16
Global and Planetary Change, 8 (1993) 203-218 203 Elsevier Science Publishers B.V., Amsterdam Three-dimensional modelling of Quaternary fluvial dynamics in a climo-tectonic dependent system. A case study of the Maas record (Maastricht, The Netherlands) A. Veldkamp a and M.W. van den Berg b Department of Geology and Soil Science, Agricultural University Wageningen, PO Box 37, 6700 AA Wageningen, The Netherlands b Geological Survey of The Netherlands c / o Winand Staring Centre, P.O. Box 125, 6700 AA Wageningen, The Netherlands (Received February 16, 1992; revised and accepted March 4, 1993) ABSTRACT Veidkamp, A. and Van den Berg, M.W., 1993. Three-dimensional modelling of Quaternary fluvial dynamics in a climo-tectonic dependent system. A case study of the Maas record (Maastricht, The Netherlands). Global Planet. Change, 8: 203-218. A three-dimensional model (MATER) simulating Quaternary terrace development along the river Maas at Maastricht was made. Simulation results indicate that Maas terraces are most probably the result of both climatic and tectonic changes in time. Terrace formation is mainly climatically determined but their long-term preservation is closely related to the prevailing tectonic regime. Quaternary uplift of the Maastricht area has known several phases with different "constant" uplift rates. These differences in uplift rate are a main cause of the general valley morphology near Maastricht. Whereas the deep-ocean climatic record is clearly dominated by the tilt cycle between 2.4 and 0.9 Ma, the continental record in the form of the cold-stage floodplain aggradation-dominated periods appears to register primarily modulations of the eccentricity cycle over this episode. During the last 0.9 Ma, when the general uplift rate was of the order of 0.1 m/ka, a "complete" climate-related terrace sequence developed. Every 100 ka climatic cycle is registered in this part of the terrace sequence. Both model and field evidence suggest that Maas terrace sediments are mainly records of the "cold" periods. As such they are valuable continental counterparts of palynological records registering "warm" episodes. Tectonism determines long term terrace preservation and the value of a specific fluvial system as record of past climatic change. Introduction Long continental and marine records are of great value in understanding global changes. Un- derstanding the correlation between the conti- nental and marine records is crucial for gaining a comprehensive view of global changes. One of the physical links between the continent and the ocean is formed by the fluvial system. As such it can help us to make a more detailed analysis of basin margins filled with continental deposits. Intra-plate stress fields relate the tectonic history of the centre of the basin with its margins (Cloe- tingh et al., 1985). Basin margins that experience uplift become dissected by the fluvial systems that are the feeder systems to these basins. So com- bined geomorphological-sedimentological and stratigraphic studies of the fluvial system in the realm of the rising basin flank may provide us with a detailed insight into the temporal evolu- tion of the basin dynamics. Over the last few million years all continents have experienced fundamental climatic changes. In Northwestern Europe not only the climate but also the lithosphere did experience important changes in its rate of deformation. Around the end of the Pliocene to the Early Pleistocene (2.7-2.4 Ma) the rate of deformation accelerated sharply (Van den Berg, 1990; Kooi et al., 1991). The close temporal correspondence in accelera- tion between subsidence curves and uplift curves (Zijerveld et al., 1992) suggests the operation of a 0921-8181/93/$06.00 © 1993 - Elsevier Science Publishers B.V. All rights reserved

Three-dimensional modelling of Quaternary fluvial dynamics in a climo-tectonic dependent system. A case study of the Maas record (Maastricht, The Netherlands)

  • Upload
    m

  • View
    212

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Three-dimensional modelling of Quaternary fluvial dynamics in a climo-tectonic dependent system. A case study of the Maas record (Maastricht, The Netherlands)

Global and Planetary Change, 8 (1993) 203-218 203 Elsevier Science Publishers B.V., Amsterdam

Three-dimensional modelling of Quaternary fluvial dynamics in a climo-tectonic dependent system. A case study of the Maas record (Maastricht, The Netherlands)

A. Veldkamp a and M.W. van den Berg b Department of Geology and Soil Science, Agricultural University Wageningen, PO Box 37, 6700 AA Wageningen, The Netherlands

b Geological Survey of The Netherlands c / o Winand Staring Centre, P.O. Box 125, 6700 AA Wageningen, The Netherlands

(Received February 16, 1992; revised and accepted March 4, 1993)

ABSTRACT

Veidkamp, A. and Van den Berg, M.W., 1993. Three-dimensional modelling of Quaternary fluvial dynamics in a climo-tectonic dependent system. A case study of the Maas record (Maastricht, The Netherlands). Global Planet. Change, 8: 203-218.

A three-dimensional model (MATER) simulating Quaternary terrace development along the river Maas at Maastricht was made. Simulation results indicate that Maas terraces are most probably the result of both climatic and tectonic changes in time. Terrace formation is mainly climatically determined but their long-term preservation is closely related to the prevailing tectonic regime. Quaternary uplift of the Maastricht area has known several phases with different "constant" uplift rates. These differences in uplift rate are a main cause of the general valley morphology near Maastricht.

Whereas the deep-ocean climatic record is clearly dominated by the tilt cycle between 2.4 and 0.9 Ma, the continental record in the form of the cold-stage floodplain aggradation-dominated periods appears to register primarily modulations of the eccentricity cycle over this episode. During the last 0.9 Ma, when the general uplift rate was of the order of 0.1 m/ka , a "complete" climate-related terrace sequence developed. Every 100 ka climatic cycle is registered in this part of the terrace sequence. Both model and field evidence suggest that Maas terrace sediments are mainly records of the "cold" periods. As such they are valuable continental counterparts of palynological records registering "warm" episodes. Tectonism determines long term terrace preservation and the value of a specific fluvial system as record of past climatic change.

Introduction

Long continental and marine records are of great value in understanding global changes. Un- derstanding the correlation between the conti- nental and marine records is crucial for gaining a comprehensive view of global changes. One of the physical links between the continent and the ocean is formed by the fluvial system. As such it can help us to make a more detailed analysis of basin margins filled with continental deposits. Intra-plate stress fields relate the tectonic history of the centre of the basin with its margins (Cloe- tingh et al., 1985). Basin margins that experience uplift become dissected by the fluvial systems that are the feeder systems to these basins. So com-

bined geomorphological-sedimentological and stratigraphic studies of the fluvial system in the realm of the rising basin flank may provide us with a detailed insight into the temporal evolu- tion of the basin dynamics.

