ter verkrijging van de graad van doctor aan de Technische Universiteit Delft,
op gezag van de Rector Magnificus prof. dr.ir. J.T. Fokkema, voorzitter van het College voor Promoties,
in het openbaar te verdedigen op dinsdag 31 januari 2006 om 15.30 uur
Robert Mark HOOGENDOORN doctorandus in de Fysische Geografie
Toegevoegd promotor: Dr. G.J. Weltje
Rector Magnificus, voorzitter
Prof. dr. S.B. Kroonenberg, Technische Universiteit Delft, promotor
Dr. G.J. Weltje, Technische Universiteit Delft, toegevoegd promotor Prof. dr. M.J.F. Stive, Technische Universiteit Delft
Prof. dr. P.L. de Boer, Universiteit Utrecht Prof. dr. ir. A. Veldkamp, Wageningen Universiteit
Dr. E. Aliyeva, Academy of Sciences, Azerbaijan
Dr. J. Kwadijk, WL Delft
Prof. dr. S.M. Luthi, Technische Universiteit Delft, reservelid
ISBN-10 90-9020310-9 ISBN-13 978-90-9020310-2
Copyright © 2006 by R.M. Hoogendoorn, Section of Applied Geology, Faculty of Civil Engineering and Geotechnology, Delft University of Technology. All rights reserved, no parts of this thesis may be reproduced, stored in any retrieval system or transmitted, in any forms or by any means, electronic, mechanical, photocopying, recording or otherwise without the prior written permission of the author.
Graag zet ik de traditie van het, vaak, veel gelezen voorwoord/dankwoord voort. Maar alvorens ik de hogere krachten bedank die mij “de weg wezen…”, zou ik graag kort een aantal mijlpalen willen noemen die niet in het proefschrift aan bod komen maar daar wel invloed op hebben gehad. Zo is het mij privé voor de wind gegaan gedurende de afgelopen 5 jaar. Hoogtepunten zijn mijn huwelijk met Tanja-anne en de geboortes van Isabelle en Lily. Graag wil ik dan ook melding maken van het feit dat Sue en Liska (resp. mijn moeder en schoonmoeder) wekelijks bijdragen aan ons gezinsgeluk door regelmatig (1 keer in de 2 weken) af te reizen naar Den Haag om een dag met de meisjes door te brengen. Zonder hun zorg zou het onmogelijk geweest zijn voor Tanja-anne en mijzelf om een dergelijke tijdsintensieve opleiding te volgen en volbrengen zoals wij doen. Ik ben ook zeer blij met het feit dat ik binnen de sectie Toegepaste Geologie de ruimte heb gekregen om voldoende tijd en aandacht aan ons gezin te besteden. Het is niet te meten maar ik ben ervan overtuigd dat het een positieve invloed heeft gehad op het onderzoek en het schrijven van het proefschrift.
Een ander punt dat ik graag wil noemen is het bezoek aan een crèche voor orang oetangs tijdens het veldwerk in Kalimantan, (Borneo) Indonesië. Daar worden jonge orang oetangs opgevangen die als gevolg van (vaak illegale) houtkap hun moeder en territorium zijn kwijt geraakt. De inzet van de mensen van het Wereld Natuur Fonds voor deze bijzondere beesten is indrukwekkend. Ik ben er dan ook van overtuigd dat het werk wat door het Wereld Natuur Fonds daar wordt uitgevoerd kan bijdragen tot het behoud van orang oetangs en het tropische woud in Borneo. Wellicht ongebruikelijk, maar anticiperend op het gebruik dat vrienden en familie een presentje geven na de openbare verdediging van een proefschrift, zou ik u willen vragen als u dat van plan bent om in plaats daarvan een bijdrage over te maken naar het Wereld Natuur Fonds in Zeist op 188.8.131.520, onder vermelding van 'promotie Hoogendoorn'.
Dit proefschrift zou er niet zijn geweest zonder de initiator en mijn promotor Salomon Kroonenberg. Salle, ik wil je bedanken voor de kans die me hebt gegeven om dit onderzoek te beginnen en de steun die je me gedurende vijf jaar onvoorwaardelijk hebt gegeven om het onderzoek af te ronden. Verder waren de Kaspische veldwerken een onvergetelijke ervaring en een essentieel onderdeel van dit proefschrift. Het succes daarvan is het resultaat van je enthousiasme en je vakkennis.
De volgende die ik wil noemen is Joep Storms, mijn kamergenoot voor vier jaar en veldwerk partner in Kalimantan. Veldwerken zijn altijd gedenkwaardig en op een bootje varen door de rimboe van Borneo was dat zeker. Je hebt ook vaak als een klankboord gefungeerd voor mijn initiële ideeën waardoor de uitwerking altijd beter werd. Daarnaast ben je in mijn ogen een bijzonder goede wetenschapper en is je werk een voorbeeld geweest voor mijn eigen werk.
A short period spent at INSTAAR in Boulder, Colorado, USA was pivotal for my work. I want to thank James Syvitski for inviting me and spending so much time discussing my work during this period.
Irina Overeem and Albert Kettner, behalve oud collega’s werkten zij bij INSTAAR gedurende mijn bezoek en hebben mij bijzonder gastvrij ontvangen in hun huis en op sleeptouw genomen. Die periode heeft ook tot de nodige wetenschappelijk samenwerking geleid, niet ten minste door jullie aanstekelijke enthousiasme voor de sedimentaire geologie.
Ik ben altijd met plezier naar Delft gegaan als gevolg van de aangename werkomgeving. Ik wil graag alle collega’s van de sectie toegepaste geologie, secretariaat en aanverwante secties hiervoor bedanken. Specifiek wil ik de “lunchclub” noemen: Joep, Klaas, Luc, Remco, Marit, Israel, Jose, Rory, Gert Jan, Jan Kees, Jon. De dagelijkse onderbreking van het werk was zeer welkom en de terugkerende onderwerpen van gesprek hoezeer voorspelbaar brachten voor mij de broodnodige luchtigheid in het dagelijkse isolement van de promovendus.
Drie grote veldwerken, staan aan de basis van dit proefschrift, waarbij veel mensen betrokken waren, Joep en Salle heb ik al genoemd, daarnaast waren Jelle Boels, Klaas Scholte, Bagir Ibrahimov, Elmira Aliyeva, Dadash Huseinov, Peter Kloosterman, Anneke Hommels en Rien Dam er ook bij. Dankzij hun inzet zijn alle veldwerken bijzonder succesvol verlopen. Triest was het overlijden van Peter Kloosterman. Peter heeft als afstudeerstudent mee gewerkt aan het tweede veldwerk in de Kura delta in Azerbeidjaan en heeft een grote rol gespeeld in het succesvol afronden daarvan.
Naast mijn werk als AIO heb ik het genoegen gehad om deel uit te maken van het organiserende comité voor het “8th International Conference on Fluvial Sedimentology”. Ik wil graag alle leden van de diverse comités en de dames van het aula congres bureau hartelijk danken voor de aangename samenwerking.
verdediging komen om het einde van deze periode te vieren. Michiel en Niek wil ik speciaal noemen en bedanken voor de oprechte vriendschappen.
