Marinus Eric Donselaar
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Research Interests
Marinus Eric Donselaar has a total of 116 co-authors. Here are the top co-authors by number of shared publications:
- Koen A. van Toorenenburg
Department of Geoscience and Engineering Delft University of Technology Delft The Netherlands
8 shared publications G.J. Weltje
7 shared publications- Ashok Kumar Ghosh
Mahavir Cancer Sansthan & Research Centre, Patna, India
4 shared publications - Cees J. L. Willems
N.V. HVC, Jadestraat 1, Alkmaar, 1812 RD, Netherlands
4 shared publications Stefan M. Luthi
4 shared publicationsGert Jan Weltje
Department of Earth and Environmental Sciences University of Leuven Leuven Belgium
3 shared publications
Latest Publications
- Open Access
- 15 May 2026
- Published by Elsevier in SSRN Electronic Journal
Detection of geogenic arsenic contamination hotspots in alluvial basins by geomorphically informed machine learning
Natural arsenic toxicity in densely-populated Holocene alluvial-basin aquifers affects millions of people and poses a global public health concern. Sedimentation in these basins forms geogenic arsenic hotspots governed by geomorphology-controlled groundwater flow paths, promoting arsenic mobilization and accumulation in permeable point bars enclosed by low-permeable oxbow-lake deposits. This study presents a geomorphologically informed, machine-learning-based framework to improve localization and interpretation of arsenic risk in such deposits, starting from remote sensing data, transferable to similar alluvial settings. The workflow is demonstrated in the highly arsenic-affected Ganges-Brahmaputra Basin. The framework integrates two stages. First, the YOLO (You Only Look Once) model is applied to 10~m resolution Sentinel-2 spectral satellite imagery to detect oxbow lakes and associated point bars, delineating geomorphologically relevant zones. Second, within these constrained regions, a Gaussian Mixture Model (GMM) classifies arsenic risk using topographic elevation and population density as input, producing probabilistic arsenic risk classes, enabling spatial risk mapping and uncertainty quantification. The YOLO model demonstrates overall good robust detection of well-defined oxbow lakes across varying seasonal and environmental conditions, though performance declines in complex geomorphological (dry) landscapes. The GMM yields spatially coherent, interpretable arsenic risk patterns consistent with geomorphological theory, highlighting elevated point-bar ridges as zones of increased risk. The results show that geomorphological constraints meaningfully guide machine-learning models for arsenic prediction, reducing reliance on field measurements and minimizing interpolation bias. The proposed framework enables a scalable, data-efficient identification of arsenic hotspots in data-scarce alluvial systems, supporting a foundation for targeted field validation.
- 15 May 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessDetection of geogenic arsenic contamination hotspots in alluvial basins by geomorphically informed machine learning
Natural arsenic toxicity in densely-populated Holocene alluvial-basin aquifers affects millions of people and poses a global public health concern. Sedimentation in these basins forms geogenic arsenic hotspots governed by geomorphology-controlled groundwater flow paths, promoting arsenic mobilization and accumulation in permeable point bars enclosed by low-permeable oxbow-lake deposits. This study presents a geomorphologically informed, machine-learning-based framework to improve localization and interpretation of arsenic risk in such deposits, starting from remote sensing data, transferable to similar alluvial settings. The workflow is demonstrated in the highly arsenic-affected Ganges-Brahmaputra Basin. The framework integrates two stages. First, the YOLO (You Only Look Once) model is applied to 10~m resolution Sentinel-2 spectral satellite imagery to detect oxbow lakes and associated point bars, delineating geomorphologically relevant zones. Second, within these constrained regions, a Gaussian Mixture Model (GMM) classifies arsenic risk using topographic elevation and population density as input, producing probabilistic arsenic risk classes, enabling spatial risk mapping and uncertainty quantification. The YOLO model demonstrates overall good robust detection of well-defined oxbow lakes across varying seasonal and environmental conditions, though performance declines in complex geomorphological (dry) landscapes. The GMM yields spatially coherent, interpretable arsenic risk patterns consistent with geomorphological theory, highlighting elevated point-bar ridges as zones of increased risk. The results show that geomorphological constraints meaningfully guide machine-learning models for arsenic prediction, reducing reliance on field measurements and minimizing interpolation bias. The proposed framework enables a scalable, data-efficient identification of arsenic hotspots in data-scarce alluvial systems, supporting a foundation for targeted field validation.