Over the last few million years all continents have experienced fundamental climatic changes. In Northwestern Europe not only the climate but also the lithosphere did experience important changes in its rate of deformation. Around the end of the Pliocene to the Early Pleistocene (2.7-2.4 Ma) the rate of deformation accelerated sharply (Van den Berg, 1990; Kooi et al., 1991). The close temporal correspondence in accelera- tion between subsidence curves and uplift curves (Zijerveld et al., 1992) suggests the operation of a

0921-8181/93/$06.00 © 1993 - Elsevier Science Publishers B.V. All rights reserved

Page 2: Three-dimensional modelling of Quaternary fluvial dynamics in a climo-tectonic dependent system. A case study of the Maas record (Maastricht, The Netherlands)

21)4 A, V E L D K A M P A N D M.W. V A N D E N B E R G

large-scale fundamental process in the plate inte- rior. There is growing evidence that plate margins in other parts of the world show dynamics similar to those in Northwestern Europe (Kalvoda, 1988; Ruddimann et al., 1989; Cloetingh et al., 1990; Chen, in: Morner, 1991). Among other records, rivers are likely to have registered both climatic and tectonic changes. The mechanism of terrace formation is a multiple process mainly related to external components of the fluvial system, al- though rivers may make terraces without there necessarily being an external change (Schumm, 1977). From a morphological point of view ter- races are quite simple features but from a sedi- mentological point of view they are very complex. Each terrace unit is usually made up of several stacked, often incomplete sedimentary cycles, representing several alternating depositional and erosional stages. Generally three major driving forces behind terrace formation are used: changes in climate, tectonism and base level. Both climate and tectonism are likely to play a significant role in terrace formation. Base level changes have only importance in fluvial reaches downstream a terrace intersection.

Models that do describe fluvial system dynam- ics for longer time spans are rare because most knowledge of fluvial systems is based on short- term, well controlled experiments (Schumm, 1977; Gregory, 1983; Dawson and Gardiner, 1987).

A conceptual, long-term (2000 ka) macro-scale (100 km 2) fluvial system model, a so-called coarse-scale model (Thornes, 1987), was devel- oped by Veldkamp and Vermeulen (1989). This model simulates the development of fluvial ter- races as the result of strongly simplified external and internal changes in the simulated fluvial sys- tem. An adapted and extended version of this model was applied to the Allier terrace sequence (Veldkamp, 1992). However, the major limitation in the Allier drainage basin was the limitation of the terrace record. In this paper the three-dimen- sional terrace formation model is adapted to the Maas terrace Sequence at Maastricht (Limburg, the Netherlands, Fig. 1). This terrace sequence comprises a very long flight of at least 30 terrace levels from the Middle Miocene onwards (Van den Berg, 1989). Most of the Pleistocene terraces

developed out of reach of base level changes, as we shall describe below. Therefore other external forcing conditions such as the changing climate and or tectonism may be responsible for the recurrence in the Maas floodplain rejuvenation.

The adapted MATER model (MAas TERraces) is a PASCAL program which runs on a VAX 8600. It simulates climate and tectonic dependent ter- race formation three dimensionally. MATER simu- lations will give insight into possible effects of climatic and tectonic changes and internal fluvial dynamics on general alluvial stratigraphy and val- ley morphology of the Maas valley near Maas- tricht.

The Maas terrace flight

The river Maas is a rain-fed river draining an area of approximately 33,000 krn 2 in Western Europe with a mean annual discharge of 260 m 3 s-~ and peak discharges up to 2800 m 3 s-1 at Maastricht. The river basin is superimposed on various morpho-teetonic units and can be classi- fied as a complicated basin as mentioned by Starkel, 1990 (Fig. 1). The study area is situated in the most southeastern part of the Netherlands near Maastricht. Here the river has left the Ar- dennes and has spread its courses over the south- ern flank of the Roer valley graben divided over two successive valleys, the East Maas and the West Maas (Fig. 2).

Maas terraces

East and West Maas valleys together contain about 30 different terrace levels with steps of about ten meters. Quartz content of terrace grav- els decreases with altitude in favour of fresh rock fragments (Van Straaten, 1946). Simultaneously the heavy mineral composition changes from an extremely stable association towards a more un- stable spectrum with a significant proportion of garnet, epidote and hornblende (Zonneveld, 1949) indicating a renewed incision of the uplands. Ter- race gravels are generally coarse and contain abundant poorly rounded gravel together with large drift blocks (boulders larger than 0.5 m in diameter) and blocks of non-cohesive sediments,

Page 3: Three-dimensional modelling of Quaternary fluvial dynamics in a climo-tectonic dependent system. A case study of the Maas record (Maastricht, The Netherlands)

THREE-DIMENSIONAL MODELLING OF Q U A T E R N A R Y FLUVIAL DYNAMICS. T H E MAAS RECORD 205

50Ol

i msterdam

Relief drainage " ~ T o u ~ / / ~ ~ 3 " /

200- 400

i >1000 ~

A

400m-] ~ , , , D / ~ g i t u d a l profile ofthe Maas I I Maastricht I 131

200m~ ~ I l ~ I

i I', / ~ ............. ~ - - ~ - ' - , , M s l

/ Paris Basin ~,~ ......................... -~;--~'~ Ardennes , i Roer Valley _-I_', ~ ......... North Sea u ............................ d;w'~ n ~ w uow Graben I ~ Km B'a~in-,.-,-~ .... ........................................................................................................................................................................................................................... i e

E - M a a s - - . . . . . . . . . . . . . . . . . . - 1 7 0 t

I% !

E ]110- i i i 9 0 -

70-

50-

3 0 -

r ......... e.

• - - - - - " . . . . . . . " F - ' - : . . . . . . . ' -I- . . . . . . . . . . , ~ . . . . . ~ . . . . . ~ , J , ..~ . . . . ~-•----__.~ . . . . . . . W - M a a s

. . . . . . . . . . . : . - . . . . . . . . . : . , ~ -~ = ~ •

. ~ ' , ~ . : , : . . . . . . . ~ . ,_ . ._~ . . . . . . . ;.: . . . . . .

--. . . . . . . . . . . . . . . . . . . . . . . . ' - -~ - . . . . . . . . . . . . . . . . . . . . . . . . . . . . , , - - - ~ ~ : ' : : " - ~ w ' - ' - . - . . . . . . . . . . . . . . . . . . . . . - . . . . . . . . . . . . . . . . . . . . . . , ~. . ' ; . . . . . . . . . . . . . . . ? : : . . ._- -~- . . , , . . . . . . ' . . . . . . . . . .

' " ' " ' : ~ . . . . . . . -'-" ~ r . - - - - - - ' - . - . . . . . . . ., , . . . . . . . . . . . ~ . . . . . . . . . . . .--.

