floodplain, behind minor dike

sampling intervals were 2.5-5 cm in the upper 20 cm

of the profiles and 5-10 cm at greater depths.

Analytical methods

The contents of clay (particles < 2 µm), organic mat-

ter, and heavy metals were determined for each sam-

ple from the sediment traps, the dike-breach cores,

and the floodplain soil profiles.The percentage of clay

in the samples was determined using the standard

pipette method after NEN 5753 (NNI, 1992a). The

organic matter contents were determined by loss on

ignition after NEN 5754 (NNI, 1992b).The concen-

trations of Cd, Cu, Pb and Zn in the samples were

measured by ICP-AES after extraction using 2M

HNO3in a microwave. Extensive testing has demon-

strated that the metal concentrations obtained using

this extraction method differed only by 3% (except

Cd: 13%) from extraction results obtained using aqua

regia (NEN 6465: NNI, 1981; Middelkoop, 1997).

The sediment collected by the traps was oven dried at

70°C and subsequently weighed. The metal deposi-

tion was calculated by taking the product of the

amount of deposited sediment measured using the

traps and the metal concentrations in the sediment.

The metal concentrations in the samples from the

dike-breach cores and the soil profiles were converted

to standard concentrations in a reference sediment

type containing 40% of clay and 8% of organic mat-

ter. This was done using linear relationships between

metal concentrations and the contents of clay and or-

ganic matter of the floodplain sediments deposited in

1993. The dry bulk densities of the sediment cores

from the ponds were determined by measuring the

weights of the samples with constant volume after

oven drying at 70°C. These densities were used to

correct the depth scales of the profiles for com-

paction. The 210Pb analyses were carried out at the

Netherlands Institute for Sea Research (NIOZ), Tex-

el, the Netherlands.The 210Pb activity in the samples

was measured by alpha-spectrometry emitted by its

granddaughter 210Po (Heijnis et al., 1987).

Sedimentation rates were obtained according to the

so-called constant sedimentation (CR) model (Do-

minik et al., 1981) after plotting the sample excess

210Pb activities (in Bq/kg) originating from deposition

on a logarithmic scale against the compaction-cor-

rected depth. Errors in the measurements of 210Pb

activities were about 5 Bq/kg and the precision of the

estimated sedimentation rates was about 0.2 cm/a.

Middelkoop (1997) and Middelkoop & Asselman

(1998) give a more complete description of the meth-

ods.

Deposition of heavy metals during the flood of

December 1993

Particulate-bound metal concentrations in the river

Figure 2 shows the suspended-sediment concentra-

tions and the concentrations of particulate-bound

Cu, Pb and Zn in the Rhine and Meuse sediment

during the flood of December 1993. Average metal

concentrations in Rhine sediment measured during

the flood at Lobith (German/Dutch border) were

about 1.6 mg/kg for Cd, 64 mg/kg for Cu, 90 mg/kg

for Pb and 400 mg/kg for Zn. These concentrations

were lower than during the preceding period of low

discharge. During the flood, the Meuse sediment at

the Belgian/Dutch border contained 5-7 mg/kg Cd,

50-60 mg/kg Cu, 220 mg/kg Pb and 670 mg/kg Zn.

414 Geologie en Mijnbouw / Netherlands Journal of Geosciences 79(4) 2000

Table 2. Contents of clay, organic matter and heavy metals in sediments deposited on the Waal and Meuse floodplains during the flood of De-

cember 1993.

minimum maximum average median standard deviation

Waal (n = 41)

< 2 µm (mass %) 5.7 47.9 27.3 30.3 10.5

org. matter (mass %) 1.5 12.2 6.5 6.8 2.2

Cd (mg/kg) 0.3 3.0 1.9 2.1 0.5

Cu (mg/kg) 10.4 81.5 57.1 63.8 17.0

Pb (mg/kg) 26.4 126 89.1 94.6 22.2

Zn (mg/kg) 152 462 344 379 76.6

Meuse (n = 17)

< 2 µm (mass %) 4.6 40.0 24.8 28.5 12.7

org. matter (mass %) 0.7 11.2 6.4 6.9 3.3

Cd (mg/kg) 0.5 12.1 7.1 7.5 3.9

Cu (mg/kg) 3.0 110 62.5 71.0 33.7

Pb (mg/kg) 17.2 344 188 202 99.5

Zn (mg/kg) 93.7 1177 684 707 339

Concentrations of Cd,Cu, Pb and Zn in overbank deposits

Table 2 summarises the heavy-metal concentrations

in the samples from the sediment deposited during

the flood of December 1993 on the floodplains along

the Waal and Meuse. A wide range of concentrations

was found for all metals. The average and standard

deviations of metal concentrations in the Waal sam-

ples are lower than those found in the samples from

the Meuse.

Figures 3-5 show typical examples of the composi-

tion and amounts of overbank sediments deposited

during the flood of December 1993. Large differ-

ences in metal concentrations were found along the

transects of the Klompenwaard (Fig. 3b), Brakel,

Keent and Bern (Fig. 5b) sections, whereas the con-

centrations in the Willemspolder and within the cen-

tral parts of the Bemmelsche Waard (Fig. 4b), the

Slijk-Ewijk section and Varikse Plaat varied less. The

strongest gradients in metal concentrations occurred

in floodplain sections with a pronounced natural

levee (e.g., Klompenwaard and Bern; Figs. 3b and

5b). The changes in metal concentrations coincide

with those of the clay and organic matter contents.

Metal concentrations in the sandy material deposited

directly behind the natural levee are relatively low, but

the contents of clay, organic matter and heavy metals

increase with distance from the natural levee. In

floodplain sections bordered by a minor dike, only the

finer sediment fraction with high metal concentra-

tions was deposited.

Variations in metal concentrations within a flood-

plain section are mainly determined by the clay and

organic matter contents of the sediment. Figure 6

Geologie en Mijnbouw / Netherlands Journal of Geosciences 79(4) 2000 415

December 1993 January 1994

1 5 10 15 20 25 30 4 9 14 19 24 29

suspended sediment concentration (mg/l) discharge (m3/s)

100

150

200

250

300

350

400

450

2000

4000

6000

8000

10000

12000

100

200

300

400

500

600

metal

concentration

(mg/kg)

700

suspended sediment concentration (mg/l)

100

200

400

300

500

600

500

1000

1500

2000

3000

3500

2500

200

400

800

600

1000

metal

concentration

(mg/kg)

suspended

sediment

concentration

discharge

discharge (m3/s)

Meuse at Borgharen

no data

suspended sediment concentration

discharge

Rhine at Lobith

Fig. 2. Discharge, suspended-sediment

concentration and metal concentrations

in the sediment of the Rhine at Lobith

(top) and the Meuse at Borgharen (bot-

tom) during the flood of December

1993. Sedimentation on the embanked

floodplains of both rivers occurred be-

tween December 21, 1993 and January

13, 1994. Source: Institute for Inland

Water Management and Waste Water

Treatment (RIZA) (unpublished data)

and Kos (1994).

shows by the example of Zn the correspondence be-

tween metal concentrations and the percentages of

clay and organic matter in the sediment. Zn was cho-

sen for illustration because its concentrations are

highest. The increments of linear regressions of the

metal concentrations on the clay and organic matter

contents (Table 3) can be considered as a measure of

the degree of pollution of the river in the upstream

basin (Middelkoop, 1997). The larger increments of

metal concentrations for the Meuse sediments show

that these are more contaminated than the Rhine se-

diments.