Hoewel het 30 jaar geduurd heeft, heb ik mijn onderwijsperiode nu afgesloten (4 maanden voordat Isabelle zich in het moderne onderwijs zal storten). Gedurende deze jaren heb ik opvallend weinig commentaar en veel steun gekregen van het thuisfront. Daarom wil ik mijn ouders, Chris en Sue, bedanken voor hun steun en inzet gedurende deze gehele periode waarin veel is voorgevallen, besproken en waar jullie deur altijd open was voor mij. Maar ook voor hun ongelofelijke gulheid, onder andere in de vorm van het diner om deze promotie te vieren. Verder denk ik dat dit succes vooral mogelijk is gemaakt door de goede basis die jullie mij hebben meegegeven. Bij deze basis horen natuurlijk ook mijn zus, Christine, en broer, Sebas, (jammer dat je er niet bij kan zijn maar geniet van je tijd in Australië!). Mijn schoonfamilie, Ber, Liska, Steven en Homa wil ik bedanken voor hun gastvrijheid en de gemeende interesse in het onderzoek en de bijbehorende discussies.
Chapter 1 Introduction
1.1 General 1
1.2 Delta studies and societal relevance 4
1.3 Objectives 5
1.4 Standard delta classification 5
1.5 Deltaic stratigraphy 7
1.6 Quantitative modelling of fluvio-deltaic Stratigraphy 10
1.7 Thesis outline 12
Chapter 2 Development of the Kura delta, Azerbaijan; a record of Holocene
Caspian sea-level changes
2.1 Introduction 13
2.2 Regional setting 15
2.2.1 Geologic setting 15
2.2.2 River system characteristics 15
2.2.3 The Caspian Sea 16
2.2.4 Delta morphology 19
2.3 Delta sediments and stratigraphy 20
2.3.1 Lithology 20
2.3.2 Depositional environments 22
2.3.3 Sparker data 24
2.4 Laboratory analyses 28
2.4.1 Radiometric Dating 28
2.4.2 Biostratigraphic dating and diatom analysis 30 2.4.3 Grain-size distribution of modern sandy surface sediments 30
2.5 Depositional history 32
2.6 Discussion 36
Chapter 3 Late-Holocene evolution of the Mahakam delta, East Kalimantan,
3.1 Introduction 41
3.2 Regional Setting 43
3.2.1 Geologic setting 43
3.2.2 River system characteristics 44
3.2.3 Marine processes 44
3.2.4 Delta morphology 45
3.3 Data collection 47
3.4 Sediments and Stratigraphy of the late-Holocene delta development 50
3.5 Facies 52
3.6 Facies succession 54
3.7 14C dates versus eustatic sea level history 56
3.8 Deposition rates 57
3.9 Discussion 58
3.10 Conclusions 60
Chapter 4 A stochastic model for simulating long time series of river-mouth
discharge and sediment load
4.1 Introduction 63
4.2 Model theory 64
4.3 Model validation, simulations based on discharge measurements 69
4.4 Model prediction of discharges 72
4.5 Kura River simulation 75
4.6 Climate change and sediment supply 78
4.7 Discussion and conclusions 81
Chapter 5 Estimating world-wide liquid and discharge and sediment delivery
to the oceans.
5.1 Introduction 83
5.2 Data 85
5.2.1 Databases 85
5.2.2 Data structure 86
5.3 Partial least square regression 87
5.3.1 Theory 87
5.3.2 Application of PLS-R to global precipitation…. 90
5.4 Determining the sediment flux 93
5.5 Precipitation control on sediment production 96
5.6 Application of the PLS-R method to the Kura River 99
5.6.2 A conceptual cold- warm transition 100
5.7 Discussion and conclusions 102
Chapter 6 Process response modelling of fluvial deltaic systems
6.1 Introduction 105
6.2 Process response model of fluvial deltaic evolution 106
6.2.1 Erosion 107
6.2.2 Deposition 108
6.3 Model variables 109
6.3.1 Time dependent variables 109
6.3.2 Initial conditions 110
6.3.3 Model parameters 110
6.4 Calibration 111
6.5 Model applications 114
6.5.1 Grain size 114
6.5.2 Rate of sediment supply 116
6.5.3 Discharge variability 117
6.5.4 Rate of sea level change 120
6.5.5 Initial slope 121
6.5.6 Probabilistic model output. 122
6.5.7 Simulation of the Kura delta 124
6.6 Discussion and conclusion 126
The impact of changes in sediment supply and sea-level on
The processes that create and change landscapes on the earth can be summarised as a cycle of events. Lithospheric plates drift apart and collide forming the major structural features of the earth such as volcanoes and mountain ranges. Erosive powers continuously, slowly and steadily degrade these high grounds. The eroded material is transported to the ocean predominantly by rivers where it is deposited as sedimentary formations in basins, especially in deltas, and remain there to be recycled by tectonic processes in due time. It is estimated in this thesis that each year ca. 15 G ton of sediment is being transported by rivers to the oceans worldwide. (15 Giga ton yr-1 is equal to ca. 750.000 truck loads with 20m3 of sediment deposited in the oceans each day or two billion shoeboxes filled with clay and sand emptied in the seas each day).
A delta is generally defined as a “discrete shoreline protuberance (a lump of sediment) formed where a river enters an ocean or lake” (Bhattacharya and Walker 1992). When a river arrives at open water the flow velocity of the river decreases abruptly. Hence the river’s competence to carry sediment decreases and most sediment in transport is deposited, forming a delta. This phenomenon was first described ca. 400 B.C. at the mouth of the Nile River, where the subaerial part of this sedimentary system had the form of the Greek letter delta (Suter 1997). It is important to realise that a delta does not only comprise the visible subaerial onshore delta plain, but also the subaqueous delta front and prodelta (Figure 1.1).
than open water) buoyancy forces resulting from the density contrast dominate. The suspended sediment is relatively fine grained, and tends to be carried far out into the basin as a buoyant plume, resulting in relatively low depositional slopes and laterally extensive deltaic deposits. Most major rivers debouch in the salty oceans and since salt water has a higher density than fresh water, even when it contains relatively large amounts of suspended load, buoyancy forces will predominate. As a result, almost all river flows entering open water are classified as hypopycnal (Bates 1953).
Figure 1-1. Schematic representation of the basic environments of a fluvial dominated delta and the two main allocyclic controls, sediment supply and sea-level, studied in this thesis
Figure 1-2. A) Eustatic level curve for the last 18.000 yr, (after Fairbanks 1989) shows the sea-level rise that has occurred since the last glacial maximum (last ice-age). This curve is combined with a prediction range (grey area) for the coming 3.000 yr based on data from the Intergovernmental Panel on Climate Change (IPCC) that predicts a probable maximum rise between 1 and 5 meters compared to present day sea-level (after Houghton et al. 2001) (k yr = 1000 yr) B) Enlargement of the prediction range for global average sea level rise (1990-2100), shows an expected rise to be between 0.1 and 0.9 meters in 2100 AD, figure B is represented by the square in curve of figure A.
A change in the volume of water driven by climate change is the major force in present day sea-level rise. Evidence for global warming is e.g. illustrated by an accelerated retreat of glaciers (Scambos et al. 2004). Changes in the amount of river discharge are directly related to changes in climate conditions and will influence the power of the river as well as the amount of sediment it can supply. Climate conditions influence the erodibility of soils in the drainage basin, and the density of stabilizing vegetation. Since climate is ever-changing it is important to realise that the liquid and solid discharge supplied to a delta varies over time (Ruddiman 2001). The dynamic interaction between (relative) sea-level change and sediment supply is viewed as the main (allocyclic) control on fluvio-deltaic evolution (Emery and Myers 2001) (Figure 1.1).