, I " " . . . . . . . . . . . . . . • - " - " ~ - - . : ~ . . , , . . , - . ~ , . . . . . . . . . . . I I I

. . . . . . . . . . . . . . . . . . . . . : . . . . . . .'- . . . . . . . . . . . . . . . . . . . . . . . . ,

. . . . . . . ---.-.. ,. : ~ . . . . . . . . . . . . . . . . . . . . - . . . . . . . . . . . . . . . . . . . . . . . - - - , . . . . . .

e°. Maast r ich t I l a b . ~ I i i 5 10 15

Altitude of - - Continuous terrace deposits terrace surface

I

20 25 30 35 km i I

............. gradlentlnterpreted floodplain . . . . . . . valleysEr°si°n i

Fig. 1. Setting of the study area within the catchment of the river Maas: (a) present-day longitudinal profile of the Maas, the major tectonic domains are indicated; (b) reconstruction of the Pleistocene longitudinal profiles of the West Maas.

Page 4: Three-dimensional modelling of Quaternary fluvial dynamics in a climo-tectonic dependent system. A case study of the Maas record (Maastricht, The Netherlands)

206 A. "vl:SI.DKAMP AND M.W. VAN DEN BER(

Maas terraces fan / / . . . . . (Paleo) Watershed W.-E. Maas [~/ ~r ~r ~r Midi-Aken Thrust ~ / / , , Section composite East - Maas /~ / < Temporal control I/I

PM palaomagnetic boundary .4t~/ C P pollen lk/~l / ~ / s soil ~'1 ~ / TL therrnoluminescence\ _~1-/ ~^~ 14cradi°carbon "~.~-~'~i~. : ~ ~ z / u u m

150 m

West - Maas ~ / _ . . PM.P

~ B ~PM 100 m o

/ m

Terrace sequence-dataset used in simulation

320

250

N W - " "- SE Not to scale

Fig. 2. Schematic cross-section of the Maastricht terrace sequence.

n

¢

QJ

t~

m

~0

-q aJ t~

O~

e~

t~ t~

Page 5: Three-dimensional modelling of Quaternary fluvial dynamics in a climo-tectonic dependent system. A case study of the Maas record (Maastricht, The Netherlands)

'HREE-DIMENSIONAL MODELLING OF Q U A T E R N A R Y FLUVIAL DYNAMICS. T H E MAAS RECORD 207

Pollen stages

ODP 677

s4, o! s,

Maastricht section

o

Mechelen ai d Maas

Eijsden-t.anklaar

Caberg 3

u,I

Rothem 2

Rothem 1

's Gravenvoeren

• Pietersberg 2

Geertruid 3

Geertruid 2

Geertruid 1

" Valkenburg -~--

Sibbe 1~..

apparently transported in frozen state, indicating cold climate fluvial transport (Krook, 1961; Brun- nacker et al., 1982).

The West Maas valley contains about 20 ter- races (Figs. 1 and 2) above the present-day flood- plain. The reconstructed longitudinal profiles are parallel to each other and to the Holocene flood- plain. They represent top levels of cold-stage aggradation in the course of the valley formation. This parallelism indicates that terraces developed out of reach of sea-level fluctuation. Moreover floodplain aggradation took place during sea-level low stands.

First-order geological age control on the West Maas terrace sequence comes from climatic (Gul- lentops, 1954), biostratigraphical (Zagwijn, 1974; Zagwijn and De Jong, 1983) and palaeomagnetic (Bruins, 1981) indications of the sediments, com- bined with 14C (De Jong, 1984) and thermolumi- nescence datings (Huxtable and Aitken, 1985) (Fig. 2). This age control also shows that at least the last three glacial stages are represented in the Maas terrace record as an individual terrace. This frequency suggests an individual terrace level for approximately each 100 ka. Nine terrace levels higher up the flight, the Geertruid 3 level is found. This older level contains in its top-sedi- ments Interglacial pollen (Zagwijn and de Jong, 1983) and the top of the Jaramillo palaeomag- neti¢ event (Bruins, 1981) and was therefore dated around 1 Ma B.P. Now we have left 8 undated terrace levels to devide over 750 ka. Assuming a terrace level for each 100 ka a match of the undated terraces with known cold stages with the ODP 677 record (Shackleton et al., 1990) is possi- ble. This correlation allows us to assign an ap- proximate age to the individual terrace bodies between Geertruid 3 and Caberg 3 levels. Of the nine terraces above the Geertruid 3 level only the Geertruid 2 and Simpelveld 1 levels contain pollen (Zagwijn 1974) and /o r palaeomagnetic bound- aries (Bruins, 1981). These data allowed Van den

Fig. 3. Correlation between the Northwestern European pollen stages (Zagwijn, 1985), the ODP 677 t so ocean record (Shackleton et al., 1991) and the Pleistocene terraces of the Maas.

Page 6: Three-dimensional modelling of Quaternary fluvial dynamics in a climo-tectonic dependent system. A case study of the Maas record (Maastricht, The Netherlands)

2 0 8 A. V E L I ) K A M P A N D M.W. V A N D E N B E R ( i

Berg (1993, in prep.) to match these older West Maas terraces with the ODP 677 record (Fig. 3). The match of the ODP 677 record with the pollen record (Zagwijn, 1985) relies partly on a correlation of Maas and Rhine deposits. This exercise revealed an interesting new positioning of some of the pollen stratigraphic stages within the Cromerian complex.

Changes in the Maas catchment area

At least since the Pliocene, the Ardennes and Vosges massifs have belonged to the source areas of the Maas system. There is ample evidence in the Vosges that glaciers were active during cold stages of the Pleistocene epoch (Autran and Pe- terlongo, 1980). Since the deposition of the Caberg 3 sediments (about 250 ka B.P., Huxtable and Aitken, 1985) the Maas lost its upper reaches in the Vosges to the Rhine system (Moselle) as evidenced by changes in heavy mineral assem- blage (Zonneveld, 1949; Bustament de Santa Cruz, 1976). Another, much smaller part of the headwaters was lost to the Aire-Aisne-Seine system (Fig. 1). These captures did not only cut off more than 10% of the catchment (Bosch, 1992), but the Vosges with their relatively high altitudes also formed a significant source area in generating rain and melt water.

Tectonic control

In the study area, the successive Maas flood- plains show a strong tectonic control. The mapped Maas terrace sequences (Van den Berg 1989; Felder and Bosch, 1989) showed a strong relation in their spatial distributions to the position of folds and overthrusts known from the top of the Carboniferous. This strain pattern, initially gener- ated by the Hercynian deformation phase (Wa- terschoot van der Gracht, 1938), is known in great detail from geophysical records and coal exploration and mapping (Patijn, 1963; Drozd- zewski et al., 1985). Known structural antiforms of Hercynian age are reflected in the limits of palaeo-lateral migration of the river. These linea- ments are parallel to the northern Ardennes thrust (the Midi-Aachen thrust). We interpreted

this topographical expression to reflect incipient crustal buckling, a strain pattern caused by the neotectonic thrust-induced foreland stress (Van den Berg, 1990) (Fig. 4). During its evolution the river valley shifted away from the thrust, crossing several thrust-parallel blocks. This tectonic his- tory is relevant to modelling because realistic simulations will require an unequal uplift for different valley slopes. In the study area, the Maas valley near Maastricht, the western and eastern valley slopes of the West Maas belong to different tectonic units, the eastern valley slope being the most uplifted one.