Metal deposition

Figures 3c, 4c and 5c demonstrate that the spatial

variation of metal deposition is dominated by the

amount of sediment deposition. The decrease in total

sediment deposition with increasing distance from

the main channel is much larger than the increase in

metal concentrations associated with changes in clay

and organic-matter contents.

High-maximum metal depositions occurred within

50 m distance from the top of the natural levee of the

Klompenwaard (0.03 g/m2Cd, 0.7 g/m2Cu, 1.1 g/m2

Pb and 5.0 g/m2Zn) and Bern (0.06 g/m2Cd, 0.75

416 Geologie en Mijnbouw / Netherlands Journal of Geosciences 79(4) 2000

{

●

0 250 m

WAAL

13.5 m + Dutch Ordnance Datum (O.D.)

13.0

12.5

12.0

11.5

11.0

10.5

10.0

9.5

WAAL

150

100

250

200

350

300

Cu, Pb, Zn

1.5

1.0

0.5

heavy metal content (mg/kg)% clay and organic matter

100 250200150500 350300

400

3.0

2.5

2.0

100 250200150500 350300

distance from natural levee (m)

sediment deposition (kg/m2)

0.01

0.02

0.03

metal deposition (g/m2)

AB CDE

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

transect of sediment traps

sample location

a. Floodplain elevation and sampling cross section

b. Cross section: contents of clay, organic matter and heavy metals in overbank sed

●

AB C D E

% clay

% organic matter

sediment deposition

heavy metal sample

sediment trap

Cu, Pb, ZnCd

distance from natural levee (m)

Fig. 3. Klompenwaard: sediment com-

position and heavy-metal deposition.

a: floodplain elevation and sampling

cross-section.

b: cross-section: contents of clay, organ-

ic matter and heavy metals in overbank

sediments.

c: cross-section: deposition of sediment

and heavy metals.

g/m2 Cu, 1.4 g/m2Pb and 6.0 g/m2 Zn) sections

(Figs. 3c and 5c). At larger distances from the river,

the metal deposition decreased to about one third

(along the Waal) or one fifth (along the Meuse) of

these maximum values. Large metal deposition due to

high sediment accumulation behind the natural levee

or minor dike was also found in the sections of

Slijk-Ewijk, Willemspolder, Stiftsche Uiterwaard and

Varikse Plaat. Within the Bemmelsche Waard (Fig.

4c), the highest metal deposition occurred in the

low-lying area around sample point C, where large

amounts of sediment with high contents of clay and

organic material were deposited.

Total metal deposition on the Waal floodplain during the

flood of December 1993

The total amount of silt and clay deposited on the

Waal floodplains during the flood was estimated from

the sediment-trap measurements to be 240•106kg.

The total sand deposition, determined in a separate

study (Van Manen et al., 1994), was about 350•106

Geologie en Mijnbouw / Netherlands Journal of Geosciences 79(4) 2000 417

7.125

6.25

5.375

4.5

3.625

2.75

1.875

WAAL

{

●

●●

●

●●

●

●●

●

●●

●

0 500 m

150

100

250

200

350

300

Cu, Pb, Zn

1.5

1.0

0.5

heavy-metal content (mg/kg)

% clay and organic matter

600 0200

400

3.0

2.5

2.0

distance from minor dike (m)

8004000 200 400

450

500

sediment deposition (kg/m2)

600 0200 8004000 200 400

Cu, Pb, ZnCd

0.01

0.02

0.03

heavy metal deposition (g/m2)

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

sample location

transect of sediment traps

>11.5 m +O.D.

11.5

11.0

10.5

10.0

9.5

9.0

8.5

<8.0

●

ABC D EF

ABC D E

F5.0

WAAL

distance from minor dike (m)

% clay

% organic matter

sediment deposition

heavy-metal sample

sediment trap

Fig. 4. Bemmelsche Waard: sediment

composition and heavy-metal deposi-

tion.

a: floodplain elevation and sampling

cross-section.

b: cross-section: contents of clay, organ-

ic matter and heavy metals in overbank

sediments.

c: cross-section: deposition of sediment

and heavy metals.

kg. It was estimated by multiplying these amounts by

the average metal concentrations measured in each of

the sediment types that the Waal deposited during the

flood about 750 kg Cd, 230 kg Cu, 360 kg Pb and

143,000 kg Zn on its floodplain (Table 4). This is

about 17-25% of the total particulate-bound metal

load transported during the flood by the Waal, and of

the order of 5-10% of the total particulate-bound met-

al load transported in 1993 into the Netherlands by

the Rhine. Metal accumulation of the embanked flood-

plains is thus an important factor when determining a

heavy-metal balance for the Rhine-Meuse delta.

Interpretation

Sedimentation processes govern the spatial patterns

of metal deposition during a flood. Variations in the

percentages of clay and organic matter in the deposit-

418 Geologie en Mijnbouw / Netherlands Journal of Geosciences 79(4) 2000

{

●

0 250 m

100

200

300

Cu, Pb, Zn

heavy-metal content (mg/kg)% clay and organic matter

100 250200150500

400

2.0

100 250200150500

MEUSEMEUSE

3.0 m +O.D.

2.8

2.6

2.4

2.2

2.0

1.8

1.6

1.4

300

distance from levee (m)

300

500

600

700

800

4.0

6.0

900

1000

8.0

10.0

distance from levee (m)

0.01

0.02

0.03

0.04

Cu, Pb, ZnCd

sediment deposition (kg/m2)

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

5.5

6.0

0.05

transect of sediment traps

sample location

BC D E FA

3 m +OD

metal deposition (g/m2)

% clay

% organic matter

sediment deposition

heavy-metal sample

sediment trap

Fig. 5. Bern: sediment composition and

heavy-metal deposition.

a: floodplain elevation and sampling

cross-section.

b: cross-section: contents of clay, organ-

ic matter and heavy metals in overbank

sediments.

c: cross-section: deposition of sediment

and heavy metals.

ed sediment determine the differences in metal con-

centrations within a floodplain section. The amount

of sediment deposition generally controls the deposi-

tion of metals. As a consequence, the spatial distribu-

tion of metal concentrations in sediments is often dif-

ferent from that of metal deposition. Spatial varia-

tions in sediment deposition are largely determined

by the flow patterns during overbank flooding (Mid-

delkoop & Van der Perk, 1998). In those sections

where a natural levee borders the main channel, sedi-

ment deposition decreases more or less exponentially

with increasing distance from the main channel.

Many floodplain sections, however, have a non-uni-

form topography with small embankments, former

secondary channels, built-up areas and other obsta-

cles to the water flow. Middelkoop & Asselman

(1998) and Middelkoop & Van der Perk (1998) have

demonstrated that this can result in complex patterns

of the deposition of sediment and thus of heavy met-

als. Over a period of many years, metal accumulation

will be highest in low-lying areas close to the main

channel. Large amounts of sediments are deposited

there already during low flow stages, when the metal

Geologie en Mijnbouw / Netherlands Journal of Geosciences 79(4) 2000 419

Table 3. Coefficients of regression between metal concentrations

on the percentages of clay and organic matter. Old uncontaminated

river sediments (deposited before 1750) and sediment deposited

during the flood of December 1993 on the Waal en Meuse flood-

plains. Metal concentration = b0+ b1*%clay + b2*%OM.

b0b1b2R2

zinc

uncontaminated 46.1 1.00 1.00 0.66

Waal 163.0 3.35 13.87 0.71

Meuse 67.6 6.75 76.20 0.95

lead

uncontaminated 22.0 0.50 0.25 0.48

Waal 38.2 0.75 4.71 0.73

Meuse 4.18 2.45 20.50 0.95

copper

uncontaminated 5.5 0.46 0.99 0.66

Waal 15.0 0.82 3.05 0.88

Meuse 0.4 0.78 7.12 0.94

cadmium

uncontaminated ––––

Waal 0.64 0.02 0.11 0.69

Meuse –0.94 0.05 0.49 0.65

200

400

600

800

1000

1200

Zn content (mg/kg)

100

200

300

400

500

Zn content (mg/kg)

0 5 10 15

% organic matter

0 1020304050

% <2µm

01020304050

% <2µm

051015

% organic matter

Fig. 6. Relationship between Zn concentration and percentages of clay and organic matter.

a and b:Waal floodplains.

c and d: Meuse floodplains.

concentrations in the river sediment are still higher

than during periods of large discharge. In addition,

large metal accumulations are expected in depres-

sions and residual channels that often inundate and

where fine humic sediments with high metal concen-

trations settle.

