• No results found

Holocene Rhine reoccupation of the IJssel valley by divide dissection north of Zutphen

N/A
N/A
Protected

Academic year: 2022

Share "Holocene Rhine reoccupation of the IJssel valley by divide dissection north of Zutphen"

Copied!
1
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

880 880 990 RWS km langs geul 980

10 km

V V V V

NAP 10

v v v v V V V V V V V

v v v v V V V V V V V

v v v v V V V

V V V V v v v v V V V V V V V

v v v v V V V V V V V

v v v v V V V V V V V

v v v v V V V V V V V V V V

V V V V V V V V V

V V V V V V V V

V V V V V V V V

V V V V V V V V

14 12

8 6 4 2

-2 -4 -6 -8

NAP +10 (m)

890 900 910 920 930 940 950 960 970 1000

890

50 100

0 km

downstream distance

V V V V

V

V

Holocene Rhine reoccupation of the IJssel valley by divide dissection north of Zutphen

K.M. Cohen k.cohen@geo.uu.nl Dept. Physical Geography, Fac. Geosciences, Utrecht University PObox 80.115 3508 TC Utrecht, Netherlands

Cohen, K.M., E. Stouthamer, W.Z. Hoek, H.J.A. Berendsen & H.F.J. Kempen (2009) Zand in banen. 3rd fully revised edition. Arnhem: Provincie Gelderland. 130 pp.

+ CD-ROM + Digital Maps. In Dutch, with a summary in English.

Late Pleistocene

Local eolian and fluvial deposition

Channel deposits: sands and gravelly sands, with lateral and vertical dissective architecture indicated.

Hippo and Mammoth skull not to scale.

Overbank deposits (Wijchen Member) Late glacial local drainage trunk systems Berkel periglacial local river deposits

Veluwe periglacial alluvial fans and fan toes

Overijsselsche Vecht periglacial local river deposits Eolian dunes (Wierden & Delwijnen Members)

Boxtel FormationKreftenheye Formation

Rhine deposition

Holocene

Local deposition

Peat (Nieuwkoop Formation) Rhine delta deposion

Level of top of IJssel natural levees Floodplain and -basin deposits, clay

Level of top of subrecent pointbar sands along Nederrijn and IJssel Rhine branches

Level maintained by Early Holocene local drainage (top of brook channel deposits)

IJssel embankment level (situation 1995)

Echteld Formation

V V V V V

V V V V V V V V V V V V V V V V V

V V V V V V V

V V V V V V V

V V V V V V V

V V V V V V V V V V

V V V V V V V

V V V V

V V V V V V V V V V

V V V V V V V V V V V V V V V V V V V V V V V V V

V V V V V V V V V V

V V V V V V V V V V V V V V V V V V V V V V V

V V V V V V V V V V V V V V V V V V V V V

V V V V V V V

V V V V V V V

V V V V

Nederrijn

Dortherbeek Berkel

Oude IJssel Zwarte Water

VELUWEZOOM

ZUIDOOST VELUWE

WOLDBERG IJsselmeer

(Zuiderzee)

Ov. Vecht

Aa KMC ‘09

OOST VELUWE

IJssel

Drainage and geomorphology (straightened map)

Annotated geological section along the axis of the IJssel valley

LIEMERS ACHTERHOEK GLD | OV SALLAND IJSSEL DELTA

GLD | OV

OVERBETUWE

Legend

IJssel embankment IJssel valley Rhine

[80-40 ka]

‘Rond-Montferland’Rhine[60-25 ka]

valley Oude IJssel

Rhine-filled glacial tongue basin: ‘Twello Mb.’

End of Saalian [145-130 ka], Termination II Valley edge 40-25 ka Rijn Valley edge

Oude IJssel Coversand ridges

Voorst-Gorssel-Wilp

Rijndelta [since 5 ka]

Nederrijn

[80-?60 ka]

[?60-40 ka]

slope and fan complex Woldberg

Berkel fan IJssel delta

since 1000 yr

[40-25 ka]

[25-15 ka]

Veluwe fan complex

top IJssel natural levee top sand IJssel pointbars

valley VechtOv.

IJsselmeer [40-15 ka]

[40-15 ka]

[13-9 ka]

11.7 5 2 1 0 ka

V V V

0 = 2000 AD Veluwe periglacial fans

Berkel periglacial fan uncovered

uncovered coveredcovered

130 80 40 25 15 11.7 ka

‘Zutphen Member’

‘Eemian’ Rhine [130-80 ka]

Former drainage, predating the IJssel

V V V V

Lower reaches of combined

Oude IJssel + Berkel local rivers Incised lower reaches of combined

Salland +Veluwe-source local rivers

IJssel

low high

valley rim topography present former

channel of the river IJssel Doesburg Arnhem

Zutphen

Deventer

Kampen

Zwolle Olst Wijhe

Doesburg

Zutphen

Deventer

Olst Wijhe

Zwolle

Kampen Arnhem

‘Gelderse Poort’