To gain insight into the rate of uplift of the study area, we reconstructed an uplift curve as a

Maas terrac~ fan Normal and strike-slip faults

. ~ Stress field orientation . . . . . . . . Watershed (crustal buckling lineaments)

Hercynian frontal thrust

,: : A'EB

km ~ 0,, ~,ym

Fig. 4. Structural setting of the Maas terrace fan on the southern flank of the Roer Valley Graben. Section B - B 1 compiled and simplified after Meissner et al. (1983).

Page 7: Three-dimensional modelling of Quaternary fluvial dynamics in a climo-tectonic dependent system. A case study of the Maas record (Maastricht, The Netherlands)

THREE-DIMENSIONAL MODELLING OF QUATERNARY FLUVIAL DYNAMICS. THE MAAS RECORD 209

t ime-al t i tude diagram, using the altitude of the various terrace surfaces as marker points relative to the present floodplain. With respect to the time-axis this marker is placed at the end of a glacial period. The resulting uplift curve for the Pleistocene shows a pattern of fluctuating uplift rates over the Pleistocene, this pattern will be discussed in another paper. For the purpose of this paper we only wanted to know the effect of variations in the rate of uplift within a realistic domain. The values of uplift rates learned from this curve fall within the range of 0.02 and 0.7 m ka-1. For the Pleistocene as a whole we arrived at an average value of 0.06 m ka-1. Although these values differ a factor ten, they still fall within a range found elsewhere in Northwestern Europe (e.g. Zuchiewitcz, 1991). Primary level- lings in the Maastricht area provided values of the order of 0.8 + 0.3 mm a-1 averaged over the last five decades (Groenewoud et al, 1991).

Because the reconstructed record of crustal uplift, as shown in Fig. 5 in detail still contains a number of uncertainties, we generalized model uplift rates by averaging over intervals between terrace levels. In the modelling procedure the following uplift rates were used: 2150-780 ka B.P.: an uplift rate of 0.05 m ka -1 780-250 ka B.P.: an uplift rate of 0.105 m ka -1 250-10 ka B.P.: an uplift rate of 0.02 m ka -1

The exercise discussed in this paper will be restricted to the development of the West Maas because the time control on the sediments of this section is much better than that on the East Maas section.

Model characteristics

General model construction, organization and operation have already been described in previ- ous papers (Veldkamp and Vermeulen, 1989;

Pollen stages

~M.

ODP 677

200(m)

150

100

Time (100 KA) 30 25 20 15 10

Fig. 5. Recons t ruc t i on of the r eg iona l uplif t , ba sed on the a l t i t u d e / a g e pos i t ion of the M a a s terraces .

5

50

Page 8: Three-dimensional modelling of Quaternary fluvial dynamics in a climo-tectonic dependent system. A case study of the Maas record (Maastricht, The Netherlands)

210

TABLE 1

Attribute domains of the entities simulation of MATER

A. TIME

B. ENTITLES 1. Entity: RIVER Att r ibu tes Discharge ln tpu t load

Form Wid th Maxload

Erosion

1 < Time < 2400 (ka) (Time steps of 2 ka)

Domain 1 × 1012 < Discharge < 1.5 × 1013 ( m 3 / k a )

0.6 × 109 < Inpu t - load < 1.4 × l 0 s

( m 3 / k a )

m e a n d e r i n g or b ra ided 0.2 < F lood-p la in-wid th < 19.6 (km) 7.05 × 1(I .5 < Maxload < 4.5 × 10 3

(m~s t)

- 2 . 1 × 10-3 < Eros ion < 4,5 × 10 3 (m3s i)

2. Entity: Landscape At t r ibu te s

Quplilt

Quplift-unequal Relief x , y , z

Valley_ depth Stratigraphy

Domain

0 < Quolift < 1).5 ( m / k a )

0 "~ Quplift-unequal < 0.05 ( m / k a ) 150 < x < 15000 (m) 15(1 < y < 15000 (m)

1 < z < 5 0 0 ( m ) 20 < Val ley-depth < 500 (m/ 0 < s t ra t ig raphy < 1000 (ka)

sed compos i t ion 1 < sed imen t -compos i t ion < = 4( - ) Val ley-width 1).4 < Val ley-width < 19.60 (kin)

Veldkamp, 1992). Only some generalities and adaptions and extensions for the Maas system will be described here. The Model was organized with entities and attributes. An entity is an inde- pendent unit in the object system. The entities in this model are LANDSCAPE and RIVER. An at- tribute is a property, mark or characteristic of an entity, such as discharge for RIVER and valley-de- pth for LANDSCAPE. The attributes have estimated values within, for the Maas system, realistic do- mains (Table 1). The interaction between the two entities are the erosion and sedimentation pro- cesses acting in time steps of 2 ka. These pro- cesses are large-scale analogies of the real pro- cesses. They are artificially constructed processes reacting to changes in the 2 ka discharge/ sediment load equilibrium and they are both de- fined as a function of climate. When the sedi-

A VI~,I.I)KAMP ,XNI) M.W. VAN DEN BERG

ment input load exceeds the sediment transport capacity, which is a function of discharge, the difference is deposited (NOT EROSION = TRUE), and in case the transport capacity exceeds the input load the difference is eroded in the simu- lated system ( E R O S I O N = TRUE). The LANDSCAPE is changed by simple straightforward processes that add grid cells to or remove them from the LANDSCAPE in a way similar to deposition and erosion processes. Before these processes start acting, the boundaries and conditions controlling these processes are determined. For instance, erosion in a meandering river can only occur within the calculated floodplain width which is related to the 2 ka discharge. The changed LAND- SCAPE is stored in a geographical information system (GIS). The model simulates the existence of a fluvial system within a valley and not the river dynamics itself.

Which processes act in the LANDSCAPE is de- termined by decision rules extracted from general literature on hydrology (Schumm, 1977). These decision rules are: iv EROSION and UPLIFI and MEANDER THEN IN(I- SION

IF EROSION and UPLIF' I and B R A I D E D THEN INCI-

SION a n d BANK-EROSION

IF EROSION and NOT-UP1JFT and M E A N D E R THEN

INCISION and BANK-EROSION

IF EROSION and NOT-I,;PLIFT and BRAIDED THEN

BANK-EROSION

1F NOT-EROSION THEN 1)EI'OSIT1ON (disregarding all other conditions) UPLIFT and NOT-UPIJI-:T are system states indicat- ing the impact of tectonic uplift. UPLIFT i s true (NO'r-UPLIVr = FALSE) when the fluvial erosion is not able to compensate for the LANDSCAPE uplift. MEANDER is true when the Maas is meandering, whereas the BRAIDED state is true when the Maas is braiding. Erosion processes are headward mi- grating along the longitudinal profile, whereas sedimentation migrates in a downstream direc- tion. Irregular headward erosion during erosion of a meandering RtVER is caused by changes in the effective floodplain width during the simula- tion. The resulting irregular bank erosion has an effect similar to that of a meandering river in a real system.