The results of the research into the deposition after

the major flood of 1993 demonstrate the causes of the

420 Geologie en Mijnbouw / Netherlands Journal of Geosciences 79(4) 2000

Table 4. Metal deposition on the entire embanked floodplain of the Waal during the flood of December 1993.

sediment Cd Cu Pb Zn

metal deposition in sand sheets (bed load)

(a) sand deposition 1(106kg) 350

(b) average metal concentration sediment (mg/kg) 0.7 20 40 150

metal deposition by suspended sediment

(d) clay and silt deposition (106kg) 240

(e) average metal concentration sediment (mg/kg) 2.1 65 90 375

(f) metal deposition (d•e) (103kg) 0.50 16 22 90

total metal deposition (c + f) (103kg) 0.75 23 36 143

metals transported by suspended sediment

* during the flood of 1993

(g) total suspended load Rhine at Lobith 2(106kg) 1,950

(h) average conc. Rhine sediment at Lobith 2(mg/kg) 1.6 65 88 395

(i) total metal load Rhine at Lobith 2(g•h) (103kg) 3.1 127 172 770

(j) total metal load Waal 3(2/3• i) (103kg) 2.1 85 115 516

* over the entire year 1993

(k) total metal load Rhine at Lobith 2(103kg) 6.4 240 335 1560

(l) total metal load Waal 1993 3(2/3•k) (103kg) 4.3 160 223 1040

metals deposited by suspended sediment

proportion of total particulate-bound metal load

of the Waal during the flood (f/j•100%) (%) 24 19 19 17

proportion of total particulate-bound metal load

of the Rhine during the flood (f/i•100%) (%) 16 13 13 12

proportion of total particulate-bound metal load

of the Rhine at Lobith in 1993 (f/k•100%) (%) 8 7 7 6

1 Van Manen et al. (1994).

2 Rijkswaterstaat (1994).

3 the Waal discharges 2/3 of the Rhine water.

trend

0100 200

0100 200 300 400

0500 1000 1500

metal content in the sediment (mg/kg)

1970

1960

1950

1940

1930

1920

1910

1900

1890

1880

1870

1860

1980

1850

year

1990

zinc lead copper

Fig. 7.Trends in the copper, lead and zinc contamination of sediment of the lower Rhine over the past 150 years.

variation in metal pollution between the various loca-

tions within a floodplain, but they cannot explain the

variation in metal concentrations in a vertical direc-

tion within the floodplain profiles. These are caused

by the temporal changes in metal pollution of the

Rhine that have occurred over the past century, as

will be illustrated in the next section.

Pollution history of the lower Rhine

The metal concentrations in the cores from the three

ponds (Zwarte Gat, Wamel and Tuil, Fig. 1b) have

been plotted in Figure 7. The chronological base of

the metal records was obtained using the 210Pb dat-

ings and by comparing the results with the chronolo-

gy of previous records established by Vink & Winkels

(1991),Winkels & Van Diem (1991) and Beurskens et

al. (1994). The upper sections of all profiles show

strongly increased metal concentrations. The profile

of Zwarte Gat shows less detail, because it is based on

fewer samples than the Wamel and Tuil profiles. Metal

concentrations in the Tuil profile are generally higher

than in the Zwarte Gat and Wamel profiles.

The solid lines in Figure 7 show the trends that

were obtained by averaging the measured metal con-

centrations at chronologically corresponding levels in

Geologie en Mijnbouw / Netherlands Journal of Geosciences 79(4) 2000 421

0 500 1000 1500

100

120

140

160

180

200

0 1020304050

% <2µm, % organic matter

heavy-metal content (mg/kg)

0 1020304050

heavy-metal content (mg/kg)

0 1020304050

heavy-metal content (mg/kg)

0 500 1000 1500

100

120

140

160

180

200

0 500 1000 1500

100

120

140

160

180

200

% <2µm, % organic matter % <2µm, % organic matter

depth below surface (cm)

HU-3 HU-2 HU-1

% <2µm

% OM

HU-1

HU-2

HU-3

m + O.D.

W E

Nederrijn

0 250 m

channel deposits

overbank deposits

residual channel deposits

embanked floodplains

overbank deposits

floodbasin deposits

peat

flood basins

water

river dike

other

Legend to the cross-sections

Fig. 8. Cross-section and heavy-metal profiles of the Huissen floodplain section.The metal concentrations have been standardised for differ-

ences in clay and organic-matter contents within the profiles.

the three cores.The scatter around the trend lines may

be attributed to year-to-year variations in metal con-

centrations, as well as changing metal concentrations

during flood events. The solid curves do not reflect

such annual fluctuations, but indicate the main trends

of the upstream metal contamination of the Rhine and

the resulting contamination of sediments, with an esti-

mated temporal resolution of about five years.

The heavy-metal pollution history of the lower

Rhine can be subdivided into several phases:

(1) before about 1860, the metal concentrations in the

Rhine sediment were at low, pre-industrial values;

(2) from 1860 onwards, the river sediment became

gradually contaminated with heavy metals;

(3) at the beginning of the twentieth century, a

strong increase in metal contamination started,

resulting in a maximum in the early 1930s. This

period of high pollution is not comprised in pre-

vious reconstructions of the pollution history of

the Rhine (a.o.,Vink & Winkels, 1991; Beurskens

et al., 1994);

(4) during the period of World War II, the metal pol-

lution temporarily decreased;

(5) after World War II, a second major increase of

pollution started until the early 1960s;

(6) between the early 1970s and about 1985, the

heavy-metal pollution of the sediment was strong-

ly reduced, as a result of the Rhine Action Plan;

(7) during the past decennium, the metal concentra-

tions have decreased more slowly.

Those parts of the floodplain profiles that have been

deposited in the twentieth century are expected to

demonstrate more or less the same fluctuations in

metal deposition as reconstructed from the ponds.

Depending on the sedimentation rates and the sub-

sequent degree of physical soil mixing processes and

chemical re-distribution of the particulate-bound

metals, the pollution may extend over shallow or

large depths, and fluctuations in metal concentra-

tions within the vertical profiles may have become

more or less smoothed (Middelkoop, 1997). This

was analysed by comparing metal profiles in flood-

422 Geologie en Mijnbouw / Netherlands Journal of Geosciences 79(4) 2000

NS AM-3AM-2AM-1

AM-1 AM-2 AM-3

Nederrijn

0 250 m

m + O.D.

0 100 200 300 400 500

100

heavy-metal content (mg/kg)

0 100 200 300 400 500

0 1020304050

0 100 200 300 400 500

0 10203040 50

% <2µm, % organic matter

0 1020304050

heavy-metal content (mg/kg) heavy-metal content (mg/kg)

100

% <2µm, % organic matter % <2µm, % organic matter

depth below surface (cm)

% <2µm

% OM

Fig. 9. Cross-section and heavy-metal profiles of the Amerongen (AM) floodplain section.

plain soils from different parts of the lower Rhine

floodplain.