Delta studies and societal relevance
Deltas are important depositional systems, both ecologically and economically. They often contain extensive wetlands, with high bio-productivity. From a socio-economical perspective deltas are important for their extensive agricultural activities, supported by fertile soils and abundant water supply. Furthermore many of the world’s ports are located in a delta, an excellent location for the transit of products. Delta plains are the areas at risk of inundation due to sea-level rise. Therefore, predicting the future development of deltas is becoming more urgent imposed by environmental, social and political concerns. Scientists are challenged by questions on how the accelerating sea-level rise in the coming decades will influence deltaic development. Besides environmental and social concerns, emerging questions are related to the technical limitations and benefits of possible measures. In the coming centuries (geologically close future) policy makers in e.g. the Netherlands will have to answer elementary questions as: “Should we stay or should we go? “. Therefore a growing effort is made to develop concepts to handle the complexity of deltaic development to changes in sea-level and sediment supply driven by climate change.
At present predictions of the development of future deltaic architecture are insufficient. This is a consequence of the lack of data on the long-term behaviour of deltas. Subsequently, there is only limited knowledge on the processes controlling long-term deltaic evolution. Many geologist have stated that the key to the past is the present, which imposes the question ”what is the key to the present?” (Bates 1953). As far as sedimentary processes are concerned, the key of the present lies in qualitative and quantitative understanding of sedimentary systems. This should lead to the development of objective methods for reconstructing environmental conditions of the past and for predicting future conditions.
This thesis addresses the problem of qualitative and quantitative understanding of sedimentary systems in two stages. In the first stage the stratigraphy of Holocene delta deposits subjected to (rapid) sea-level change is studied. Results from these studies can provide insights as an analogue to ancient systems or serve as an example for expected future conditions. Information obtained from studies in a similar context has the potential to reduce uncertainties and increase confidence in models. The second stage of this thesis covers the development of a process-response simulation model to assess the impact of changes in sediment supply and sea-level on fluvio-deltaic stratigraphy. Anticipating on the dominant role of sediment supply on fluvial dominated deltas, a choice is made for a detailed investigation in how to incorporate good estimates of sediment supply in the numerical model.
Besides the individual objectives of each chapter, this thesis uses data of the Kura delta and River for each chapter as a theme. The objective of this exercise is to show that data driven field studies and the numerical models should be viewed as complementary. More specifically, numerical models are as much part of the geologist toolbox in sedimentary geology as e.g. sequence stratigraphy.
Standard delta classification
A much applied classification scheme based on typical morphological characteristics of deltas was suggested by Galloway (1975). This scheme identifies three major forces that act upon delta morphology; river, wave and tidal forces (Figure 1.3). In theory, all three forces are present at any oceanic delta, but their proportional contribution will differ for each specific delta, giving rise to the tripartite scheme of Galloway (1975). The dominance of one of these forces in delta development leads to a recognizable end-member stage in delta morphology.
Wave dominated deltas in which shallow marine processes rework and disperse the sediments tend to be destructive (Abdulah et al. 2004; Storms 2002). These deltaic systems are recognised by beach ridges along the delta front, caused by the redistribution of the sediments by waves. For instance, the Rhine-Meuse delta in the Netherlands is subjected to reworking of sediments by waves.
morphology and sedimentary patterns (Chang-shu and Jia-song 1988). The morphology of tide-dominated deltas is characterised by funnel-shaped channels which accommodate the in- and out-flow of water caused by the tidal forces. An example of a delta affected by tidal forces is the Mahakam delta in Indonesia, which is presented in chapter 3.
In contrast, fluvial-dominated deltas tend to display a constructive evolution, meaning that in theory they will prograde (relative) more rapidly due to the absence of reworking of sediments by marine processes and the (relative) high sediment input. The morphology of fluvial-dominated systems tends to be lobate (semi-circular) if they prograde in shallow water, or elongated (stretched) when they evolve in deeper water near the shelf edge (Bhattacharya and Walker 1992). The most thoroughly studied modern delta in the world is the fluvial dominated, elongated, delta of the Mississippi River (Fisk 1961; Frazier 1967; Roberts 1997; Törnqvist et al. 1996). The fluvial dominated Kura delta presented in chapter 2 is an example of a more lobate delta compared to the Mississippi delta. Galloway’s scheme proves to be a valuable tool to interpret ancient and modern delta deposits, since each of the end members of Galloway’s classification are also believed to be characterized by specific depositional patterns (Galloway 1975).
Stratigraphy is a technique for working out parts of earth’s history. It integrates a diversity of data such as outcrop-, core-, seismic data and radiometric dating techniques into a coherent view of how a system evolved (Brookfield, 2004), e.g. the description of depositional patterns in fluvio-deltaic settings. Stratigraphy can tell much about the processes and the conditions affecting the deposition of sediments, although even simple depositional processes can result in very complex stratigraphy. Thus, long profiles or data from a number of wells are necessary to avoid misinterpretation. A well-known expression states that, when you show three geologists the same field data, you will receive at least four interpretations, none of them conclusive. Disagreement on the interpretation of geological data is most often the result of different insights between geologists dealing with the (incomplete) stratigraphic record.
In stratigraphy there is a broad range of subjects, sequence stratigraphy among them. The essence of sequence stratigraphy is the recognition of regional unconformities and conformities on a basin-scale level that bound sequences of genetically related strata (Emery and Myers 2001). Many of the concepts and principles of sequence stratigraphy are primarily based on the observation from seismic data that prograding basin-margin systems often have a consistent depositional geometry. The term clinoform is used to summarise this geometry, and can be subdivided into top-, fore and bottomsets (Figure 1.5). The clinoform profile results from the interplay between sediment supply, wave and tidal energy in the basin (Emery and Myers 2001). In order to understand the evolution of the clinoform through time it is necessary to consider the accommodation space. The rate of change of accommodation space is a (local) function of sea-level change, subsidence and sediment supply. Progradational geometries occur when sediment supply exceeds the rate of relative sea-level rise. When local sea-level is stable or falls, a relative low rate of sediment supply will also result in progradational clinoform architecture. Progradation is expressed as clinoforms that show basinward migration. Aggradation occurs when sediment supply and relative sea level rise are balanced and result in an architecture where clinoforms are stacked vertically. Retrogradational clinoform geometries occur when sediment supply is less then the rate of relative sea-level rise, as a result clinoform geometry will migrate landward. The clinoform geometries can be further related to depositional sequences, also know as system tracts, based on similarities in geometry, facies and depositional environments.
Figure 1.4. Grain size characteristics from a well in deltaic setting showing a coarsening upward sequence (after Van Wagoner et al. 1990), with the interpreted depositional environments.
Figure-1.5. Conceptual cross section of fluvial dominated delta, showing the clinoform geometry (after Kenyon and Turcotte 1985), with the interpreted depositional environments that are horizontally adjacent in stead of vertical as in Figure 1.4, this is in accordance with Walther’s Law (Middleton 1972).
Summarising, generally a prograding fluvio-deltaic system displays a clinoform geometry that consists of a subaerial and subaqueous part. The subaerial topset consists of low gradient fluvial and delta plain deposits, whereas subaqueous fore- and bottomsets consist of delta front and prodelta environments (Gilbert 1890; Kenyon and Turcotte 1985; Swenson et al. 2005). In addition Walther's Law states that vertical facies successions also occur laterally to one another (Middleton 1972) (Figure 1.4 and 1.5). Therefore the clinoform is a measure of the lateral variability of deposits and predicts that seaward deposits become progressively finer-grained. This concept suggests that it is possible to relate deltaic stratigraphy to sediment dispersal patterns and transport processes.
Apart from the nature of the fluvial input and the reworking of deposits, other factors influence delta formation, e.g. the geometry and nature of the receiving basin, the tectonic setting, the gradient of the shelf; changes in climate and (relative) sea-level. Controls on delta formation can be divided into autocyclic and allocyclic controls on deltaic development.