Page 9: Three-dimensional modelling of Quaternary fluvial dynamics in a climo-tectonic dependent system. A case study of the Maas record (Maastricht, The Netherlands)

THREE-DIMENSIONAL MODELLING OF QUATERNARY FLUVIAL DYNAMICS. THE MAAS RECORD 211

Model input

Climate

As shown by various environmental records on the two hemispheres, the Quaternary climatic changes are global and synchronous. The astro- nomical parameters included in Milankovic's the- ory satisfactorily describe the waxing and waning of continental ice-caps as evidenced from tuning procedures carried out on both continental (Kuk- la, 1987; Kukla et al., 1988) and long deep-sea records (e.g. Imbrie et al., 1984). Continental ice-caps and related sea-ice fields in the Northern Hemisphere strongly influenced the ocean-conti- nent heat transfer and the jet stream related depression tracks (COH members, 1988; Broecker and Denton, 1990) thus effecting Northwestern European climate.

The relationship between climatic behaviour and fluvial dynamics depends on the rate of cli- matic change and the effect of such changes on discharge distribution and sediment load (Lowe and Walker, 1984).

During glacial episodes the southward expand- ing high pressure areas forced the depression tracks and the vegetation belts towards lower latitudes. The shift of the vegetation belts led to a vegetation type (tundra) in the Maas catchment which was particularly sensitive to high frequency changes in the climatic system and caused an increase in sediment supply to the Maas. The southward shift of the depression tracks led to more continental and drier conditions (Guiot et al., 1989; Ruddiman et al., 1989) causing rela- tively low mean discharges in the rivers in these periods.

Interglacials yielded an opposite picture, an increase in the mean discharge and a comple- mentary decrease in sediment supply to the Maas.

As model simulations require a climatic input with a constant reliability during the simulated time span a simplified Milankovic curve was used as basic climatic input. The mean discharge per 2 ka is simulated as the sum of three sine (SIN) functions with the periodicities of the precession (23 ka), obliquity (41 ka) and eccentricity (96 ka). As the sediment load seems to have a similar

behaviour complementary to the 2 ka discharge, it was simulated as the sum of the cosine (COS) functions of the same three periodicities.

The resultant harmonic functions are not weighted by the 400,000 eccentricity modification. This signal is clearly present before 1 Ma B.P.; also the dominance of the obliquity in this period (Raymo et al., 1989) is not accounted for in our climate-related functions.

Although the climate functions used do not exactly match the climatic curves derived from deep-sea cores and lake sediments, they suffi- ciently describe climatic changes during the Qua- ternary for our long-term modelling purposes. This is because these functions are used in a qualitative way. The finite-state model MATER determines states using the climate-related func- tions.

The effects of the climatic dynamics are thus simulated as changes in discharge and sediment load in the Maas. Except the difference in be- haviour of the fluvial dynamics in glacial and interglacial periods the transition from a glacial to an interglacial environment was also taken into account. As a result of glacier melting in the Maas headwaters (Vosges) an extra sediment flux of 1.5 times a normal flux is supplied to the fluvial system. This extra sediment is only sup- plied to the Maas when the glacial duration is longer than 10 ka. This subjective criterium which can be seen as a threshold between the impact of a glacial and stadial, is indirectly derived from a sediment flux reconstruction in the river Allier (Veldkamp, 1991). This river within the Loire basin (Central France) showed a strong climate control over its sediment fluxes during the Late Quaternary due to glaciers in its headwaters.

I

\

Fig. 6. T e c t o n i c c o m p o n e n t s u sed in MATER s imula t ions .

Page 10: Three-dimensional modelling of Quaternary fluvial dynamics in a climo-tectonic dependent system. A case study of the Maas record (Maastricht, The Netherlands)

212 A VELI)KAMP AND M.W. VAN DEN BERG

Tectonism

Two different components of tectonism have been incorporated in the MATER model (Fig. 6). A component of gradual uplift of the whole simu- lated landscape (OUPLIFT), and an uplift compo-

nent describing the difference in uplift rate of the landscapes on both sides of the Maas (OUPLI~W- UNEOUAL). It was already argued that both sides of the Maas valley at Maastricht belong to two different tectonic units with slightly different movements in time. Based on the slight differ-

Initial landscape 1900 KA

1400 KA 900 KA

400 KA Actual landscape Fig. 7. Three-dimensional valley relief development during simulation, in time steps of 0.5 Ma from 2.4 Ma BP (a) to the

present (f).

Page 11: Three-dimensional modelling of Quaternary fluvial dynamics in a climo-tectonic dependent system. A case study of the Maas record (Maastricht, The Netherlands)

ences in terrace altitudes in the Maas valley we assumed a difference in uplift rate of only 0.0015 m ka - I whereby the eastern valley slope experi- enced the highest uplift rate.

The general uplift reconstruction showed al- ready the changing rates in the Maas basin in time. During the simulation the following con- stant uplift rates were used: from 2.4 to 0.78 Ma B.P. a rate of 0.04 m ka-1, from 0.78 to 0.25 Ma B.P. a rate of 0.105 m ka -1, and during the last 0.25 Ma a rate of 0.02 m ka-1.

Initial relief

Model output

The chosen initial relief (Fig. 7a) has a surface area of 400 km 2 (20 km × 20 km) a maximum altitude of 180 m and a minimum altitude of 30 m. The initial relief consists of a valley with a width of 2000 m and a depth of 5 m only. Terrace stratigraphy displays only age 0, indicating that no fluvial sediments occur in the initial LAND- SCAPE.

/ . . . . . . . . . . . . . . . . . 0 . , , I I . . . .

Fig. 9. A g e o f s u r f a c e s e d i m e n t s in t he e n d rel ief .

The output of the MATER model is a data file (GIS) with LANDSCAPE altitudes and stratigraphy for each time step (2400). With these data, cross- sections, maps, three-dimensional relief graphs, or three-dimensional graphs of the development of one cross section in time can be drawn. It is also possible to make a time-related graph show-

40

Fig. 8. Three-dimensional relief development of one cross section in time.

THREE-DIMENSIONAL MODELLING OF QUATERNARY FLUVIAL DYNAMICS. THE MAAS RECORD 213

!I

ing changes in climate and the accompanying sediments and terraces. (Figs. 7, 8 and 9)

Model tuning

The model was tuned using the mean 2 ka discharge and 2 ka sediment load and their am- plitudes. Two known system conditions were used as references during tuning: the total net volume eroded during the Quaternary, and the occur- rence of alternating erosion and sedimentation. The tectonic inputs were relatively well known and not used for model tuning.