Heavy-metal profiles in floodplain soils

Typical examples of heavy-metal profiles are shown in

Figs. 8-11. Metal concentrations in the deepest parts

of the profiles are at low, pre-industrial levels. They

increase to a maximum with decreasing depth. Near

the surface of many profiles, the decreased contami-

nation of the Rhine sediment can be recognised.

Maximum metal concentrations in the floodplain

soils vary from 30 to 130 mg/kg for Cu, from 70 to

490 mg/kg for Pb, and from 170 to 1450 mg/kg for

Zn. To summarise the metal-profile characteristics,

five indicators were defined:

Di: depth above which the metal concentrations

strongly increase;

Dx: depth of the maximum metal concentration;

Cx: maximum metal concentration within the pro-

file;

C10: average metal concentration in the upper 10

cm of the profile.This indicator was used in sanita-

tion surveying studies of the embanked floodplains

to determine the extent of the floodplain soil pollu-

tion;

M: total amount of heavy metals present per m2in

the entire profile being the excess over the back-

ground value. M is calculated by multiplying the

depth-averaged excess concentration by the bulk

density (1200-1400 kg/m3).

The values or estimated ranges of the characteristics

of the Zn profiles from the investigated floodplain

sites are listed in Table 5. The ranges found for the

characteristics are very large. For the example of Zn:

Di, = 0.175-2 m; Dx, = 0.05-1.0 m; Cx= 170-1450

mg/kg; C10 = 156-890 mg/kg; M = 22- 1400 g/m2.

The main differences between the profiles may be

summarised as follows:

Both the total and maximum metal contents in the

Waal floodplain are generally higher than those in

the Nederrijn-Lek and IJssel floodplains. The

Geologie en Mijnbouw / Netherlands Journal of Geosciences 79(4) 2000 423

0 250 500 750 1000 0 250 500 750 1000

heavy-metal content (mg/kg)

100

0 1020304050

% <2µm, % organic matter

heavy-metal content (mg/kg)

100

0 1020304050

% <2µm, % organic matter

depth below surface (cm)

VO-2 VO-1

VO-1VO-2

0 250 m

m + O.D.

NW SE

O.D.

IJssel

% <2µm

% OM

Fig. 10. Cross-section and heavy-metal

profiles of the Vorchten (VO) floodplain

section.

smallest amounts of metals are found behind a mi-

nor dike and in the distal parts of the floodplain of

the Nederrijn-Lek (see, for instance, the Ameron-

gen profile: AM3, Fig. 9).

Low floodplain sections along the Waal that are not

bordered by a minor river dike are contaminated

by heavy metals to a great depth, usually more

than 1 m below surface. As a result of the decreas-

ing sediment contamination over the past thirty

years, the metal concentrations in the upper parts

of these profiles are lower than the concentrations

at several dm depth.The Klompenwaard (Fig. 11)

and Varikse Plaat profiles demonstrate this.

The metal pollution is much greater in the natural

levees and at a short distance from the main chan-

nel than in the distal parts or behind a minor dike.

These differences are demonstrated in the profiles

of Huissen (Fig. 8), Amerongen (Fig. 9) and

Vorchten (Fig. 10). In most cases, the metal con-

tent of the profiles decreases with increasing dis-

tance to the main channel (e.g. Amerongen, Fig.

9), but these differences do not show up in the

Bronkhorst and Slijk-Ewijk areas. Metal accumu-

lations in depressions are slightly greater than at

nearby elevated sites.The absence of the pollution

trends within the KL3 profile (Fig. 11) suggests

that the depression from which it was taken has

been partly filled-in by human activities.

The temporal changes in the metal contamination of

the Rhine sediment are reflected in all investigated

profiles. The profiles where the increased metal con-

centrations extend to a great depth and with highest

maximum concentrations (e.g., in Klompenwaard)

have developed at sites where sedimentation rates

have been high.These profiles also display well the re-

cent improvement of the chemical quality of the

Rhine. The metal concentrations in the upper 10 cm

may be 2-3 times as low as the maximum concentra-

424 Geologie en Mijnbouw / Netherlands Journal of Geosciences 79(4) 2000

m + O.D.

KL-2 KL-3 KL-1

KL-1

KL-2 KL-3

0 250 m

Waal

0 500 1000 1500 2000

100

120

140

160

180

200

heavy-metal content (mg/kg)

0 10 20 30 40 50 0 10 20 30 40 50

heavy-metal content (mg/kg)

0 500 1000 1500

heavy-metal content (mg/kg)

100

120

140

160

180

200

0 500 1000 1500

0 1020304050

% <2µm, % organic matter

100

120

140

160

180

200

% <2µm, % organic matter% <2µm, % organic matter

depth below surface (cm)

% <2µm

% OM

Fig. 11. Cross-section and heavy-metal profiles of the Klompenwaard (KL) floodplain section.

tions. The profiles that are only contaminated in the

upper 10-20 cm reflect low sedimentation rates (e.g.,

in the Amerongen profile AM3) where the maximum

metal concentrations are lower. This may be due to

the relatively wide sample depths compared to the

metal fluctuations in the profiles. In addition, the

smoothing effect of soil mixing on the metal concen-

trations is relatively strong when sedimentation rates

are low.

Estimation of sedimentation rates

By assigning the ages of characteristic changes in the

metal pollution to the corresponding depths in the

metal profiles, an estimate of the average sedimenta-

tion rates over the past century was obtained. This

was done on the basis of (1) the depth above which

the concentrations strongly increase (Di), and (2) the

depth of the maximum concentration in the profile

(Dx).

As a result of soil mixing and the sample widths,

the two maxima in the pollution history could not be

distinguished in profiles with a low sedimentation

rate. The depth of maximum concentrations in these

profiles may, therefore, correspond to the 1935-1960

period. The results given in Table 6 indicate that the

sedimentation rates range from less than 1 mm/a for

the floodplains of the IJssel and Nederrijn-Lek rivers

to about 15 mm/a for low lying parts of the Waal

floodplain.

Implications for pollution surveys

Complex patterns of water flow and deposition that

Geologie en Mijnbouw / Netherlands Journal of Geosciences 79(4) 2000 425

Table 5. Zn-profile characteristics of the investigated floodplain profiles.

Di(cm) Dx(cm) Cx(mg/kg) C10 (mg/kg) M (g/m2)

Nederrijn – Lek

1 AM1 25–35 10–12.5 466 421 100

2 AM2 25–30 10–12.5 214 199 32

3 AM3 17.5–20 5–12.5 170 158 23

4 KB2 30–35 15–20 200 187 29

5 KB3 27–32 20–25 169 156 22

6 HU1 35–45 20–30 1094 600 529

7 HU2 17.5–20 7.5–10 499 465 106

8 HU3 25–30 10–12.5 568 520 119

IJssel

9 BH1 25–35 10–12.5 647 530 109

10 BH2 17.5–20 5–10 499 478 61

11 VO1 17.5–20 5–7.5 698 622 90

12 VO2 17.5–20 5–7.5 450 433 53

Waal

13 KL1 100–150 40–50 1295 642 953

14 KL2 100–150 70–100 1352 548 1398

15 WI4 50– 60 10–15 819 651 281

16 WI5 80–150 30–50 1394 683 ?

17 SE1 25–50 7.5–10 1132 890 325

18 SE2 50–80 15–40 1129 790 ?

19 WP1 45–60 15–25 512 461 203

20 WP2 50–70 20–30 640 490 258

21 WP3 50–70 5–15 800 635 334

22 WP4 30–45 20–25 521 472 172

23 VP1 150–200 40–60 1309 518 1081

24 BR1 150–200 45–50 1453 654 1373

25 BR2 50–60 10–15 514 455 175

26 BR3 60–80 15–20 1166 832 455

27 BR4 50–60 10–15 540 420 208

28 BR5 40–60 5–10 491 469 184

Di= depth of lower boundary of increased Zn concentrations; interval indicates uncertainty range.

