Search

Combining field reconstruction and landscape evolution modelling can be useful to investigate the relative role of different drivers (tectonics, climate, local base level) on long term catchment evolution. In this thesis, field reconstruction and landscape evolution modelling are combined to unravel the long-term (300 ka) response to lava damming events of evolution of the Geren Catchment, a tributary of the upper Gediz river near Kula, Western Turkey. This catchment was considered suitable for such a study because its high preservationof remnant landscape surfaces and fluvial terraces which could be dated, while its base level evolution could be reconstructed by identifying and dating lava flows. In Chapter 2, landscape evolution modelling of an idealized catchment revealed long-term (15 ka) catchment response to natural damming. Evolution of a high erodible and low erodible landscape was simulated using landscape evolution model (LEM) LAPSUS (LandscApe procesS modelling at mUlti dimensions and Scales). The natural dam was given four different erodibilities, to mimic both the potentially more erodible landslide dams and resistant lava dams. In a low erodible landscape damming led to persistent preservation of the sediment wedge formed behind the dam, while in a high erodible landscape, damming additionally led to knickpoint persistence, hampered incision of the main river and stream rerouting. The highest erodible dam was almost removed after 15 ka, while its sediment wedge was still partly present. Comparison of results with natural dam events from literature showed that modelled response characteristics are observed in actual situations and that simulations on Quaternary timescales are useful. In Chapter 3, field reconstruction resulted in a young lava flow being age constrained to the late Holocene (3.0 – 2.6 ka), by luminescence dating of fluvial sands below and on top of the flow. This lava flow dammed the Gediz river at two locations. the upstream lake was silted, while the downstream lake was not. Dams were breached catastrophically and possibly in a cascading event. The Gediz created an epigenetic gorge and its current river bed is still not at its pre-lava flow level. Results are summarized in a conceptual diagram. Furthermore, field reconstruction and 40Ar/39Ar dating revealed multiple lava dam events which have infrequently raised and lowered the base level of the Geren Catchment in the middle to late Pleistocene (311 – 175 ka). Sediment-capped palaeosurfaces in the Geren suggest change from an active fluvial system to a more lacustrine environment in the middle Pleistocene, followed by fluvial reactivation and stepped incision in the late Pleistocene. A second landscape evolution modelling study was conducted in Chapter 4, on a 300 ka timescale, with a larger catchment. Four scenarios have been applied on a reconstructed paleodem of the Geren Catchment. In the first scenario, the palaeodem was given constant rainfall for 300 ka. In the second scenario, three short (1 ka) damming events were added at its catchment outlet. In the third scenario, the palaeodem endured gradual base level lowering at its outlet, based on the known incision rate of its base level, the Gediz river. In the fourth scenario, base level lowering and damming events were combined. Results were interpreted by evaluating 1 ka-averaged net erosion, catchment sediment storage, longitudinal profile development and spatial differences in net erosion and sediment storage. Results showed that the net erosion signal of the catchment is complex in all cases. However, average net erosion and its variability increased due to constant base level lowering. Additionally, alternating phases of high and low variability occurred in net erosion, where high variability coincided with a strong decrease in total catchment sediment storage. Adding damming events to the gradual base level lowering scenario generated similar average net erosion as the base level lowering scenario, however its temporal pattern showed significantly different alternation of high and low variability periods. Furthermore, dampened upstream erosion was observed. Over time, this dampening migrates upstream indicating a long-term legacy of short term dam events. Field reconstruction and landscape evolution modelling were combined in Chapter 5, to be able to reconstruct and understand actual Geren catchment response to identified base level evolution over a 300 ka period. In all simulations, rainfall and vegetation are varied over time based on arboral pollen. Because exact significance and duration of dam events were not known, three scenarios of landscape evolution in the Geren Catchment were investigated: i) uplift driven gradual base level lowering, ii) gradual base level lowering and short damming events and iii) gradual base level lowering and long damming events. Output was evaluated for erosion-aggradation evolution in trunk gullies at two different distances from the catchment outlet. Climate influences erosion – aggradation activity in the upstream reach, although internal feedbacks influence timing and magnitude. Scenario i shows the most correlation with the climate signal, although its correlation is weak. Lava damming events leave an aggradation signal in the downstream reach, while complex and lagged response to these dams obscure correlations with climate and leave a legacy of the past in current landscape evolution. Catchment response of the long dam scenario correspond best with field reconstruction and dating. The combination of climate and base level explains a significant part of the landscape evolution history of the Geren Catchment. In Chapter 6, a reflection and synthesis of Chapters 2-5 is presented. Indications for response to tectonics, climate and damming events are discussed separately for both field and modelling results. It is concluded that (lava) damming events of Pleistocene age can hamper, but also enhance incision on a 300 ka timescale. Furthermore, they can still have effect on current and future catchment evolution. However, catchment response to this evolution is complex and catchment specific and model results do not exactly reproduce its catchment history. An aggregated landscape evolution model output such as stream bed elevation change can be useful for comparison with fluvial terrace sequences. Combining field reconstruction and modelling suggests that the 300 ka incision history of the Geren is best explained if the catchment endured prolonged dam events. The combination of field reconstruction, dating and landscape evolution modelling therefore can enhance our understanding of long-term evolution of a specific landscape and increases knowledge on long term impact of past events on current catchment complexities and it is suggested to embed this research approach more structurally in long-term landscape reconstructions.

This research looks at landscape dynamics – erosion and deposition – from two different perspectives: long-term landscape evolution over millennial timescales on the one hand and short-term event-based erosion and deposition at the other hand. For the first, landscape evolution models (LEMs) are often used, which describe landscape forming processes by geomorphic transport laws, usually on annual temporal resolutions. LEM LAPSUS is used in this research to evaluate the landscape dynamics in a study area in south-east Spain: the Guadalentín Basin. The model is calibrated on dated river terrace levels, which show an erosion – deposition – erosion sequence that the model could reproduce. Annual precipitation in this dryland area shows large inter-annual variability and erosion is supposed to be mainly the results of low-frequency, high magnitude rainfall events. Therefore, in this research, landscape dynamics are also assessed using the event-based erosion model OpenLISEM. Eventually, the role of extreme events in long-term landscape evolution are explored by comparing the two models and by incorporating annual rainfall variability into LEM LAPSUS. Another issue that is being addressed in this study is the relative influence of humans as compared to erosion as a natural process. A conceptual model, derived on the basis of dated sediment archives, is tentatively correlated to periods of human impact on the land. Using LAPSUS, the potential influence of historical tillage erosion is simulated, showing that the relatively slow process of tillage erosion added to floodplain aggradation over thousands of years.