Previous sensitivity testing (Veldkamp and Vermeulen, 1989) of the river terrace formation model showed the model to be sensitive to changes in its input variables: discharge, sediment load, uplift rate and unequal uplift rate for differ- ent valley sides, all in 2 ka time steps, as well as the decision rules described above. Model sensi- tivity could not easily be assessed as certain input variables are interrelated, such as the two uplift components and the 2 ka discharge and sediment load. The decision rules cannot be tested to stan-

Page 12: Three-dimensional modelling of Quaternary fluvial dynamics in a climo-tectonic dependent system. A case study of the Maas record (Maastricht, The Netherlands)

214 A "~ E I . I ) K A M P A N D M.W. VAN D E N B E R G

dard procedures as their validity is based on the literature used. Terrace preservation is especially sensitive to landscape uplift rates from 0.05 to 0.2 m ka -1.

Simulation results

A simulation with plausible realistic simulation results is presented in more detail. This simula- tion had the climatic inputs as described by the model input. The LANDSCAPE development in time during this simulation is visualized three-dimen- sionally for each 500 ka in Fig. 7a-f.

To illustrate the incision and sedimentation dynamics in more detail a cross-section develop- ment in time is shown in Fig. 8, showing relief changes during the 2400 ka of the simulation. This figure clearly shows the alternating incision and sedimentation of the Maas in time, causing different terrace levels.

The role of climatic change is obvious. The simulated Maas system tends to form terraces at most changes from glacial to interglacial although terraces were also formed without a major change in climate. Such intrinsic threshold crossings are not very frequent and cause relatively small ter- races which disappear during the simulation be- cause climatic-induced terraces overrule such events. The general terrace formation mechanism is deposition at the end of a glacial period fol- lowed by an incision in the subsequent inter- glacial period. Therefore it is not surprising that, although the river has known a meandering state many times, the accompanying interglacial sedi- ments are rarely found in the stratigraphical record. This limited occurrence of meandering

sediments is true for both the described simula- tion and the actual Maas terrace sequence. This illustrates that terrace sediments mainly contain a registration of the colder periods and their transi- tion to the warmer periods. Terrace sediments seem therefore complementary to palynological records which are mainly recording the warmer climatic spells.

The general role of different uplift rates be- comes obvious from the valley morphology as displayed in Fig. 7. The higher uplift rate be- tween 780 and 250 ka B.P. caused a stronger incision shaping steeper valley slopes and smaller terraces. In Fig. 9 displaying the age of the sur- face sediments in the end relief, the general decrease in terrace area with age during the simulation is obvious. The slower uplift rate dur- ing the last 250 ka diminished the chances of a terrace to survive the erosional stages.

Another interesting observation can be made from Fig. 10 in which the terrace sediment ages are given on the Quaternary time scale with the simulated derived climate curve, the 2 ka dis- charge. The lower mean discharges represent glacials and the higher interglacials. It can be seen that most sediments are deposited during glacials. Especially during the period with an uplift rate of 0.105 m ka 1 between 780 and 250 ka almost every glacial is registered as a terrace. This illustrates that under certain tectonic and climatic conditions a "complete" terrace record is possible. A similar result was found for a compa- rable model for the Allier terraces (Massif Cen- tral, France). This model demonstrated that an apparently complete terrace sequence can have large time gaps (Veldkamp, 1991, 1992). It is

~ 5 ~ 4 -

¢ j 0 ~ 0 . . 3 "

~ 2 -

t i ~ g

2400 2000 Time ka 0 ,soo ,ooo

Fig. 10. Terrace sediment ages on the Quaternary time-scale with the simulated 2 ka discharge. Horizontal bars indicate terrace bodies, vertical bars indicate sediment sequences separated by erosional phases.

Page 13: Three-dimensional modelling of Quaternary fluvial dynamics in a climo-tectonic dependent system. A case study of the Maas record (Maastricht, The Netherlands)

T H R E E - D I M E N S I O N A L MODELLING OF Q U A T E R N A R Y FLUVIAL DYNAMICS. T H E MAAS RECORD 215

interesting to notice that the older terraces repre- sent a much longer time span and are polycyclic and therefore difficult to correlate directly with the climatic history.

Comparison of s imulation results with the Maas- tricht terrace sequence

In Fig. 11, a simulated schematic cross-section after 2400 time steps (ka) has been plotted to- gether with the schematic cross-section near Maastricht to compare the MATER simulation re- sults with the actual Maas system. Both cross-sec-

tions display terraces as they occur along a few kilometer of the Maas projected on one cross section. It has to be realized that the Maastricht cross section in Fig. 11 is only a schematization of reality. The effects of loess cover, mass move- ments on slopes and the dissection of terraces by minor tributaries are not included in the schematic cross-section. Because MATER has only conceptual value the correspondence between the model and the Maastricht sequence must be con- sidered qualitatively only.

The major points of agreement between MATER output and the Maastricht sequence are the num-

Height above sea-level (m)

Maas

150

Pietersberg-2

100 ~eter~erg-3

l 's Gravenvoeren

i ! . . . . . . . . . iRothem-2

C;iberg-2

50 Caberg-3 •

Simpelveld -2

rgraten

g-1 -2

~ Geertmid-1

I GeeRruid-2

Geertruid-3

etersberg-1

othem-2

Model

. . . . . . . . . . . . . . . . . . . . . . "~

. . . . . . . . . . . . . . . . . q ~-2T~ ~ 1526-

0 2kin

Fig. 11. Comparison of the Maastricht terrace sequence and the model output.

~60-

Page 14: Three-dimensional modelling of Quaternary fluvial dynamics in a climo-tectonic dependent system. A case study of the Maas record (Maastricht, The Netherlands)

216 A X/I~LI)KAMP AND M.W. VAN DEN BERG

ber of terraces, general terrace sediment thick- ness, and the relative altitudes of the terraces on the valley slopes. There are also some clear dif- ferences between the model and reality. The dif- ference in distribution of the older terraces and the number of younger terraces in the model and reality are the main diverging results.

The older terraces (above P-l) are more regu- larly distributed in the Maas valley than in the model. A possible explanation can be found in the change in lithology in which the Maas valley has incised. In reality the oldest terraces have eroded in Tertiary sediments whereas the younger terraces have incised in more resistant Creta- ceous chalk. This difference in lithology was not incorporated in MATER. Besides lithology, climate and valley depth effects may also account for the observed differences. The difference in distribu- tion between reality and MATER output, and the observed poly-cyclicity of the older terraces may be due to a valley depth effect. A fluvial system in a shallow valley has more lateral (erosional) dy- namics than a similar system in a deeper valley. But one must realize that we also excluded an extra weight upon the harmonics of the 400 and 40 ka climatic cycles as found in deep sea record registering climatic changes older than 1 Ma.