Dx= depth at which maximum Zn concentration occurs; interval indicates uncertainty range.

Cx= maximum Zn concentration within the profile.

C10 = Zn concentration in the upper 10 cm of the profile.

M = total excess amount of Zn in the profile.

occur during floods (Middelkoop & Asselman, 1998;

Middelkoop & Van der Perk, 1998) may cause that

correlations between metal pollution of a soil versus

factors such as distance to the main channel, local el-

evation or flooding frequency are low when larger ar-

eas are considered. Because of the reduction in pol-

lution over the past decades, inventories of the metal

pollution of the floodplains of the Rhine should not

be based on shallow samples (see Hoogerwerf,

1992).This can be well illustrated when the total (ex-

cess) Zn content, M, in each soil profile is plotted

against the average Zn concentration in the upper 10

cm of the profile, C10 (Fig. 12). Particularly in those

areas where the sedimentation rate is more than

about 3-5 mm per year, the reduction of the river

pollution over the past 25 years has resulted in rela-

tively low metal concentrations in the upper parts of

the soil profiles.

In a few profiles (KL1, WI5, SE1, SE2 and BR3),

metal concentrations in the upper 10 cm seem to be

high when compared with the total metal contents in

the profiles. At these sites, the sedimentation rate has

probably decreased during the period of improving

sediment quality.

Conclusions

Spatial variations in metal concentrations in overbank

sediment deposited during a flood are controlled by

differences in the clay and organic matter contents in

the sediment. The patterns of metal deposition, how-

ever, are generally determined by the amounts of sed-

iment deposited. Sedimentation patterns, in turn, are

controlled by the flow patterns during overbank

flooding. In floodplains where a natural levee borders

the main channel, sediment deposition – and hence

metal deposition – decreases more or less exponen-

tially with the distance from the main channel. In the

many floodplain sections with a non-uniform topog-

raphy, however, the pattern of metal deposition may

be more complex. On the long term, the largest

amounts of metals will accumulate nearby the natural

levee, and in low-lying areas that are frequently inun-

dated and where large amounts of fine sediments are

deposited.

The metal contamination of the sediment that the

Rhine has deposited on its floodplain has varied

greatly over the past hundred years. Maximum pollu-

tion occurred in the 1930s and 1960s, when Cu, Pb

and Zn concentrations were about 6-10 times as high

as background values. A considerable reduction in the

contamination has been achieved in the 1975-1985

period.

The resulting spatial variation in the concentrations

of Cd, Cu, Pb and Zn in the soils of the embanked

floodplains along the lower Rhine distributaries is

426 Geologie en Mijnbouw / Netherlands Journal of Geosciences 79(4) 2000

✖

❇

❇❇

✖

❇✛

❇

✖✖

❇

▲

▼

◆

▲

◆

❇

▼

✛

❇

✖

▲

◆

1500

1000

500

00 200 400 600 800 1000

Zn concentration in the upper 10 cm (C10) (mg/kg)

total excess Zn content (M) (g/m2)

estimated average

sedimentation rate

(mm/year)

Waal: river bank/levee

Waal: central part low floodplain

Rhine/IJssel: levee

Waal: behind levee/dike, depression

Rhine/IJssel: behind levee/dike, depression

Waal: behind levee/dike, elevated

Rhine/IJssel: behind levee/dike, elevated

symbol site type

estimated error bar

site number, see Table 1

Fig. 12. Relationship between total zinc excess in the entire floodplain soil profile (M) and average zinc concentrations in the upper 10 cm

(C10) for several sites along the Rhine distributaries.

Table 6. Estimated ranges of sedimentation rates for different

groups of floodplain sites, based on the depths of onset of pollution

and maximum pollution.

location average

sedimentation

rate (mm/a)

Rhine/IJssel floodplain, natural levee <1 – 2.5

Rhine/IJssel, central part of floodplain 2.5 – 6

Rhine/IJssel floodplain, depression 2.5 – 6

Waal, central part of floodplain 4 – 6

Waal, low floodplain 15 –18

large, both along the river and across overbank de-

posits. Maximum metal concentrations in the lower

Rhine floodplain soils vary from 30 to 130 mg/kg for

Cu, from 70 to 490 mg/kg for Pb, and from 170 to

1400 mg/kg for Zn. The lowest metal pollution is

found in floodplain sections with low flooding fre-

quencies along the Nederrijn-Lek and IJssel, at sever-

al hundreds of metres distance from the main chan-

nel. There, average sedimentation rates over the past

century have been less than about 2 mm/a. The

largest metal accumulations occur along the Waal in

low-lying floodplain sections without a minor dike,

where sedimentation rates have been in the order of

10 mm/a.

Due to the complex patterns of metal deposition

that may occur in floodplains with a non-uniform

topography, correlations between metal pollution in a

soil against variables such as distance to the main

channel, local elevation or flooding frequency will be

low when large floodplain areas are considered. Be-

cause of the variation in pollution over the past centu-

ry, inventories of the metal pollution of the lower

Rhine floodplains should not be based on shallow

samples. Pollution surveys should therefore account

for the water-flow patterns that may occur over the

floodplain when designing a sampling scheme. The

depth over which the pollution extends within the soil

profiles must be determined at several representative

sites within the floodplain, and the overall inventory

should be based on the total excess metal content in

the entire soil profile.

Acknowledgements

The author wishes to thank the Dutch Institute for

Inland Water Management and Waste Water Treat-

ment for providing data on river discharge, suspended

sediment concentrations and heavy metals for the

Rhine and Meuse. Two reviewers are acknowledged

for their useful comments on this paper.

References

Beurskens, J.E.M., Winkels, H.J., De Wolf, J. & Dekker, C.G.C.,

1994. Trends of priority pollutants in the Rhine during the last

fifty years.Water Science & Technology 29:77-85.

Burrough, P.A., Van Rijn, R. & Rikken, M., 1993.Spatial data qual-

ity and error analysis issues: GIS functions and environmental

modelling. Proceedings of the Second International Conference

on Integrating Geographic Information Systems and Environ-

mental Modelling (Breckenridge, Colorado September 1993):

14 pp.

Dominik, J., Mangini, A. & Müller, G., 1981. Determination of re-

cent deposition rates in Lake Constance with radioisotopic

methods. Sedimentology 28: 653-677.

Heijnis, H., Berger,G.W. & Eisma, D., 1987.Accumulation rates of

estuarine sediment in the Dollard area; comparison of 210Pb

and pollen influx methods. Netherlands Journal of Sea Research

21: 295-301.

Hoogerwerf, M.R., 1992. Onderzoek naar de relatie tussen veron-

treiniging van de uiterwaarden en relatieve terreinhoogten. Inter-

nal Report CSO (Den Dolder) CSO.92.030: 61 pp.

Kos, Th.J.M., 1994. Meetverslag zwevend stof metingen in de

Maas tijdens hoogwater December 1993. Internal Document

RIZA (Arnhem) 94.160X: 25 pp.

Leenaers, H., 1989. The dispersal of metal mining wastes in the

catchment of the river Geul (Belgium – the Netherlands).

Netherlands Geographical Studies 102: 1-230.

Middelkoop, H., 1997. Embanked floodplains in the Netherlands.