Autocyclic processes are intrabasinal in origin and are related to the behaviour of the depositional systems. This concerns the behaviour of a system that will always occur even if the allocyclic controls remain constant. In deltaic systems they include lobe switching and river avulsions, in which a distributary is abandoned for a hydraulically more favoured route (Stouthamer and Berendsen 2001). Autocyclic processes and specifically the timing of events remain hard to quantify.
Allocyclic processes are extrabasinal in origin and include eustasy, tectonics in the source area and receiving basin, climate and other factors (Watney et al. 1999). It can be difficult to separate autocyclic and allocyclic controls on stratigraphy in outcrops (Burns et al. 1997). However allocyclic processes tend to affect deltaic deposition more dramatically and produce laterally more extensive effects upon the stratigraphy (Bhattacharya and Walker 1992). The main focus in this study will be on identifying the influence of the allocyclic controls on the parasequence scale in fluvio deltaic settings.
Quantitative modelling of fluvio-deltaic Stratigraphy
Quantitative modelling of sedimentary geology has a history of over 40 years and has become indispensable for the industry and scientific communities. Numerous models, varying in applicability and functionality, have been presented over the years and are discussed in the excellent review of Paola (2000). He concluded that there are three main areas for development: more and extensive testing, a better understanding of non-linear and chaotic effects, and the construction of more accurate numerical models of long-term dynamic processes. Recently Slingerland (2005) noticed that these questions are still open and wondered why progress has been so slow. An important reason is that the long-term processes are still not fully comprehended (Tipper 1991).
directly, because they occur over time intervals that exceed historical records. Historical records of even the most basic data are usually limited to 100-150 years (if any data are available at all). Data needed to understand the long-term evolution of fluvio-deltaic systems must be derived from the geological record, which is by definition incomplete, arbitrary and open to discussion. This can be illustrated by examining the various scales on which depositional systems seem to act. Sadler (1982) showed that deposition rates measured at short periods (100-101 years) are commonly an order of magnitude higher than those derived from the stratigraphic record. This difference can be explained by comparing and combining the magnitude and frequency of rare events relative to normal conditions of deposition (Dott 1988). The time-ratio between these two is the recovery time for a given environment and determines the preservation potential of deposits, i.e. low-frequency / high-magnitude or low-magnitude / high-frequency deposits. As a result, the differences in deposition rates can be explained by the preservation potential of sediments, i.e., the probability that sediments will survive erosion due to high-magnitude low-frequency events. On the other hand, specific deposits in the sedimentary record can be present (or absent) when high magnitude events do not occur during a certain period (Storms 2002), making interpretations very difficult..
Models are defined as simplifications of reality. The most important part of modelling is a (self-) critical assessment of the applicability of a model at the temporal and spatial scale of the problem under investigation (a reality check). The modelling objective of the research presented in this thesis is to highlight the basic, long-term interplay between fluvial input and delta deposition and to accentuate the importance of fluctuations in sea-level change and sediment supply on deltaic stratigraphy. As such, the analysis is restricted to simple basins without waves and tides. A specific choice is made to use the model for experiments, by accentuating singular controls on deltaic development. Unravelling the effects of each factor can be accomplished in the field by studying systems that show a strong dominance of one factor over the others (Galloway 1975; Schumm 1991). By simulating this behaviour in experiments these concepts and the model itself can be reviewed.
Numerical experiments and field studies both contribute to a better understanding of long-term depositional processes. This thesis includes descriptive (data-driven) research, as well as numerical and stochastic modelling. Chapters 2 and 3 are devoted to geological descriptions of the Kura and Mahakam deltas. The Holocene history of these deltas has been reconstructed using data from various sources (field measurements and shallow seismics).
DEVELOPMENT OF THE KURA DELTA, AZERBAIJAN; A
RECORD OF HOLOCENE CASPIAN SEA-LEVEL CHANGES.
The interaction between (rapid) sea-level change and deltaic systems has mainly been examined in outcrop studies (Burns et al. 1997; Naish and Kamp 1997; Reynolds et al. 1996). The Kura delta presents the possibility to study the effects of rapid sea-level changes on active delta environments in a well constrained setting. The Kura delta is located along the southwestern shore of the Caspian Sea, Azerbaijan (Figure 2.1). It represents a fluvial-dominated delta according to Galloway’s (1975) classification, with redistribution of delta sediments through wave-action on the northern shore. Beside the fluvial and shallow marine processes, rapid sea level change has a strong influence on the formation of the Kura delta. The present day subaerial delta covers ~200 km2 of largely undeveloped arid lowland and shoreline swamps. This area is the result of the latest phase of delta development which started at the beginning of the 19th century (Mikhailov et al. 2003). Major human developments that have affected the delta dynamics in the last 50 years have been the building of the Mingechaur Reservoir, ca. 150 km upstream of Kura River mouth, and development of industrial fish and rice ponds in the delta plain.
Earlier studies of the Kura delta undertaken by Belyayev (1971), focussed on hydrology and delta growth over the last 200 years. These have recently been updated and expanded by Mikhailov (2003). Limited work has been done on the (late) Holocene development of the Kura delta, as well as on the determination of its depositional environments and lithofacies. The data on delta growth and the detailed hydrological data, combined with the work of Rychagov (1997), on the fluctuations of Caspian sea-level do provide a narrow constraint on the sedimentation models and interpretations of the Kura delta lithology.
* Chapter 2 is based on: R.M. Hoogendoorn, J.F. Boels, S.B. Kroonenberg, M.D. Simmons, E. Aliyeva, A.D.
Figure 2-1. A) Schematic overview of the Kura delta study area with locations of acquired field data. B) ASTER satellite image of the Kura delta (24 January 2004) C) location map including the bathymetry of the southwestern Caspian Sea (rectangle indicates where the Kura delta is located), major faults, syncline and anticline structures and oil and gas fields of the Lower Kura basin from Inan et al. (1997.). Inset shows location of the Kura basin in relation to the Caspian Sea.
fall and rise, during the last 200 years, in the development of the Kura delta. Whereas fluvial and marine processes are the primary forces affecting the formation and morphology in of most major deltas (Galloway 1975), the Kura delta has formed in response to a combination of fluvial processes and the rapid, high frequency sea-level changes in the Caspian Sea.
2.2 Regional setting2.2.1 Geologic setting
The modern Kura delta is located on the border between the Kura and South Caspian basins (Figure 2.1). The South Caspian basin is part of an active tectonic zone in which the Greater and Lesser Caucasus are being uplifted (Mitchell and Westaway 1999), while the Caspian seafloor subsides at a rate of 2.5 mm yr -1 (Inan et al. 1997). The Kura basin is situated in the eastern part of the depression between the Greater Caucasus to the North and the Lesser Caucasus to the South. Middle Jurassic volcanism together with shallow-marine Jurassic and Cretaceous sediments form the base of the succession in the Kura basin, which has been encountered at a depth of more than 8000 m in the Saatly ultra deep borehole in the centre of the Kura basin (Khain 1984; Khain and Shardanov 1952; Levin 1995). From Miocene times onwards, shallow-marine and deltaic sedimentation has been dominant. Major uplift occurred at the end of the Miocene as a result of underplating of the Transcaucasian micro continent under the European plate. Folding of the Kura basin sediments, and older units, into NW-SE oriented anticlinal structures took place mainly at the end of the Pliocene, leading to the development of numerous mud volcanoes, still active today. These mud volcanoes are unique geological features and give rise to significant gas, water, and oil seepages (Guliyev and Feizullayev 1997). A mud volcano is found several kilometres offshore, northeast of the Kura delta.
