Title: Sediment yield from European catchments: exogenic, endogenic and anthropogenic controls.
Other Titles: Sediment export van Europese riviersystemen: exogene, endogene en anthropogene factoren.
Authors: Vanmaercke, Matthias
Issue Date: 7-May-2012
Abstract: This study aims at better understanding the factors that control annual catchment sediment yield (SY, [t km-2 y-1]) at the intra-continental scale of Europe. Special attention is given to 1) the inter-annual variability of SY and its effect on the reliability of average SY-values; 2) the potential control of seismic activity on SY; and 3) the factors determining the human impact on SY. Specific objectives and the hypothesis behind these objectives are discussed in chapter 1.As a first step, a database was compiled based on an extensive literature review and through contacting various institutes. The resulting database consists of SY-measurements from 1,794 catchments throughout Europe, representing a total of 29,203 catchment-years of observations. Most of these data were not published in international scientific journals and were measured before 1990. A description of the database and a discussion of potential sources of errors on the measured SY-data is provided in chapter 2. Based on SY time series for 726 catchments, the inter-annual variability of SY is explored in Chapter 3. Apart from a relatively weak climate effect, no factors could be identified that control this variability. As a result, the inter-annual variability of SY and its resulting uncertainties on average SY-values remain difficult to predict for specific catchments. Nonetheless, the large dataset of time-series allowed for a general assessment of the likely ranges of uncertainty on average SY-values associated with their inter-annual variability. Monte Carlo simulations indicated that this uncertainty can be very large (&gt; 1 order of magnitude) when the measuring period (MP) is short (< 5 y). Furthermore, average SY values based on short MPs have a larger probability to underestimate rather than to overestimate the long-term mean. For instance, the SY-value of a median catchment after a 1-year MP has a 50% probability of underestimating the long-term mean by about 22%. Uncertainties quickly decrease after the first few years of measurement but can remain considerable, even after 50 years of monitoring.Despite these important uncertainties, the large constructed database of SY-measurements allowed for a general exploration of the spatial patterns and scale-dependency of SY in Europe (Chapter 4). Clear regional patterns in SY were identified: northern and western Europe are generally characterized by low SY-values (< 40 t km-2 y-1), while southern Europe and most of Europe’s mountain ranges are characterized by high SY-values (&gt; 200 t km-2 y-1). Catchment Area (A, km²) – SY relationships were calculated for various regions within Europe but generally revealed very weak correlations. This concurs with the findings of previous studies indicating that other factors controlling SY may be of larger importance and that catchment area alone cannot be used to predict SY. Nonetheless, observed differences in A-SY relationships indicate some important regional differences in scale-dependency of SY. Whereas lowland and temperate regions are generally characterized by weak but significantly negative A-SY relationships, this is not the case for Mediterranean and mountainous regions. Previous studies suggested that negative A-SY relationships mainly occur in regions where sheet and rill erosion is the dominant sediment source, while none-negative A-SY relationships may indicate that other sediment sources (e.g. landslides, river bank erosion, gully erosion) are probably more important. A comparison between SY and sheet and rill erosion rates at the plot scale supported this hypothesis (Chapter 5). It was found that, despite the fact that eroded sediments may be deposited again before reaching the catchment outlet, SY-values are generally larger than modelled sheet and rill erosion rates. This was especially the case for the Mediterranean and mountain regions, but also for other regions throughout Europe. To obtain a more thorough insight into the factors controlling SY and its sensitivity to human impact, detailed analyses were made for a subset of 146 catchments which were not significantly impacted by glaciers, lakes or anthropogenic land use changes (Chapter 6). It was found that the spatial variability in SY of these catchments is mainly explained by differences in topography (average catchments slope), catchment lithology and the degree of seismic activity in a catchment. Although previous SY-models implicitly assumed that tectonic controls on SY are reflected in the catchment topography, our results show that this assumption does not hold. Even after controlling for topography, lithology or other factors, relatively small differences in seismic activity explain a significant part of the observed spatial variation of SY in Europe. Its effects should therefore be explicitly considered when simulating SY at the regional or continental scale. This is not only the case for undisturbed catchments: also for a wide range of land uses and environmental conditions at the global scale, a clear seismic control over SY could be detected.Building on these analyses, a simple regression model was constructed that is able to simulate the baseline SY of a catchment, based on its topography, lithology and estimated degree of seismic activity. This model accounted for 56% of the observed variation in SY from the considered undisturbed catchments. By applying this model to (disturbed) catchments for which the contemporary SY was measured, the degree of human impact on SY could be estimated (Chapter 7). It was found that this impact is strongly and positively correlated with the fraction of arable land in a catchment and negatively correlated with the catchment area. As a result, the contemporary SY of small agricultural catchments (< 10 km²) may be several orders of magnitude larger than their baseline SY. However, human impact is much less apparent in larger (&gt; 100 km²) catchments. Differences in climate, topography, land use history or lithology explained little of the variation of human impact on SY for different catchments. This may be partly attributed to the uncertainties involved in estimated degrees of human impact or the data used to describe the potential controlling factors. However it also indicates that these factors are only of secundary importance compared to catchment area and the contemporary land use. The findings of this research have important implications for our understanding of land degradation and desertification. These are discussed in chapter 8. Most studies assessing the link between erosion and desertification mainly focussed at the (on-site) plot scale. However, this involves the risk of underestimating the true extent of erosion problems at the catchment scale and neglects many of the off-site consequences of erosion (e.g. reservoir sedimentation, flooding). Furthermore, also considering the catchment scale may provide valuable additional insights in the actual factors controlling erosion. It is therefore recommended that SY is included in desertification assessment studies, in combination with other indicators.A synthesis of our findings and scope for further research is given in chapter 9.
Publication status: published
KU Leuven publication type: TH
Appears in Collections:Division of Geography & Tourism

Files in This Item:
File Status SizeFormat
Compiled_WithTOC.pdf Published 5191KbAdobe PDFView/Open Request a copy

These files are only available to some KU Leuven Association staff members


All items in Lirias are protected by copyright, with all rights reserved.