Rhine [25-15 ka]

Gelderse IJssel

DIVIDE CONTROL ON REOCCUPATION

The channel deposits of the Gelderse IJssel show a two-staged division. The oldest stage (‘initial stage’) saw multiple shallow channels and occupied a relatively wide belt within the valley. The younger stage (‘mature stage’) occupies a relatively narrow channel belt of greater thickness, and showed all the phenomena of a meandering river. The river was embanked in the 14th century, continued to meander for two more centuries (still ‘mature stage’), and then lost most of its discharge due to discharge redistribution events at the upstream delta apex.

divide area

GEOLOGY OF DIVIDE FORMATION

Since ~40,000 years ago - after the Rhine abandoned its IJssel valley course between 60,000 and 40,000 years ago - local periglacial depositional sedimentary systems buried the valley floor. Orphaned former Rhine tributaries built out low gradient alluvial fans from both sides of the valley. The fans were fed from catchments draining the Veluwe ice-pushed ridge complex and the eastern Netherlands uplands, respectively.

Through deposition the fans raised the local surface elevation. Around 20,000 years ago, the largest two fans grew so large that they met in the axis of the abandoned valley. At this time lobe avulsions of the Berkel fan redirected this river to a position south of the most extensive Veluwe-sourced fan. As a result, the orphaned Berkel tributary once again became a Rhine tributary. Coversand ridges formed on top of the abandoned part of the fans, sourced from the ephimeral streams on the active parts. This caused further raising of the surface at a local level.

In sequence, these periglacial developments (from ~40,000 to ~11,000 years ago) created the saddle topography that became the drainage divide. The associated drainage network stayed active until late in the Holocene, until the Gelderse IJssel branch reoccupied the area as part of a deltaic avulsion-by-annexation event (in early medieval times).

DATING THE GELDERSE IJSSEL

An apparent gap exists in the dated age of the IJssel ‘upstream’ and ‘downstream’

of the divide area. This is due to the (i) the static presence of the divide until ca.

550 AD and (ii) the centuries necessary for the IJssel breach to mature.

Reaches upstream of the divide were gradually flooded by Rhine waters from 2500 yr ago onward. Rhine-flood slackwater deposits accumulated up the tributary valleys of the rivers Oude IJssel and Berkel, between 2000 and 1500 AD in the Zutphen vicinity. The reaches downstream of the divide record hydrological changes from 350 AD and 550 AD onwards. Rhine clay deposition began later, around 750 AD (recent 14C results; Cohen, in prep.) and expanded around 900 AD (Makaske et al. NJG 2008; 14C dating). Historic sources and archeological evidence imply that the IJssel exists through the divide region since at least 600 AD (e.g. Fermin &

Groothedde, ZAP, Gem. Zutphen).

Combined sedimentary and dating evidence constrains the stages as follows:

1) The ‘Initial Stage’ occurred between ca. 500 and 900 AD, with 550 AD as a best guess age for an initial rare-magnitude flood triggering divide breaching to commence, based on dendrochronology and paleohydrology from a former oak swamp at Zwolle (Sass-Klaassen & Hanraets NJG 2006);

2) The ‘Mature Stage’ occurred between 900 and 1550 AD, based on increased clay deposition in the lower IJssel, increased sedimentation in the IJssel delta since that time, and historical information on measures mitigating ongoing channel migration near the medieval cities of Zutphen, Deventer, and Doesburg.

METHOD The reconstruction is based on systematic lithogenetic mapping, coring to 5 meters below surface depth (and occassionally deeper). 10-20 boreholes / km2 were strategically placed based on high resolution digital elevation data (AHN 1 5x5m; RWS-AGI, 2005). The geomorphological analysis and dating methods included: separation of in-situ Rhine deposits from locally reworked units on architectural and sedimentological grounds; palaeohydrological indications from periglacial and interglacial times; OSL dating of Rhine and local periglacial strata; 14C and pollen-biozone dating of Late Glacial, Early Holocene and later Holocene brook valleys buried below IJssel overbank deposits; separation of mature-stage meandering IJssel channel deposits from initial-stage crevasse deposits on architectural and sedimentological grounds;

14C, pollen-biozone and archeology-based dating of younger brook deposits and IJssel deposits, and application of established coversand stratigraphy and climate-hydrology-discharge-sediment yield morpho-stratigraphical schemes for the time period for the Netherlands.

Thanks to Wim Hoek, Esther Stouthamer, Henk Berendsen, Henny Kempen, Freek Busschers, Michel Groothedde, Nico Willemse, Maarten Kleinhans Annelies van Hoesel, Rens Quak, Marieka Brouwer and many others. Thanks to ~100 BSc and 10 MSc students since 2005.