Geomorphological evolution over various time scales. Nether-

lands Geographical Studies 224: 1-352.

Middelkoop, H. & Asselman, N.E.M., 1998. Spatial variability of

floodplain sedimentation at the event scale in the Rhine-Meuse

delta, the Netherlands. Earth Surface Processes and Landforms

23: 545-559.

Middelkoop, H. & Van der Perk, M., 1998. Modelling spatial pat-

terns of overbank sedimentation on embanked floodplains. Ge-

ografiska Annaler 80A: 95-109.

NNI (Nederlands Normalisatie Instituut), 1981. NEN 6465 – Wa-

ter and air sample preparation for sludge, water-containing

sludge and air dust for the determination of elements by atomic

adsorption spectrometry – Destruction with nitric acid and hy-

drochloric acid. NNI (Delft): 4 pp.

NNI (Nederlands Normalisatie Instituut), 1992a. NEN 5753 –

Bepaling van de korrelgrootteverdeling met behulp van zeef en

pipet (ontwerpnorm). NNI (Delft): 22 pp.

NNI (Nederlands Normalisatie Instituut), 1992b. NEN 5754 –

Bodem-bepaling van het gehalte aan organische stof in grond

volgens de gloeiverliesmethode. NNI (Delft):4 pp.

Rang, M.C., Klein, C.E. & Schouten, C.J., 1987. Mapping of soil

pollution by application of classical geomorphological and pedo-

logical field techniques. In: Gardiner,V. (ed.): International Geo-

morphology Part 1.Wiley & Sons (New York): 1029-1044

Rijkswaterstaat, 1994. Jaarboek monitoring rijkswateren 1993.

Rijkswaterstaat (Den Haag): 163 pp.

Rikken, M.G.J. & Van Rijn, R.P.G., 1993. Soil pollution with heavy

metals. An inquiry into the spatial variation, cost of mapping and

risk-evaluation of copper, cadmium, lead and zinc in the flood-

plains of the Meuse West of Stein, the Netherlands. Internal

M.Sc. thesis, Department of Physical Geography, Utrecht Uni-

versity: 73 pp.

Van Manen, G.R., Onneweer,Y. & Berendsen, H.J.A., 1994. Inven-

tarisatie van de sedimentatie van zand op de oeverwallen van de

Waal tijdens het hoogwater van December 1993 – januari 1994.

Internal Report GEOPRO (Department of Physical Geography,

Utrecht University) 1994.06: 11 pp.

Vink, J.P.M. & Winkels, H.J., 1991. Opbouw en kwaliteit van de

waterbodem van het IJsselmeer. Flevobericht 326 (Rijkswater-

staat, Directie Flevoland, Lelystad): 112 pp.

Winkels, H.J. & Van Diem, A., 1991. Opbouw en kwaliteit van de

waterbodem van het Ketelmeer. Flevobericht 325 (Rijkswater-

staat, Directie Flevoland, Lelystad): 98 pp.

Geologie en Mijnbouw / Netherlands Journal of Geosciences 79(4) 2000 427

Advertising in Netherlands Journal of Geosciences?

Contact: Mrs.Tonny Handgraaf at Retra Publiciteitsservice b.v.

Telephone: +31-(0)23-571 84 80

Fax: +31-(0)23-571 60 02

E-mail: retra@retra-ps.demon.nl

Trophic Transfer of Cd, Cu, Pb, Zn, P and Se in Dutch Storage Water Reservoirs

Article

Full-text available

Jan 2024
ARCH ENVIRON CON TOX

Heavy metals are naturally omnipresent in aquatic systems. Excess amounts of heavy metals can accumulate in organisms of pollution impacted systems and transfer across a food web. Analysing the food web structure and metal contents of the organisms can help unravel the pathways of biomagnification or biodilution and gain insight in trophic linkages. We measured heavy metals and other elements in mussel bank detritus and organisms of the Biesbosch reservoirs (the Netherlands) and linked those to stable isotopic signatures. The heavy metal contents (cadmium, copper, lead, and zinc) were often lowest in benthivorous, omnivorous and piscivorous species (mainly fish); whereas, phosphorus contents were lower in the autotrophs. Mussel bank detritus contained the highest amounts of heavy metals. The heavy metals were negatively correlated with δ¹⁵N values. For selenium no clear trend was observed. Furthermore, there was a negative correlation between fish length and some heavy metals. Based on all 20 analysed elemental contents, similarities between species became apparent, related to niche or habitat. This study confirms that elemental contents of species can differ between feeding guilds and/or species, which can be attributed to metabolic and physiological processes. The organisms in higher trophic levels have adaptations preventing metal accumulation, resulting in lower contents. Within the fish species biodilution occurs, as most metal contents were lowest in bigger fish. Overall, the metals did not seem to biomagnify, but biodilute in the food web. Metal analyses combined with isotopic signatures could thus provide insights in metal transfer and possible trophic linkages within a system.

Exploring Spatial Data Mining Techniques: Predicting Zinc Concentration with Kriging Methods and Geographically Weighted Regression: Spatial data mining methods were used to predict zinc concentration in the Meuse River

Article

Full-text available

Apr 2025

After years of contamination, rivers may get large amounts of heavy metal pollution. Our investigation's goal is to identify the river's hazardous locations. In our study case, we select the zinc-contaminated floodplains of the Meuse River (Zn). Excessive zinc levels may lead to various health issues, including anemia, rashes, vomiting, and cramping in the stomach. However, there isn't a lot of sample data available about the Meuse River's zinc concentration; as a result, it's necessary to generate the missing data in unidentified regions. This study employs universal Kriging in spatial data mining to explore and predict unknown zinc pollutants. The semivariogram is a useful tool for representing the variability pattern of zinc. This captured model will be interpolated using the Kriging method to predict the unknown regions. Regression with geographic weighting makes it possible to see how stimulus-response relationships change over space. We use a variety of semivariograms in our work, such as matern, exponential, and linear models. We also propose Universal Kriging and geographically weighted regression. The experimental findings show that: (i) the matern model, as determined by calculating the minimum error sum of squares, is the best theoretical semivariogram model; and (ii) the universal kriging predictions can be visually demonstrated by projecting the results onto the real map.

Assessing heavy metal contamination in agricultural soils: a new GIS-based Probabilistic Pollution Index (PPI)–case study: Guarda Region, Portugal

Article

Jan 2025

Soil contamination by heavy metals is a global agricultural problem, as these elements can be absorbed by plants and transferred to humans through the food pathway. Current contamination assessment methods rely on in situ sample collection, which restricts the evaluation process due to the number of samples that can be analysed and the associated costs. This study addresses the need for a more efficient and cost-effective approach to identifying areas at risk of heavy metal contamination without the logistical constraints of physical sampling. To meet this challenge, we developed a new Probabilistic Pollution Index (PPI), calculated by integrating GIS tools with an 8-parameter probability-risk matrix to identify agricultural areas potentially contaminated by heavy metals. The factors considered included roads, industrial sites, pH levels, soil organic matter content, terrain slope, soil texture, mining areas, and drainage. Each parameter was classified and reclassified to produce a contamination risk map, categorizing each pixel into five levels of risk. To test the PPI, data analysis, parameter classification, and reclassification were applied in the district of Guarda, Portugal. The PPI map revealed that the central portion of the Guarda municipality, along with specific zones in Celorico da Beira, Sabugal, Mêda, and Pinhel, exhibited a high-probability risk of contamination. Given the agricultural expanse within each municipality, enhanced monitoring of heavy metal levels was recommended for Mêda, Pinhel, and Vila Nova de Foz Côa. This index provides a scalable, cost-effective, and easily replicable tool for identifying potential contamination hotspots and the sources and processes of contamination. Designed as a first-line method for large-scale assessments, it supports governmental decision-making and facilitates targeted risk mitigation strategies through a low-cost approach.

