Zoek binnen publicaties
Twenty years of satellite and in situ observations of surface chlorophyll-a from the northern Bay of Biscay to the eastern English Channel : is the water quality improving?
The variability of the phytoplankton biomass derived from daily chlorophyll-a (Chl-a) satellite images was investigated over the period 1998–2017 in the surface waters of the English Channel and the northern Bay of Biscay. Merged satellite (SeaWiFS-MODIS/Aqua-MERIS-VIIRS) Chl-a was calculated using the OC5 Ifremer algorithm which is optimized for moderately-turbid waters. The seasonal cycle in satellite-derived Chl-a was compared with in situ measurements made at seven coastal stations located in the southern side of the English Channel and in the northern Bay of Biscay. The results firstly showed that the satellite Chl-a product, derived from a suite of space-borne marine reflectance data, is in agreement with the coastal observations. For compliance with the directives of the European Union on water quality, time-series of 6-year moving average of Chl-a were assessed over the region. A clear decline was observed in the mean and 90th percentile of Chl-a at stations located in the mixed waters of the English Channel. The time-series at the stations located in the Bay of Biscay showed yearly fluctuations which correlated well with river discharge, but no overall Chl-a trend was observed. In the English Channel, the shape of the seasonal cycle in Chl-a changed over time. Narrower peaks were observed in spring at the end of the studied period, indicating an earlier limitation by nutrients. Monthly averages of satellite Chl-a, over the periods 1998–2003 and 2012–2017, exhibited spatial and temporal patterns in the evolution of the phytoplankton biomass similar to these observed at the seven coastal stations. Both the in situ and satellite Chl-a times-series showed a decrease in Chl-a in the English Channel in May, June and July. This trend in phytoplankton biomass is correlated with lower river discharges at the end of the period and a constant reduction in the riverine input of phosphorus through improvements in the water quality of the surrounding river catchments.
Numerical analysis of the Shiaolin landslide using Material Point Method
The Shiaolin landslide in southern Taiwan was triggered by Typhoon Morakot in 2009. The landslide carried out a volume that more than 22 million m3 and travelled a distance that over 2 km. Many researchers conducted studies in the field investigations, laboratory experiments, and numerical simulation. However, the simulations of initial failure stage and post-failure stage had to implement separately due to the problem of mesh distortion. In this study, a fully dynamic and coupled two-phase formulations using material point method is used to simulate the entire landslide process. The topographies, field investigations, and laboratory experiments are used in the model. In the last section, the evolutions of sliding surface and the varying moving distances of landslide are discussed by the different thickness of infiltration zone and coefficients of friction along ground surface.
Proceedings of the 7th International Symposium on Geotechnical Safety and Risk - ISGSR 2019 (Taipei, Taiwan, 12-13 December 2019)
De haarvaten van het ecologisch riviernetwerk : consequenties van fragmentatie en het belang van landbruggen en oeverzones
Fragmentatie heeft invloed op de habitatkwaliteit. Ecologische verbetering van beken en rivieren is daarom ook afhankelijk van het omringende land: natuurlijke oeverzones bieden schaduw en structuur en landbruggen verbinden bovenlopen. Een herwaardering van het denken in ecologische netwerken, op basis van de huidige kennis over migratiebehoeftes en habitateisen van soorten, onderschrijft de noodzaak voor additionele maatregelen. Dit artikel gaat in op de onderbelichte rol van het wijdere stroomgebied. Vanuit de grote rivieren kijken we stroomopwaarts naar de rol van zijrivieren en beken, de consequenties van fragmentatie en de rol van de oeverzones en het landschap rondom de kleinste bovenlopen, oftewel tot in de haarvaten van het riviersysteem.
Klimaatadaptatie in het rivierengebied : een geo-ecologisch perspectief
Door klimaatverandering verandert het afvoerregime van onze grote rivieren. Hoogwaters worden hoger en frequenter, laagwaters lager en langduriger. Hoe we daarop reageren hangt af van hoe we klimaatverandering zien: als opgave, of als kans om onvolkomenheden aan te pakken. In dit artikel presenteren we aanzetten voor een meer geo-ecologisch gefundeerde inrichting, of -naar McHarg- voor design with nature.