Preserved channels of the initial stage are ‘avulsion-splay crevasse channels’. They reworked lows in coversand ridges of the divide, and had splays downstream of these locations. As time progressed and repeated Rhine floods forced themselves to the divide breach, the number of active channels dropped, and the width and incision-depth of the remaining channels increased. This caused the Gelderse IJssel to mature into a single channel meandering system. In the divide region the mature meander belt occupies a slightly incised architectural position: the top of IJssel sand bars is encountered decimeters to a meter lower than equivalent initial stage deposits. To the south and the north of the divide, patches of the brook valley deposits occur preserved buried between crevasse channels. They are dated and used to constrain the begin of avulsion splay formation.

ABSTRACT

The Gelderse IJssel is the youngest natural distributary of the Rhine delta.

When it came into existence between 300 and 700 AD, it annexed a valley that had not previously been part of the Holocene delta plain before. Understanding the drainage configuration of the valley and the position of the main divide, which separated ‘Rhine tributaries’ from ‘drainage going north’ prior to the formation of the IJssel, is key when reconstructing the timing and mechanics of the annexation process. The initiation of the IJssel is of archeological- historical interest, and the features from the initial stage - preserved some meters above deposits from later stages - offer a great opportunity to quantify a major flood of the Rhine.

During 5 years of field campaign covering the full river valley length and width, we mapped and dated landscape features related to the IJssel diversion and the area’s many dissected older elements. This allowed us to position the paleodivide between Zutphen and Deventer, and to explain how it formed. The outcomes of this research falsify former paleodrainage reconstructions that assume the divide lay further south and that showed the rivers Berkel and Oude IJssel draining northward, instead of heading west towards the Rhine.

divide area

Present river IJssel

Minimal elevation(m O.D.) breached by crevasse channels

All ages based on oldest (known) historic toponym mentioning; ‘<’ = ‘existed from before’; † = church / chapel founding

Netherlands’ national RD grid (km)

= historic church move

Gelderse IJssel channel belt maturisation Avulsion splay formation [350-700 AD]

Early-Holocene inherited drainage situation

periglacial outwash fan coversand dune ridge coversand relative lows

Brook valleys

Brook channels, traced

Brook channels projected Early Medieval fortified settlements ‘Zutphen’ and ‘Deventer’

Crevasse channels

Splays and initial multi-channel avulsion belt formation Channel belt

Silted-up splay floodplain Isolated breach splays

Active channelsand moats Residual channels

Medieval cities, first dikes

205 210 215 205 210 215 205 210 215

460465470

460465470

460465470

7.5

Z V

V VV

D

G

W

7.5 7.5 7.5

7.5 7.5 7.5

7.5 7.5 7.5

7.5 7.5 7.5

Wp

Cs

Dt

E

Z

Deventer † ~768 AD fortified 866 / 882 AD

Zutphen, since ~300 AD fortified 866 / 882 AD

D

Voorst †

< 893 AD

Brummen 794 AD

Eerbeek <1096 AD Wilp †

~965 AD

Dorth

<1311 AD

Warnsveld †

< 1121 AD Berkel canals, water mill

1225-1250 AD kunstmatige bocht- afsnijding 1357 AD

N

? ?

Hunnepe Dortherbeek

Berkel

Voorsterbeek Voorsterbeek

Former valley of river Berkel active c. 40.000-20.000 yr ago ther

eafter coversand bur

ied Pingo-remnant cluster:

paleo-seepage zone

Holocene seep

age zone (FeO) Higest coversand ridges last active 12.500-1

1.700 yr ago sourced from periglacial alluvial fans

Withmundi 794 † 797 AD

? Wychmoith

† < 1500 AD Holocene

seep age zone (FeO) Pingo-remnant cluster:

paleo-seepage zone

river Berkel

BEFORE DURING AFTER

Zutphen

Doesburg Deventer

Oude IJssel Berkel

Zwolle Kampen

Dortherb.

Arnhem

Referenties

GERELATEERDE DOCUMENTEN

Research papers and interviews will be analysed by applying a coding scheme, this provides a methodologically grounded approach to investigate the influence of scientific

Development of a 2D hydraulic model for the Rhine valley using open data 25 That the coarse grid model simulates more flooding is supported by figure 16, which shows the total

more prominent were the contacts from the end of the fifth millennium onward, with the in many respects differently organised neolithic communities of the loess

Non-linear features – trend surface method, aided by geomorphological data – void fill Active zone – geological data, aided by LiDAR – IDW between points River –

We compared avulsion history to human activities, such as historical dam construction, lower delta peatland reclamation, catchment deforestation leading to enhanced erosion

We present a full 3D interpolation describing groundwater rise in two neighboring river valleys based on age- depth relations of basal peat index points. for GW levels through

• During the early Holocene, deposition of fines is high because of i) better trapping conditions, ii) a delayed arrival of sediment produced during glacial times, slowly

We must Interpret this site at Bergschenhoek äs an extraction camp, subsidiary to (semi-)permanent settlements like those near Swifterbant and on the Hazendonk, that is,