Support Vector Machines With Uncertainty Option and Incremental Sampling for Kriging

Article

Full-text available

Oct 2024
EXPERT SYST

This paper presents a novel approach to pollution assessment by investigating support vector machines (SVM) with an uncertainty option to overcome the limitations of traditional kriging. While kriging is a major tool for geostatistical modelling, allowing to estimate the distribution of contaminants in a region from a small set of samples, it does not allow to extract also the uncertainty map. An uncertainty map is of great interest, as it allows to identify regions of high uncertainty where one should sample in order to reduce high level of uncertainties. In this paper, we propose two variants of the SVM with an uncertainty option, each using a different hinge loss to improve the accuracy and efficiency. These losses allow to estimate different levels of contaminations, as well as uncertainty, such as the three levels: positive, uncertain and negative, namely for pollution estimation: high‐pollution, uncertain and low‐pollution. In addition to the exploration of SVM variants, we propose an innovative active sample selection strategy based on the uncertainty criterion. This strategy is designed to systematically reduce uncertainties in pollution assessment, thus providing adaptability to dynamic environmental changes. An incremental SVM with an uncertainty option is introduced to further optimise the sample selection process. Furthermore, the decision‐making process is refined through the introduction of a novel three‐hinge loss. The corresponding optimization problem and its resolution allow for a more nuanced contamination assessment with multiple levels of estimation, providing a valuable tool for characterising contamination levels with increased granularity. Extensive experiments on synthetic and real data validate the proposed methodology. Synthetic data simulations assess the quality of the approach, while real data from a two‐dimensional porosity measurement demonstrate practical applicability. This research contributes to the advancement of pollution assessment methodologies, providing an adaptable solution for environmental monitoring.

Trace metal accumulation through the environment and wildlife at two derelict lead mines in Wales

Article

Full-text available

Jul 2024

Trace metal pollution is globally widespread, largely resulting from human activities. Due to the persistence and high toxicity of trace metals, these pollutants can have serious effects across ecosystems. However, few studies have directly assessed the presence and impact of trace metal pollution across ecosystems, specifically across multiple environmental sources and animal taxa. This study was designed to assess the environmental health impacts of trace metal pollution by assessing its extent and possible transfer into wildlife in the areas surrounding two abandoned metalliferous mine complexes in Wales in the UK. Water, sediment, and soil at the mine sites and in areas downstream had notably elevated concentrations of Pb, Zn, and, to a lesser extent, Cd and Cu, when compared to nearby control sites. These high trace metal concentrations were mirrored in the body burdens of aquatic invertebrates collected in the contaminated streams both at, and downstream of, the mines. Wood mice collected in contaminated areas appeared to be able to regulate their Zn and Cu tissue concentrations, but, when compared to wood mice from a nearby control site, they had significantly elevated concentrations of Cd and, particularly, Pb, detected in their kidney, liver, and bone samples. The Pb concentrations found in these tissues correlated strongly with local soil concentrations (kidney: ρ = 0.690; liver: ρ = 0.668, bone: ρ = 0.649), and were potentially indicative of Pb toxicity in between 10 % and 82 % of the rodents sampled at the mine sites and in areas downstream. The high trace metal concentrations found in the environment and in common prey species (invertebrates and rodents) indicates that trace metal pollution can have far-reaching, ecosystem-wide health impacts long after the polluting activity has ceased, and far beyond the originating site of the pollution.

Geochemical analysis of overbank fine sediments of the Upper Rhine (Rhinau island, France): Evidence of metal (Pb, Zn, Cu & Sn) enrichments over the last century

Article

Feb 2025

Modern technologies for remediation of bottom sediments contaminated with heavy metals

Article

Full-text available

Nov 2024

Relevance. The need for timely control of the state of water basins and prevention of their catastrophic pollution due to the processes of handling bottom sediments. Aim. To generalize normative and published data devoted to the problem of assessment of bottom sediments pollution and application of modern technologies of remediation of sediments contaminated with heavy metals. Results and conclusions. The article shows that the most significant transformations in the composition and properties of bottom sediments naturally occur in the conditions of the highest load, typical for historically industrialized and urbanized regions. The authors have carried out a review of regulatory support for the assessment and remediation of contaminated bottom sediments. The paper highlights the gaps in the existing regulatory framework applied in Russia related to the lack of requirements for safe concentrations of toxic metals in bottom sediments. Systematization and analysis of applied technologies for bottom sediments remediation are carried out. The main directions of such works include application of physical, physico-chemical, chemical and biological methods of bottom sediments remediation and purification. These methods can be implemented either in situ or ex situ. The advantage of the first option is the lower cost of the works, while the remediation of sediments outside the water body allows achieving a higher degree of purification from heavy metals. The review shown the prospect of using hybrid remediation technologies with the selection of a set of methods that meet the requirements for treatment efficiency and economic feasibility, as well as the possibility of using contaminated sludge for the production of marketable products.

Historical construction, quantitative source identification and risk assessment of heavy metals contamination in sediments from the Pearl River Estuary, South China

Article

May 2024
J ENVIRON MANAGE

Temporal evolution of plastic additive contents over the last decades in two major European rivers (Rhone and Rhine) from sediment cores analyses

Article

Full-text available

Mar 2024

Methods for microplastic analysis in marine matrices, microplastic exposureexperiments and microplastic degradation experiments, as developed, optimized and/or applied in the JPI Oceans Project Andromeda.

Spatial Data Mining for Prediction of Unobserved Zinc Pollutant using Various Kriging Methods

Preprint

Full-text available

Dec 2023

After years of contamination, rivers may get large amounts of heavy metal pollution. Our investigation's goal is to identify the river's hazardous locations. In our study case, we select the zinc-contaminated floodplains of the Meuse River (Zn). Excessive zinc levels may lead to a variety of health issues, including anemia, rashes, vomiting, and cramping in the stomach. However, there isn't a lot of sample data available about the Meuse River's zinc concentration; as a result, it's necessary to generate the missing data in unidentified regions. This study employs universal Kriging in spatial data mining to explore and predict unknown zinc pollutants. The semivariogram is a useful tool for representing the variability pattern of zinc. To predict the unknown regions, this captured model will be interpolated using the Kriging method. Regression with geographic weighting makes it possible to see how stimulus-response relationships change over space. We use a variety of semivariograms in our work, such as matern, exponential, and linear models. We also propose Universal Kriging and geographically weighted regression. The experimental findings show that: (i) the matern model, as determined by calculating the minimum error sum of squares, is the best theoretical semivariogram model; and (ii) the accuracy of the predictions can be visually demonstrated by projecting the results onto the real map.