The other difference between model and real- ity concerns the younger terraces. In the simu- lated valley only a few terrace remnants are found witnessing the last 200 ka. In reality the younger terrace records appear very complete. This differ- ence may also be explained by an event not incorporated in the MATER model: the capturing of the Maas headwaters in the Vosges by the Rhine approximately 250 ka ago. This capturing caused a decrease in both discharge and sediment load.

Despite the discussed differences it is striking that relatively simple model inputs can already accomplish such a complicated terrace sequence. Despite the differences between the MATER out- put and the Maas terrace sequence the simula- tion shown is thought to display a possible sce- nario for the development of the Maastricht ter- race s e q u e n c e .

The simulation results clearly suggest that even an apparently regular and complete terrace se-

quence as found in the Maas valley at Maastricht is only partly complete. Only a part of the se- quence at Maastricht from terrace level Geer- truid-3 down to the Caberg-3 level (1000-250 ka B.P.) displays a "complete" terrace chronology since they were formed in a period with stronger uplift rates than the older and younger terraces. In a "comple te" terrace sequence younger than 1 Ma the glacials of every 100 ka cycle have been registered by the occurrence of a terrace. The older terraces ( > 1 Ma BiR) in the simulation are polycyclic and thus represent a longer time span than the terraces from l Ma to 250 ka B.P. This simulation result may also apply to the older Maas terraces in the so-called East Maas valley.

C o n c l u s i o n s

Long-term simulations of the climo-tectonic dependent Maas system by the MATER model suggest that the Maas terraces at Maastricht are mainly the result of general discharge and sedi- ment load ratio response to both climatic and tectonic factors. MATER illustrates that in the Maas system tectonism must have played a domi- nant role in determining the terrace preservation and the actual general valley morphology, whereas climatic dynamics seem to have strongly deter- mined terrace formation by causing changes in water and sediment supply. The differences be- tween model simulation results and the actual Maas system may be well explained by the strong simplifications of climatic en tectonic parameters of the simulated system. The observed differ- ences may also indicate that variations in lithol- ogy and valley depth did also contributed to the terrace distribution in the Maas valley. A plausi- ble explanation of the deviation between mod- elling output and the actual Maas system for the youngest terraces is the capturing of the Maas headwaters by the Moselle system 250 ka ago.

In general it can be concluded that a climo- tectonic dependent fluvial system can register Quaternary oscillations as terraces. Essentially, terraces are records of cold periods. The past and prevailing tectonic settings determine the sensitiv- ity of the fluvial terrace record. It can therefore be concluded that a successful at tempt to corn-

Page 15: Three-dimensional modelling of Quaternary fluvial dynamics in a climo-tectonic dependent system. A case study of the Maas record (Maastricht, The Netherlands)

THREE-DIMENSIONAL MODELLING OF QUATERNARY FLUVIAL DYNAMICS. THE MAAS RECORD 217

pare and correlate different fluvial systems can only be made when their tectonic setting and history is well known.

Acknowledgements

Part of this research was supported by the Netherlands Foundation for Earth Science Re- search and financed by the Organization for the Advancement of Pure Research (NWO). We owe thanks to Prof. Dr. S.B. Kroonenberg from the Department of Soil Science and Geology, Agri- cultural University Wageningen, Drs. C. Staudt, the director of the Geological Survey of the Netherlands, to colleagues of the Geological Sur- vey office at Heerlen for field assistance, and for supplying data, and to Dr C.G. Langereis of the Palaeomagnetic Laboratory, University of Ut- recht, The Netherlands for the palaeomagnetic data.

References

Autran, A. and Peterlongo, J.M., 1980. G6ologie des Pays Europ6ens, France, Belgique, Luxembourg. Dunod, Paris.

Bosch, P.B., 1992. De herkomstgebieden van de Maasgesteen- ten, Grondboor en Hamer, 3: 57-64.

Broecker, W.S. and Denton, G.H., 1990, What drives glacial cycles. Sci. Am., 262: 42-50.

Bruins, H.J., 1981. Paleomagnetische datering van de Maasterrassen en vroeg-Pleistocene loessafzettingen in Zuid Limburg. Int. Rapp. Vakgroep Bodemkd. Geol. Landbouwuniv., Wageningen.

Brunnacker, K., Loscher, M., Tillmans, W. and Urban, B., 1982 Correlation of the Quaternary terrace sequence in the Lower Rhine valley and the northern Alpine foothills of central Europe. Quat. Res., 18: 152-173.

Bustamente Santa Cruz, L., 1976, L'evolution Plio-Pleistocene du basin Mosa d'apr~s ses min6raux lourds. Rev. G6ogr. Phys. G6ol. Dyn., 2, 18 (4): 291-300.

Cloetingh, S., McQueen, H. and Lambeck, K., 1985. On a tectonic mechanism for regional sealevel variations, Earth Planet. Sci. Lett., 75: 157-166.

Cloetingh, S., Gradstein, F.M., Kooi, H., Grant, A.C. and Kaminski, M., 1990. Plate reorganization: a cause of rapid late Neogene subsidence and sedimentation around the North Atlantic? J. Geol. Soc. London, 147: 495-506.

COHMAP Members, 1988. Climatic change of the last 18,000 years: Observations and model simulations. Science, 241: 1043-1052.

Dawson, M.R. and Gardiner, V., 1987. River terraces, the

general model and palaeohydrol ogical and sedimentologi- cal interpretation of the terraces of the Lower Severn. In: K.J. Gregory, J. Lewin and J.B. Thornes (Editors), Palaeo- hydrology in Practise. Wiley, London.

De Jong, J., 1984. Pollenanalytisch onderzoek en CI4-bepaling aan een boring in bet Maasdal by Rijckholt. Rijks Geol. Dienst Rapp., 949.

Drozdzewski, G., Engel, H., Wolf, R. and Vrede, V., 1985. Beitr~ige zur Tiefentektonik westdeutscher Steinkohlen- lagerst~itten. Geol. Landesamt Nordrhein-Westfalen, Kre- feld.

Felder, W.M. and Bosch, P.B., 1989. Geologische kaart van Zuid-Limburg en omgeving, schaal 1:50,000. Afzettingen van de Maas. Rijks Geol. Dienst, Heerlen.

Gregory, K.J. (Editor), 1983. Background to Palaeohydrology. Wiley, London.

Groenewoud, W., Lorenz, G.K., Brouwer, F.J.J. and Molendijk, R.E., 1991. Geodetic determination of recent landsubsidence in The Netherlands. In: Land Subsidence, Proc. 4th Int. Symp. Land Subsidence. IAHS Publ., 200: 463-471.

Guiot, J., Pons, A., De Beaulieu, L.J. and Reille, M., 1989. A 140,000 years continental climate reconstruction from two European pollenrecords. Nature, 338: 309-313.