A framework for predicting rainfall-induced landslides using machine learning methods
Landslides are catastrophic geo-hazards that threaten urbanization. Growth in population besides construction of critical infrastructures such as roads and pipelines in landslide-prone areas elevates the risk associated with landslides. Therefore, a system that is able to predict landslides and issues warning in a timely manner is very appealing. Various factors influence the stability of natural and engineered slopes and cause landslides, including topography, geology of slopes, precipitation, temperature changes, snowmelt, seismic activities, vol-canic activities, and human actions. It is widely accepted that precipitation is one of the most influential factors for triggering landslides. In this paper, we present the preliminary results of a practical research study that has been carried out in Deltares, The Netherlands. To that end, we have set up a framework that combines geo-engineering, remote sensing, hydrology with machine learning to predict the onset of landslides under the effect of precipitation. In this data-driven approach, Machine Learning (ML) methods are used to predict landslides by exploiting multiple Earth observation datasets, including rainfall data (e.g. TRMM 3B42) and Digital Elevation Models (e.g. SRTM1) and the NASA Global Landslide Catalogue. A detailed inventory of 10,988 landslides at a global level is built out of which 4,542 cases are used to train a supervised machine learning algorithm. The trained ML model is then fed by rainfall data, topography features such as slope and elevation relief, soil and bedrock data, and vegetation index of target regions to assess the stability of the studied area.
Proceedings of the 17th European Conference on Soil Mechanics and Foundation Engineering - ECSMGE 2019 : geotechnical engineering, foundation of the future (Reykjavik, Iceland, 1-6 September 2019)
A Bayesian approach to ecosystem service trade‑off analysis utilizing expert knowledge
The concept of ecosystem services is gaining attention in the context of sustainable resource management. However, it is inherently difficult to account for tangible and intangible services in a combined model. The aim of this study is to extend the definition of ecosystem service trade-offs by using Bayesian Networks to capture the relationship between tangible and intangible ecosystem services. Tested is the potential of creating such a network based on existing literature and enhancement via expert elicitation. This study discusses the significance of expert elicitation to enhance the value of a Bayesian Network in data-restricted case studies, underlines the importance of inclusion of experts’ certainty, and demonstrates how multiple sources of knowledge can be combined into one model accounting for both tangible and intangible ecosystem services. Bayesian Networks appear to be a promising tool in this context, nevertheless, this approach is still in need of further refinement in structure and applicable guidelines for expert involvement and elicitation for a more unified methodology.
Rheological analysis of mud from Port of Hamburg, Germany
An innovative way to define navigable fluid mud layers is to make use of their rheological properties, in particular their yield stress. In order to help the development of in situ measurement techniques, it is essential that the key rheological parameters are estimated beforehand. We investigated the changes in the rheological properties of mud from along the river stream in the Port of Hamburg, Germany, using a recently developed laboratory protocol. A variety of rheological tests was performed including: stress sweep tests, flow curves, thixotropic tests, oscillatory amplitude, and frequency sweep tests. The yield stresses of sediments from different locations were significantly dissimilar from each other due to differences in densities and organic matter content. Two yield stresses (termed static and fluidic) were observed for every sample and linearly correlated to each other. The thixotropic studies showed that all mud samples, except from one location, displayed a combination of thixotropic and anti-thixotropic behaviors. The results of frequency sweep tests showed the solid-like character of the sediments within the linear viscoelastic limit. The yield stresses, thixotropy, and moduli of the mud samples increased by going deeper into the sediment bed due to the increase in density of the sediments. This study confirmed the applicability of the recently developed protocol as a fast and reliable tool to measure the yield stresses of sediments from different locations and depths in the Port of Hamburg. The fluid mud layer, in all the locations it was observed, exhibited relatively small yield stress values and weak thixotropic behavior. This confirms that despite the fact that rheology of fluid mud is complex, this layer can be navigable.
Modelling the torque with artificial neural networks on a tunnel boring machine
The performance of earth pressure balanced tunnel boring machines (EPB-TBM) is dependent of a variety of parameters. Moreover, these parameters interact in a rather challenging way, making it difficult to adequately model their behaviour. Artificial neural networks have the aptitude to model complex problems and have been used in a variety of construction engineering problems. They can learn from existing data and then be used to predict the results, which makes them adequate for modelling problems where large amount of data is generated. In this work, a multilayer feedforward artificial neural network has been used to predict the torque at the cutter head of an EPB-TBM. A time series neural network has been used, where torque was predicted as a function of the measured torque and the volume of the injected foam on previous time steps. Results indicate that feedforward artificial neural network can be used to predict the torque at the cutter head in a EPB-TBM.