Embanked floodplains in the Netherlands; geomorphological evolution over various time scales

Article

Hans Middelkoop

Embanked floodplains in the Netherlands; geomorphological evolution over various time scales

Article

Jan 1997

Hans Middelkoop

Trends of Priority Pollutants in the Rhine during the Last Fifty Years

Article

Feb 1994

Sediment cores from Lake Ketelmeer, a sedimentation area of the River Rhine were analyzed for priority pollutants (8 metals, 8 polycyclic aromatic hydrocarbons (PAHs), 7 polychlorinated dibenzo-p-dioxins, 10 polychlorinated dibenzofurans, 6 planar and mono-ortho polychlorinated biphenyls and 11 polychlorinated benzenes) in order to determine trends in the last 50 years. Post-depositional changes in pollutant levels were verified. Metal and PAH levels reflected the unchanged historical inputs. Concentrations were high in the late 1930s, low during the second world war and reached a maximum between 1955 and 1975. As an example, mercury levels changed from 3 to 2 mg/kg and reached 13 mg/kg in these successive periods. Levels in recently deposited sediment are ca. 2 mg/kg again. Some chlorinated aromatics proved to be persistent as well. For example, the highly toxic 2,3,7,8-tetrachlorodibenzo-p-dioxin had concentrations ≤ 10 ng/kg in the early 1940s, a maximum of ca. 400 ng/kg in the 1960s and again very low levels (≤ 10) in recently deposited sediment. On the other hand, post-depositional changes were observed for a few higher chlorinated dioxins and furans and several biphenyls and benzenes. This appeared upon comparison of concentrations in sediment layers collected recently and 20 years ago respectively and was possibly caused by microbial dechlorination. The historical inputs of these pollutants will be underestimated by the sediment core data.

Mapping of soil pollution by application of classical geomorphological techniques and pedological field techniques.

Article

Jan 1987

Irregularly but frequently, parts of the floodplain of the River Meuse are inundated and fresh deposits of black reduced river-bed materials, mixed with yellowish brown silts from locally eroded soils, are left behind. The concentration of zinc, cadmium, copper, lead, cobalt, nickel and chromium were shown to be positively correlated with inundation frequency, with carbonate content and with the thickness of the layer of sediment deposited after the industrial revolution. -from Authors

Modelling Spatial Patterns of Overbank Sedimentation on Embanked Floodplains

Article

Aug 1998

This paper presents a GIS-based mathematical model for the simulation of floodplain sedimentation. The model comprises two components: (1) the existing hydrodynamic WAQUA model that calculates two-dimensional water flow patterns; and (2) the SEDIFLUX model that calculates deposition of sediment based on a simple mass balance concept with a limited number of model parameters. The models were applied to simulate floodplain sediment deposition over river reaches of several kilometres in length. The SEDIFLUX model has been calibrated and validated using interpolated raster maps of sediment deposition observed after the large magnitude December 1993 flood on the embanked floodplain of the lower river Rhine in the Netherlands. The model appeared to be an adequate tool to predict patterns of sediment deposition as the product of the complex interaction among river discharge and sediment concentration, floodplain topography, and the resulting water flow patterns during various discharge levels. In the investigated areas, the resulting annual average sedimentation rates varied between 0.5 mm/year and 4.0 mm/year. The role of the most important mechanisms governing the spatial patterns of overbank deposition, i.e. inundation frequency, sediment load, floodplain topography and its influence on the flow patterns over the floodplain, are discussed.

Spatial Data Quality and Error Analysis Issues: GIS Functions and Environmental Modeling

Article

Spatial Variability of Floodplain Sedimentation at the Event Scale in the Rhine–Meuse Delta, the Netherlands

Article

Jun 1998
EARTH SURF PROC LAND

This article addresses spatial variability of comtemporary floodplain sedimentation at the event scale. Measurements of overbank deposition were carried out using sediment traps on 11 floodplain sections along the rivers Waal and Meuse in The Netherlands during the high-magnitude flood of December 1993. During the flood, sand sheets were locally deposited behind a natural levee. At distances greater than 50 to 100 m from the river channel the deposits consisted mainly of silt- and clay-sized material. Observed patterns of deposition were related to floodplain topography and sediment transporting mechanisms. Though at several sites patterns were observed that suggest transport by turbulent diffusion, convection seems the dominant transporting mechanism, in particular in sections that are bordered by minor embankments. The average deposition of overbank fines ranged between 1·2 and 4·0 kg m−2 along the river Waal, and between 1·0 and 2·0 kg m−2 along the river Meuse. The estimated total accumulation of overbank fines (not including sand sheets) on the entire river Waal floodplain was 0·24 Mton, which is 19 per cent of the total suspended sediment load transported through the river Waal during the flood. © 1998 John Wiley & Sons, Ltd.

Accumulation Rates of Estuarine Sediment in the Dollard Area: Comparison of 210Pb and Pollen Influx Methods

Article

Dec 1987
Neth J Sea Res

Profiles of excess 210Pb activity in sediment cores collected on the muddy tidal flats of the Dollard indicate deposition rates ranging from 0.14 to 0.27 cm·y−1. One profile was also analysed for pollen: the pollen profile in the core provided a record of the sedimentation rate because two independent historical events were found recorded in the profile. The first, a sudden abundance of an Aster-type pollen, reflected, at a depth of 33 cm in the sediment, the large land reclamation of 1862 A.D. The second, the appearance of Zea mays (corn) at a depth of 6 cm, reflected the increase of the area in which corn was cultivated: from 44 ha in 1970 up to 552 ha in 1973. The pollen data gave an estimated sedimentation rate of 0.25 cm·y−1. Excess 210Pb analysis of the same core yielded a sedimentation rate of 0.27 cm·y−1.

Determination of recent deposition rates in Lake Constance with radioisotopic methods

Article

Oct 1981
SEDIMENTOLOGY

Sedimentation rates were determined with the 210Pb method in eight sediment cores from Lake Constance. The rate of deposition in the main basin (Obersee) varies from about 0.06 g cm−2 y−1 in the central part to 0.13 g cm−2 y−1 in the eastern part of the lake and then increases rapidly towards the Rhine delta. In the central lake area the rate of deposition has been approximately constant since 1900, and dating with the 210Pb method is in good agreement with sedimentological observations. In the Konstanzer Trichter area, the deposition rate has been increasing since about 1955 as a result of eutrophication and subsequent high carbonate production. Dating with 137Cs is fairly accurate for sediments deposited at a high rate, but is questionable for slowly accumulating ones. A positive correlation of 210Pb fluxes and sedimentation rates indicates that 210Pb flux into sediments follows the distribution pattern of solids. 210Pb profiles in four sediment cores interpreted in terms of a constant flux model display synchronous fluctuations of the sedimentation rate; however, their relation to long-range particulate input variations remains to be proved. Sedimentation rates determined with the 210Pb method were used to calculate recent nutrient and heavy metal fluxes. Anthropogenic fluxes of Zn and Pb are in the same range of magnitude as in other polluted areas in Europe and America.

The Dispersal of Metal Mining Wastes in the Catchment of the River Geul (Belgium – The Netherlands)

Article

Jan 1989

H. Leenaers

This thesis is a compilation of papers previously presented at scientific meetings and articles published in or submitted to international journals. The study area covers the catchment of the transboundary River Geul which flows from northeastern Belgium into the Netherlands province of Limburg where it eventually flows into the Meuse. The fluvial dispersal of mining wastes has caused enhanced concentrations of lead, zinc and cadmium in aquatic sediments and floodplain soils in both countries. As the Geul continues to migrate across its floodplain, contaminated sediments are reworked and ultimately discharged into the River Meuse, which is an important source of drinking water in the Netherlands. The objective of this study is to investigate quantitatively the dispersal of metal mining wastes in this catchment. Eventually, the data collected and knowledge obtained may be used to compile mass balances of sediments and heavy metals in the alluvial area. -from Author

Heavy-Metal Pollution of the River Rhine and Meuse Floodplains in the Netherlands

Abstract and Figures

Recommended publications

Reconstructing floodplain sedimentation rates from heavy metal profiles by inverse modelling

Floodplain Sedimentation - Methods, Patterns, and Processes: A Review with Examples from the Lower R...

The impact of climate change on the sedimentation rates on the embanked floodplains in the Netherlan...

Modelling Spatial Patterns of Overbank Sedimentation on Embanked Floodplains