Periodic transgressions and regressions of the Caspian Sea since the Late Pleistocene changed the coastline configuration in the area of the present Kura Lowland where the Kura River formed a number of delta lobes. During significant transgressions, this lowland turned into an inland shallow water bay, in which ancient deltas of the Kura River were formed. The traces of several deltas can be found in the present topography of the lowland. Over the period of large-scale regressions, the delta of the Kura River protruded into the sea far more to the east of the modern delta (Mamedov 1997).
2.2.2 River system characteristics
Kura River is 1515 km and the total area of the catchment is 188 000 km2 (including the Araks River). The catchment occupies the greater part of the Lesser Caucasus and the south-eastern Greater Caucasus.
The Kura water discharge at the river mouth averaged around 17.1 km3 yr-1 (550 m3 sec-1) between 1938 to 1984 (Bousquet and Frenken 1997). The sediment (bedload and suspended load) of the Kura River upon entering the delta is predominately clay, silt and fine sand (<200 μm). The annual sediment volume reaching the delta averaged 11.3 *106 m3 yr-1 between 1967 and 1976 and from 1977 to 1986 the sediment volume dropped to 8.8 *106 m3 yr-1 (Aybulatov 2001; Mikhailov et al. 2003).
Figure 2-2. A) Part of the map of Europe of Joseph Scheda (1845). At the location of the Kura River mouth (centre of the square) there is no evidence of a subaerial body. To the south of the river mouth, an active delta body can be seen, consistent with the deltaic remains observed in the Landsat TM7 Satellite image (2001) of the Kura delta and the Caspian Sea (B). These deltaic remains south of the modern Kura delta indicate active channel switching of the Kura River.
2.2.3 The Caspian Sea
middle and southern parts. The northern part is a shallow shelf region reaching a maximum depth of about 10m. The middle and southern regions are deeper areas, separated by an east-west oriented underwater range near the Apsheron peninsula. The depth of the southern Caspian Sea is approximately 1025m and the shelf edge is located 20 to 40 km offshore of the present day Kura delta. The sea level of the landlocked Caspian basin, presently at approx. 27 m below GSL, fluctuates rapidly on several time scales, seasonal to centuries. The measured seasonal sea-level change is up to 0.4 m (Cazenave et al. 1997) while the maximum measured inter-annual Caspian Sea level change in the records has been 0.34 m yr-1. These fluctuations are a result of the interaction between differences in river discharge (predominantly the Volga River), evaporation, precipitation and water temperature (Kosarev and Yablonskaya 1994; Rodionov 1994)
Figure 2-3. A) Estimated Holocene sea-level fluctuations of the Caspian Sea (Rychagov 1993a; Rychagov 1993b; Rychagov 1997) and measured Caspian Sea-level fluctuation (Klige and Myagkov 1992).
1995, it rose at a rate of 15 cm yr-1 (Kaplin and Selivanov 1995). Numerous transgressions and regressions of the Caspian Sea have also occurred in the more distant past (Ignatov et al. 1993; Svitoch 1991). The Holocene sea-level history has been reconstructed from a marine terrace section along the Dagestan coast. Results from these studies show five transgressive phases that have been dated around 8000, 7000, 6000, 3000 and 200 BP (Rychagov 1993a; Rychagov 1993b; Rychagov 1997). The lowest documented sea level is estimated at 50 m below global sea level at the end of the Pleistocene or early Holocene (Mangyshlak regression). The Derbent regression, around 1500 BP, reached a minimum of at least -32 m. The highest level reached by the Caspian Sea during the Holocene is around -22 m, the elevation of the present delta apex.
2.2.4 Delta morphology
2.3 Delta sediments and stratigraphy2.3.1 Lithology
Figure 2-5. Lithologic profiles of selected cores, note that the vertical scale of well 4 (F) is in meters while the others are in centimeters
cores are 7-8 m deep. The piston cores were obtained offshore using a 3.5 m long, 10 cm wide piston corer profiler, and wells were drilled up to 20 metres deep, in 2 m sections.
A typical onshore core (Figure 2.5 A, B and C) consists of massive dark grey clays and silty sands at the base. These pass up into laminated clays and silty clays overlain by layered fine sand, silts and clays. These deposits are often intercalated with sandy beds, massive and heterolithic sands, which vary in thickness (10 - 100 cm.) and sorting. The massive sands are relatively poorly sorted, dark reddish brown in colour, very fine to medium silty/clayey sands, with a uniform grain size. The thickness of the massive sand layers varies between 10 cm and 1.3 m. The heterolithic sands are brownish and reddish and vary in grain size resulting in fining- or coarsening up sequences. They are well sorted with a thickness that varies between 10 cm and 0.5 m. The top of the cores consist of massive clays and silty clays (homogenous) which are rich with rootlets. The onshore cores located away from the main channel often lack sandy deposits. The cores towards the delta front feature a set of layered fine sand, silts and clays on top of the homogenous massive clays.
The offshore piston cores (3.5 m) consist of fine sediment (Figure 2.5 D and E). The distal piston cores are homogenous and consist primarily of laminated clays. The colour-laminated clays and silty clays are found with abundant mm- to cm-scaled colour transitions and their colour varies from yellowish brown to olive black. The continuous thickness of the laminated clays reaches 260 cm. In some cases, in proximal as well as distal cores, thin shelly, sandy deposits can be found. Sometimes complete shells occur within these coarser layers and vary in size from 1 mm to 5 cm. The locations proximal to the delta shoreline generally show laminated clays at the bottom of the proximal piston cores which are overlain by layered clays and silts. The layered clays and silts are sporadically intercalated with films of fine sand, the thickness of the sand and silt layers varies from 1 mm to 1 cm. There are also silty clays containing significant amounts of organic material forming black layers, 5 - 20 cm thick.
Table 2-1 Characteristics of the lithofacies of the Kura delta
Lithofacies Texture Observed sedimentary characteristics
Location Massive clays and Silts Silty Clay and Clay Massive, roots, desiccation cracks Onshore Massive Sands Medium to Very Fine
Sand Massive, sharp boundaries Onshore Heterolithic Sands Medium to Very Fine
Sand, Silty Sand Coarsening up or fining up, badly sorted Onshore Interstratified Clays, Silts
and Sands Fine to Very Fine Sand, Silt and Clay Layers and lamination Onshore and Offshore Laminated Clays and Silts Silt and Clay Lamination, Mud dominated Onshore and
Offshore Dark Grey Clays Clay and Medium to
Very Fine Sand Dark Grey colour, Well sorted layers, Mud dominated Onshore and Offshore Shelly Sands Medium to fine sand Shells Offshore Cemented Shells Shells fragmented and
cemented 100% shells Offshore
Figure 2-6 Depositional environments of the modern Kura delta
2.3.2 Depositional environments
Satellite images, field observations and surficial sediments were used to classify the delta into several depositional environments. Figure 2.6 illustrates the spatial distribution of these depositional environments.
subaerial deposits formed by fluvial processes. The mottled clays which have been observed in the lower portion of the well cores have also been related to a lower deltaplain depositional environment. Sandy sediments on the subaerial delta surface were only found at the bifurcation of the northern and southern distributaries, forming a point bar on the inside of the river bend. Sandy deposits in the subsurface samples of the delta plain are represented by massive and heterolithic sands locally deposited in higher energy environments in the vicinity of the distributaries, e.g. channel fills, crevasses, or levees. However, the poor sorting, ranging from fine sands to clays, indicate an environment where flow energy is sporadically high enough to deposit such a mixture, but not continuously high enough to effectively sort the material.