Gullentops, F., 1954. Contributions ~t la chronologie du Pleis- tocene Inf6rieur en moyenne Belgique. Bjul. Peryglac., 9: 91-95.

Huxtable, J. and Aitken, J., 1985, Thermoluminescence dating results for the palaeolithic site Maastricht-Belvedere. Meded. Rijks Geol. Dienst 39 (1): 41-44.

Imbrie, J., Hays, J., Martinson, D., McIntyre, A., Mix, A., Morley, J., Pisias, N.., Prell., W. and Shackleton, N.J., 1984. The orbital theory of Pleistocene Climate: Support from a revised chronology of the marine 8180 record. In: A. Berger et al. (Editors), Milankovith and Climate. Rei- del, Dordrecht, pp. 269-305.

Kalvoda, J., 1988. Recent orogeny of the Himalayas--A re- markable geomorphological event. J. Geodyn., 9: 319-329.

Kooi, H., Hettema, H. and Cloetingh, S., 1991. Lithospheric dynamics and the rapid Plio-Quaternary subsidence phase in the southern North Sea basin. Tectonophysics, 192: 245-259.

Krook, L., 1961. Verslag van het onderzoek naar de kli- maatscondities tijdens de afzetting van de Maasterrassen. Thesis. Univ. Utrecht.

Kukla, G., 1987. Loess stratigraphy in central China. Quat. Sci. Rev., 6: 191-219.

Kukla, G., Heller, G., Liu, X.M., Xu, T.C., Liu, T.S. and An, Z.S., 1988. Pleistocene climates in China dating by mag- netic susceptibility. Geology, 16: 811-814.

Lowe, J.J. and Walker, M.J.C., 1984. Reconstructing Quater- nary Environments. Longman, New York.

Meissner, R., Springer, N., Murawski, H., Bartelsen, H., Fluh, E.R. and Durschner, H., 1983. Combined seismic reflec- tion-refraction investigation in the Rhenish Massif and their relation to recent tectonic movements. In: K. Fuchs et al. (Editors), Plateau Uplift. Springer, Berlin, pp. 276- 287.

Page 16: Three-dimensional modelling of Quaternary fluvial dynamics in a climo-tectonic dependent system. A case study of the Maas record (Maastricht, The Netherlands)

218 a V H _ D K A M P A N D M.W. V A N D E N BIER(;

M6rner, N.A. (Editor), 1991. Uplift of the Tibetan plateau: A short review. Special proceedings. Rev. Rep. Syrup. 13th INQUA Congr., Beijing.

Patijn, R.J.H., 1963. Tektonik yon Limburg und Umgebung. Verb. K. Ned. Geol. Mijnb. Gen., 21-2: 9-24.

Raymo, M.E., Ruddiman, W.F., Backman, J., Clement, B.M. and Martinson, D.G., 1989. Late Pliocene variations in northern hemisphere ice sheets and north Atlantic deep water circulation. Paleoceanography, 4: 413-446.

Ruddiman, W.F., Raymo, M.E., Martinson, D.G., Clement, B.M. and Backman, J., 1989. Pleistocene evolution: North- ern Hemisphere icesheets and North Atlantic Ocean. Pa- leoceanography, 4: 353-412.

Schumm, S.A., 1977. The Fluvial System. Wiley, New York. Shackleton, N.J., Berger, A. and Peltier, W.R., 1990. An

alternative astronomical calibration of the Lower Pleis- tocene time scale on ODP site 677. Trans. R. Soc. Edin- burgh Earth Sci., 81: 251-261.

Starket, L., 1990. Fluvial environment as an expression of geo-ecological changes, Z. Geomorphol. N.F., 79: 133-152.

Thornes, J.B., 1987. Models for palaeohydrology in practice. In: K.J. Gregory, J. Lewin and J.B. Thornes (Editors), Palaeohydrology in Practice. Wiley, London.

Van den Berg, M.W, 1989. Toelichting op kaartblad 59-62, Geomoffologische Kaart Nederland. DLO-Staring Cen- trum, Wageningen/Rijks Geologische Dienst, Haarlem, 32 pp.

Van den Berg, M.W., 1990. Morphotectonics of the Roerval- Icy Graben. In: European Cenozoic Rift Valley systems, ECRIS Symp., Basel (abstract).

Van den Berg, M.W., 1993. River Maas terraces as a long continental record of climo-tectonic changes in north-west Europe. In: 7th EUG Biannu. Meet., Strasbourg, 1993 (abstract).

Van Straaten, L.M.J:U., 1946. Grindonderzoek in Zuid-Lim- burg. Meded. Geol. Sticht. Ser. C.I, 3.1.

Veldkamp, A., 1991. Quaternary river terrace formation in the Allier basin, France. A reconstruction based on sand bulk geochemistry and 3-D modelling. Thesis. Agric. Univ. Wageningen.

Veldkamp, A., 1992. A 3-D model of Quaternary terrace development, simulations of terrace stratigraphy and val- ley asymmetry, a case study in the Allier terraces (Limagne, France). Earth Surf. Proc. Landforms, 17: 487-500.

Veldkamp, A. and Vermeulen, S.E.J.W., 1989. River terrace formation, modelling, and 3-D graphical simulation. Earth Surf. Proc. Landforrns, t4: 641-654.

Van Waterschoot van der Gracht, W.A.J.M., 1938. A struc- tural outline of the Variscan front and its foreland from south-central England to eastern Westphalia and Hessen. C.R. 2~me Avanc. Stratigr~ Carbonif~re, Heerlen, 1938-3, pp. 1485-1565.

Zagwijn, W.H., 1974. Pollenanalytisch onderzoek van een humeuze kleilaag in de groeve bij Susterseel en een op- merking n.a.v, rapp. 369. Rijks Geol. Dienst, Haarlem, Rapp., 719, Paleobot.

Zagwijn, W.H., 1985. An outline of the Quaternary stratigra- phy of the Netherlands. Geol. Mijnb., 64 (1): 17-24.

Zagwijn, W.H. and De Jong, J., 1983. Die Interglacial yon Bavel und Leerdam und ihre stratigraphische Stellung in Niederl~indische Friih-Pleistoz~in. Meded. Rijks Geol. Dienst, 37 (3): 155-169.

Zijerveld, L. Stephenson, S_ Cloetingh, S., Duin, E. and Van den Berg, M.W., 1992. Subsidence analysis and modelling of the Roer Valley Graben (S.E. Netherlands), Proc. Geo- dynamics of Rifting Symp. Tectonopbysics, 208: 159-171.

Zonneveld, J.I.S., t949. Zand-petrologische onderzoekingen in de terrassen van Zuid-Limburg. Meded. Geol. Stichting. N.S., 3: 1948-1949: 103-123.

Zuchiewitcz, W., 1991. On different approaches to neotecton- ics: A Polish Carpathians example. Episodes, 14 (2): 116- 124.