The proximal delta front comprises the southern, low angle dipping seafloor, and that part of the lower floodplain that is occasionally submerged. Deposits of the proximal delta front environment comprise layered clays, silts and fine sands. Depending on the local topography and hydrodynamics of the channels, parts of the delta front deposits may contain organic material, representing the marshy freshwater environment, representing a transition zone between the delta plain and the proximal delta front. The distal delta front is the part of the delta comprising a high angle slope varying between 0.3 – 0.5º. The laminated clays are found here. The distal delta front gradually changes into the prodelta where sedimentation rates are low.
Between the prograding northern distributaries an interdistributary bay has formed, ca. 0.5 to 1 m deep, in which clay and silt has been deposited during floods. This bay was dry during the last lowstand (1977) and is currently submerged and overgrown with aquatic vegetation.
Mouth bar deposits occur seaward of the river mouth (South). This facies is characterized in the sub surface samples by laminated sands, and interbedded sands and muds (85% sand).
Figure 2-7. Sparker profile 5 (0105) showing the down lap of the modern delta to the NE, and the ‘drape’, comprising several, parallel reflectors in the middle of H2 and clinoform stacking in the southern part of the profile indicating the aggrading phase of H2.
2.3.3 Sparker data
Figure 2-8 Sparker profile 2 (0102) coast-parallel profile, illustrating the two channel types the most S-SE channel (right-hand side of the figure) is horizontally filled with sediments and incised the underlying strata. The most N-NW channel (left-hand side of the figure) serves as an example for an aggradation channel since it is still visible in the surface topography.
Five sparker profiles (Figures 2.7 to 2.11) show the typical features of the Kura delta. Four main features can be recognized within the profiles: 1) horizontal / subhorizontal reflectors (deltaplain), 2) clinoform reflectors (delta front, prodelta) 3) concave-upward reflectors (distributary channels and possible incised channel), which are often associated with 4) hyperbolic reflectors (levees, barrier).
The horizontal / subhorizontal reflectors represent the topset facies. The stratigraphic position and the parallel character of these reflectors implies vertical aggradation in a deltaplain depositional setting (Figures 2.7, 2.8 and 2.9). The cores which penetrate these reflectors show layered clays, silty clays and fine sands that are interpreted as proximal delta front deposits. Typical palaeo-floodplain deposits are only found at the base in wells 4 and 5. Other subhorizontal reflectors are interpreted to represent the mud volcano dynamics.
Figure 2-9. Sparker profile 11 (0111) Overview of the northern offshore part, with the sediment drape of the modern delta extending to the slope of the mud volcano. The data also shows the subhorizontal and clinoform reflectors representing the progradational phase of H2. Furthermore the transgressive surface 2 (TS2) can be correlated to a facies change in well 3, TS1 is correlated to the deeper sandy shelly horizon of well 3, which were dated at ca.1400 BP (Figure 12).
The concave-upward reflectors are mainly associated with the topset deposits of Figure 2.8. Two channel types can be recognized. 1) Channel type 1 incises and fills horizontally. The incised channel is associated with a regressive system as it is filled up when the sea level rises. 2) Channel type 2 aggrades vertically, this channel type is associated with a transgressive phase of delta development. When the regressive channel is suffocated, other, smaller, channels develop. These channel features are preserved under a drape of sediment, in a similar way to the processes observed in the modern delta.
Figure 2-11. Sparker profile 14 (0114) in line with the progradational direction of the modern delta shows a high seafloor gradient (0.5°) and also depth-related erosive features. Steep clinoform features of the H2 phase appear at the slope.
2.4 Laboratory analyses2.4.1 Radiometric Dating
Figure 2-12 Lithologic profiles with summary of all age data (based on 210Pb, 14C, Biostratigraphic
and diatom data) and interpreted depositional environment. Note that the vertical scale of well 3 is in meters while the others are in centimeters. In well 3 the datum level of the interpreted TS2 of Figure 9 is shown.
Table 2-2. Overview of the 14C analyses and biostratigraphic results (Caspian reservoir age is ca.
290 yr, K. van den Borg personal Comment).
sample depth (cm) material years BP cal years Well3#3 1050-1060 Dreissena polymorpha/andrussovi *1844+/-32 1409-1335 Well3#4 1530-1550 Dreissena polymorpha/andrussovi *2829+/-33 2674-2539 Well3#5(1) 1600-1630 Dreissena polymorpha/andrussovi fresh 1368+/-36 947-888 Well3#5(2) 1600-1630 Dreissena polymorpha/andrussovi old *1914+/-32 1495-1420 Well3#6(1) 1710-1715 Dreissena polymorpha/andrussovi fresh 1414+/-37 984-918 Well3#6(2) 1710-1715 Dreissena polymorpha/andrussovi old 1443+/-29 1009-944 Site depth (cm) Invasive species Year
Piston 7 244-246 Balanus improvisus 1954 AD Piston 5 109-111 Balanus improvisus 1954 AD Well 2 305-310 Abra ovata 1939 AD Well 3 920-925 Mytilaster lineatus 1920 AD
2.4.2 Biostratigraphic dating and diatom analysis
Biostratigraphic analysis of the shells resulted in the recognition of invasive species (Table 2.2). The timing of first occurrence of invasive species in the Caspian Sea is well known (Kosarev and Yablonskaya 1994), and therefore can be used to date historic deposits. Mytilaster lineatus invaded the Caspian Sea around 1920-1930, attached to ships coming from the Black/Azov Sea region. Abra ovata was deliberately introduced in order to raise biological production and consequently fish productivity in 1939. The barnacle Balanus improvisus was introduced with the opening of the Volga-Don Canal in 1954. The age estimates and subsequent average sedimentation rates derived from lowest occurring depth of invasive species are given in table 2.2. Piston cores 7 and 9 show an average rate of approx. 2.3 cm yr-1. This is in agreement with the results from the 210Pb analysis.
A total of 10 offshore sediment samples were analysed for diatom content. Of these, 8 were found to contain diatoms, with 5 containing sufficient numbers to allow counting and detailed environmental interpretation. The samples typically contained a proportion of inorganic material, including quartz and mica, and diatom recovery from these samples was low. Recovery of diatoms tended to be highest in samples with a high proportion of clay. Table 2.2 and Figure 2.12 show the results of the diatom analysis from the piston core samples, and the interpretation of their depositional environment. Although there were limited data the diatom analyses of layered silts and clays from the piston cores indicate that the depositional environment is related to a delta front, confirming the sedimentological interpretations of the depositional settings.
2.4.3 Grain-size distribution of modern sandy surface sediments
the N-S barrier which formed the coast before being breached by the Kura 200 years ago. Location is situated, approx. 500 m N of the present apex, in a sandpit. Site B is located on the beach along the north shore of the delta, ranging in width from 50 to 400 m. Site C is an active point bar at the inside of the channel bend at the bifurcation of the Kura River. Sixty-four undisturbed samples were collected in plastic tubes (35 mm photo film cylinders) by carefully pushing the tubes into the unconsolidated sediments. All samples were analysed in duplicate using laser grain particle analysis (Konert 1997) at the Free University, Amsterdam (VU). The principal parameters of mean, standard deviation, median and skewness were derived from the data and presented in Table 2.3.
Table 2-3. Overview of average texture parameters per sample location divide into subsets based on sample location Please provide a caption for this table
Location Grain-size classification Mean (phi) Mean (mm) Median (phi) Median (mm) No.of Samples Site A 1 Medium sand 2.60 0.165 2.36 0.195 26 Site A 2 Fine Sand 3.20 0.109 1.91 0.267 16 Site B I Fine Sand 3.26 0.105 2.80 0.144 42 Site B II Fine Sand 2.95 0.129 1.83 0.282 20 Site C Very fine sand 3.50 0.088 1.75 0.298 24
The samples of the three sites reflect time sequence and sediment source information. (Miall 1996). Beach deposits are exposed to two forces of unequal strength acting in opposite directions, the incoming waves and outgoing wash. Of which the outgoing wash will transport only the fine-grained particles in a seaward direction. Subsequently the grain size distribution of winnowed sand lacks the tail at the fine-grained end of the distribution. Winnowing also tend to increase the sorting of the sediments. So, if the provenance of the beach sands of site A and B are Kura River sands then the results of winnowing should show in the grain size distributions in relation to the grain size distribution of the samples of site C. The results of the grains size analysis shown in Figure 2.13 suggest such a relation between the samples. Where an increasing mean is associated with a better sorting and associated with a shallow marine environment. And secondly there is a separation between the beach and the river sands for the two fluvial and shallow marine depositional environments. The analysed river sands are more evenly distributed, and the analysed beach sands are, generally, negatively skewed.
Figure 2-13. Three typical grain size histograms of Kura delta sediments representing the fine-grained sands of the three locations where sandy deposits were found on the surface of the delta.
Based on the grain size information we can now construct a model for the development and source of the beach sands. Despite the fact that the Kura River mainly transports fine sediments there is an influx to the delta of coarser sediment. Material found in the point bar at site C proves this hypothesis. Therefore we assume that Kura River sand is transported to the river mouth where it is deposited and reworked by wave actions forming the barriers. If Kura River sediments build up the barriers at site B, the sediments of site B and C should be comparable. The trend shown in Figure 2.13 suggests that the two samples of the two sites are associated with each other. The differences in grain size properties can be explained as the effects of winnowing. That resulted in sorting of the sediments and a decrease in skewness. The old barrier (site A) is even better sorted and more skewed due to a probable longer exposure to wave action. Although the spatial onset of the present delta is not related to the old barrier, it is possible that the relation with the river sands and the old barrier is similar to the relation of site B and C. Therefore we assume that the both barriers are a result of the influx of river sands transported by the Kura River.
2.5 Depositional history
exceptional rate of change of the Caspian sea-level restrains us from using sequence stratigraphic terms, although similarities between the deposits and systems tracts will be mentioned.
Figure 2-14. Geochronological representation of the different phases of delta development superimposed on the Holocene Caspian sea-level curve (Rychagov, 1997). Giving close constraint to the interpretation of the (late) Holocene deposits of Kura delta.
In order to be able to refer certain stratigraphic features to former sea levels, all datum levels are stated in absolute values calculated as h= d+w+z, in which d is the depth of the feature in the well or piston core, w is the water depth at the top of the well, and z the datum level of the sea in 2001 with respect to the Kronshtadt gauge in the Baltic (-27 m). A schematic overview of the geochronology for the Kura delta deposits in relation to the Holocene Caspian sea-level curve (Rychagov 1997) is shown in Figure 2.13. The overall depositional patterns, calculated datum levels and 14C datings combined show 4 phases of deposition, one pre-Holocene and three late Holocene phases. These phases are characterised by different stages of deltaic deposition associated with erosive marine surfaces, which are interpreted to represent cessation of the sediment supply and transgression.
Phase 1, Pre-Holocene deposits (PH, TS1).
deposits occur on top of the reddish clays at absolute depths between -83 m and -76 m which are associated with shoreface environments, Sparker profile 7 (Figure 2.10) shows a reflector at this level. The results from Rychaokov (1997), show that, after the lowstand, at the Pleistocene-Holocene transition, a transgression occurred. TS1 reflector is interpreted as a marine erosion surface formed during this transgression. Since well recovery is very poor, the TS1 reflector only occurs in one profile and no biostratigraphic information is available for this phase, the position of these deposits in the overall stratigraphy remains inconclusive.
Phase 2, late Holocene deposits 1, (H1, TS2).
The second phase consists of sedimentation of the unit underlying the Transgressive Surface indicated by TS2 in the sparker profiles. The seismic facies within this unit is indistinct. The wells that intersect the TS2 continue 10-14 m down into this sedimentary unit. The bottom of the unit is unknown except for the TS1 reflector of sparker profile 7. The unit consists mainly of layered silty clays, with minor intervals of laminated clays and silts, and, in Well 3 (Figure 2.9), several shell-rich horizons. Microfauna in Well 2 indicate a decreasing depositional depth (with some fluctuations). Depositional depth in Well 3 fluctuates between 10 and 15 m, in harmony with the actual water depth of 11.4 m. In the uppermost part of the unit the ostracod Iliocypris brady was found, indicative of fresh-water influence. Together these data indicate a generally falling sea level during deposition. Six 14C datings were obtained from the shell-rich horizons in Well 3, located underneath the boundary between H1 and TS2, and indicated deposition at ca. 1400 BP. Therefore the H1 deposits are thought to be associated with the forced regression preceding the Derbent lowstand of 1500 BP (Rychagov 1997). Since the Derbent regression did not start before 3500 BP and no other depositional units were found between TS1 and TS2 it is probable that no deposition took place at the study site during the early-Holocene.
The TS2 is a prominent reflector in the sparker sections, especially NE and E of the present delta. The surface is highly irregular in shape between absolute depths of 45 and 60 m, and truncates H1 sediments. Below that it slopes smoothly down to 75 m, the deepest level it has been recognised (Figure 2.11), and parallel to the stratification of the underlying unit. The irregular topography of the reflector indicates either an erosive origin, or an accumulative origin as overstepped barriers, or both, but in any case features that occur in a coastal to onshore setting. The shell horizon recovered in Piston Core 5 at an absolute depth of -47 m can also be related to this phase. This is the deepest appearance of TS2 in the sampled data and suggests that this may be a lowstand. On the basis of 14C datings this lowstand must have taken place around 1400 yr BP. While a lowstand of -34 m is inferred for the Derbent regression (Rodionov 1994; Rychagov 1997; Varushchenko et al. 1987), our data suggests that the lowstand fell, to an estimated depth of -37 to -42 m when it is assumed that the shells from Piston Core 5 were deposited at a water depth of 5-10 m. Hence this surface is interpreted to be associated with the Derbent lowstand and the subsequent transgression.
Phase 3 late Holocene deposits 2 (H2, TS3)
consists of latterly varying facies that are syndepositional. Proximally, organic-rich silty clay was deposited in a delta front environment. A prograding deltaic sequence with clinoform-shaped reflectors can be seen on the distal, more seaward side, shown in profile 5 and 11 (Figure 2.7 and 2.9). Furthermore this phase is found at the base of the onshore cores and consists of massive dark grey clays and silty sands, similar to the modern delta plain deposits. The depositional depth indicated by the microfauna in Well 2 first increases and, then decreases back to its initial level of 25 m. In Well 3 the depositional depth is uniformly about 15 m. Both figures are similar to the present water depth of -26.3 and -11.4 respectively. Organic-rich clays at -38 m absolute depth in Piston core 7 contain fresh water diatoms, and have higher vegetal organic compounds than organic clays from organic clays in piston cores sampled in deeper water. This suggests that also this unit reflects an overall falling sea level. The aggrading stacking patterns of sparker profile 5 (Figure 2.7) indicate a transition from regression to transgression during this phase, which suits the definition of a Regressive Systems Tract. In view of the position of the deposits between the two transgressive surfaces the age of phase 3 is probably between the 11th and 16th century AD when several alternating stages of rapid regression and rapid transgression occurred (Rychagov 1997).
The TS3 reflector truncates the H2 unit with an irregular topography with ridges, benches and depressions, especially between -47 and -57 m absolute depth. The reflector is smooth at -37 m depth along the shallow SW part of the delta. The age of the Transgressive Surface can only be established indirectly, since datings from this unit are not available. The overlying unit is known to have been deposited from the start of the 19th century onward following the 200 BP highstand (Rychagov 1997). During the period preceding this highstand major barrier complexes were formed (Storms 2002). Because at that time the Kura River did not discharge at its present location, but much further south, in the Qızılağaç bay, the barrier complex at the apex of the present-day delta and subsequently TS3 were formed during the 16thand 17th century (Mikhailov et al. 2003).
Phase 4 modern delta (H3)
The history of the Kura delta can be traced back only for a relative small time interval in this study. The present position of the delta corresponds with the axis of the rapidly subsiding Kura basin (Khain and Shardanov 1952) and it is also situated at the head of a prominent submarine valley, suggesting a period of deep incision in its early history. The bulge shape of the submarine delta suggests a cumulated thickness of at least 50 m. All these facts suggest that its history must go back much further than can be retrieved from our data. With respect to the available data, the historic and newly collected data form only a part, although the network of core- and sparker data compliment one another and clearly characterize the Kura delta
For a better understanding of the significance of the development of the late Holocene Kura delta it is useful to compare it to other deltas, as many sedimentological investigations of modern fluvial-dominated deltas have concentrated on the Mississippi delta (Coleman et al. 1998; Fisk 1961; Frazier 1967; Gould 1970; Roberts 1997), it is logical to use it as a reference point. A number of similarities exist between the Mississippi- and the Kura deltas. 1) All sediment is concentrated in a single channel with a limited number of outlets. 2) Both deltas build out on their own unconsolidated sediments. 3) Present day delta fronts are being (partly) redistributed by waves. 4) Different phases of delta development can be recognized. However, some differences in the sequence of events leading to the above mentioned analogues are also recognized. Primarily, a scale difference, both spatial and temporal is evident. The Mississippi delta is bigger in all aspects, water and sediment discharge, delta surface and delta volume and has been at its current location for at least 2000 years (Coleman et al. 1998), whereas the current Kura delta has switched at least 4 times during the same period. Secondly, the Mississippi delta started to develop after a major avulsion has occurred (Törnqvist et al. 1996). Avulsions are caused by decrease of the river gradient as a result of several processes that interact, such as subsidence, rise of floodplain lake levels or relative sea-level rise (Overeem et al. 2003). The sequence of events in the Kura delta seems to be different: progradation starts as a result of sea-level fall and the subsequent sea-level rise causes aggradation and eventually a switch of the delta lobe. Despite the differences the Kura delta could be described as a “baby birdfoot” delta (Figure 2.1) based on the similar morphological development patterns as seen in the Mississippi delta.
environment (Mikhailov 1997). So despite the major influence of the Caspian sea-level on these deltas, they all evolved differently. The general shelf edge setting bathymetry of the Kura delta (Figure 2.1), and the minor influence of waves, make the Kura delta more comparable with other delta settings. It is therefore the better suited as a natural laboratory to test conceptual models of sea-level change in deltas.
The early Pliocene Productive Series in Azerbaijan consist of fluvial deltaic sediments deposited in the isolated South Caspian Basin by several large river systems, which were also subjected to an unstable sea-level regime (Hinds et al. 2001; Reynolds et al. 1996). Many offshore and onshore hydrocarbon occurrences are in this unit (Aliyeva 1988; Bagirov and Lerche 1998). The Productive Series in the southwest of the South Caspian Basin has volcanogenic heavy mineral assemblage, indicating provenance form the Kura River, which rains Jurassic and Cenozoic volcanic deposits in the Lesser Caucasus (Pashaly 1964). Despite similarities with regards to depositional setting of the early Pliocene Productive Series, this study shows that any comparison between the Kura delta and the Productive Series is difficult because of the differences in lithology. Except for the thin and narrow sand bodies in the channels and barriers of the onshore plain, the whole late Holocene Kura delta consists of clays and silts, while the Productive Series is characterized by the presence of large amounts of sand (Reynolds et al. 1996). Several reasons can be put forward for the absence of sand in the present delta. The Kura River occupies the axis of a very rapidly subsiding basin. The strongest subsidence in the past has occurred not close to the coast, but ca. 100 km inland near Kurdamir (Inan et al. 1997). Here part of the sand may be trapped before it reaches the coast.
Figure 2-15. (A), interpretation of the main (D-AV) onshore core section through the Kura delta. Vertical dashed lines approximate the Caspian sea-level and the delta front location at the time indicated. The delta front moves along stream constructing a progradational delta sequence during the relatively stable sea level (1800–1933) and during the level fall (1933–1977). The major sea-level rise (1977–present day) can be recognized as an overlap on top of the progradational sequence near the delta front. The aggradational sequence is confined to the tip of the delta and along the southern shoreline. (B) Schematic summary of the Kura delta development, in order to illustrate all phases/stages of delta deposition and marine erosion.
The interpretation of the late Holocene Kura delta development can be summarised in seven stages (Figure 14A and B):
Stage 1 (PH): Pre-Holocene fluvial sedimentation (probably top of a lowstand deposits) identified by reddish soil in the delta plain sediments recovered from deep well data associated with the Mangyshlak regression based on depositional depth.
Stage 2 (TS1): PH was followed by formation of a Transgressive Surface after the Mangyshlak lowstand, with no sediment transported to this location by the Kura River.
Stage 3 (H1): Reactivation of sediment supply resulting in progradational deltaic deposition of a shallowing-upwards sequence of clays, silts and shell horizons deposited during the Derbent regression (before 1500 BP), Depositional depths at the start of the regression were comparable to the present-day and estimated at a maximum 42m below GSL at the end of this stage (forced regressive deposits).
Stage 4 (TS2): The absence of sediment supply and transgression, following the Derbent lowstand (maximal, ca. -42m absolute depth, 1500 yr BP), resulted in a period of marine erosion. Identified in the sparker profiles as Transgressive Surface 2 (TS2).
Stage 6 (TS3): A next phase of no sediment supply and transgression resulting in an erosive surface, possibly related to a 16th century lowstand and following transgression ending in the highstand of 200 BP. This stage probably related to the 17th-19th century, when the Kura river was diverted southwards to the Qızılağaç bay.
Stage 7 (H3; modern delta): Deltaic sedimentation resumed at the present-day position from the start of the 19th century onwards, depositing a series of prograding sandy to clayey bodies in the present delta plain, and a veneer of clayey and silty sediments offshore on top of the last erosional discontinuity (TS3). The last 1929-2000 sea-level cycle is expressed onshore by progradation during base level fall, and aggradation due to flooding of the delta plain during sea-level rise in most recent times. This single sea level cycle can be distinguished in the cross section of figure 14A
The Kura delta evolution shows cyclic behaviour; a progradational delta body is formed during 3 regressions at or near the present location while during transgression the delta body probably shifts to the Qızılağac Bay. The resulting erosional phases at the present location are good markers for the Caspian Sea lowstand. Therefore it can be concluded that the major control of the Kura delta is the rapid sea-level change of the Caspian Sea as well as the Kura River dynamics.