- Open Access
- 15 July 2024
- Published by American Chemical Society (ACS) in ACS ES&T Water
- Vol. 4 (8) , 3110-3114
Machine-Learning Approach for Identifying Arsenic-Contamination Hot Spots: The Search for the Needle in the Haystack
- 15 July 2024
- Published by American Chemical Society (ACS) in ACS ES&T Water
- Vol. 4 (8) , 3110-3114
Open AccessMachine-Learning Approach for Identifying Arsenic-Contamination Hot Spots: The Search for the Needle in the Haystack
- Open Access
- 23 December 2021
- Published by Wiley in Earth Surface Processes and Landforms
- Vol. 47 (5) , 1212-1228
Spatio‐temporal reconstruction of avulsion history at the terminus of a modern dryland river system
Fluvial depositional architecture in an unconfined environment is governed by sediment dispersal across the alluvial plain through river‐path switching by avulsion. Documented inter‐avulsion periodicity from modern rivers ranges from tens to over a thousand years. In this study, a quantitative spatio‐temporal reconstruction of avulsion history is presented of the non‐vegetated and pristine modern Río Colorado dryland river system in the semi‐arid Altiplano Basin (Bolivia), based on the integrated analysis of satellite imagery and absolute age dating using optically stimulated luminescence, complemented with sedimentological and geomorphological ground‐truth data. This approach enables us to reconstruct the chronological order of channel belts of the Río Colorado, to determine avulsion recurrence time and inter‐avulsion periodicity, to identify mechanisms for flow path changes, and to present a morphodynamic model for the spatio‐temporal evolution of fluvial deposits in a semi‐arid environment. In a maximum timespan of 12.71 ± 1.5 ka, successive avulsions of the Río Colorado created a sheet of interconnected fluvial deposits, consisting of diverging and juxtaposed alluvial ridges that formed by sediment aggradation in point bars, crevasse splays, levees, and on the channel floor. The ridges show lateral onlap and amalgamation as the result of repeated avulsion and compensational stacking, whereby the river avoided the positive alluvial‐ridge relief of its precursors. The resultant morphology is fan‐shaped, convex‐up with a surface area of approximately 500 km2 and a maximum observed thickness of 3 m. The results show inter‐avulsion periods of the river of up to 1.28 ± 0.34 ka. A paucity in fluvial activity around 2 ka BP, and at present, is interpreted as the result of low river discharge related to long‐term dry periodicity in the El Niño Southern Oscillation circulation system. Each river path started as a low sinuous, single‐thread channel in a narrow belt, and in time increased its width and sinuosity by point‐bar expansion and rotation.
- 23 December 2021
- Published by Wiley in Earth Surface Processes and Landforms
- Vol. 47 (5) , 1212-1228
Open AccessSpatio‐temporal reconstruction of avulsion history at the terminus of a modern dryland river system
Fluvial depositional architecture in an unconfined environment is governed by sediment dispersal across the alluvial plain through river‐path switching by avulsion. Documented inter‐avulsion periodicity from modern rivers ranges from tens to over a thousand years. In this study, a quantitative spatio‐temporal reconstruction of avulsion history is presented of the non‐vegetated and pristine modern Río Colorado dryland river system in the semi‐arid Altiplano Basin (Bolivia), based on the integrated analysis of satellite imagery and absolute age dating using optically stimulated luminescence, complemented with sedimentological and geomorphological ground‐truth data. This approach enables us to reconstruct the chronological order of channel belts of the Río Colorado, to determine avulsion recurrence time and inter‐avulsion periodicity, to identify mechanisms for flow path changes, and to present a morphodynamic model for the spatio‐temporal evolution of fluvial deposits in a semi‐arid environment. In a maximum timespan of 12.71 ± 1.5 ka, successive avulsions of the Río Colorado created a sheet of interconnected fluvial deposits, consisting of diverging and juxtaposed alluvial ridges that formed by sediment aggradation in point bars, crevasse splays, levees, and on the channel floor. The ridges show lateral onlap and amalgamation as the result of repeated avulsion and compensational stacking, whereby the river avoided the positive alluvial‐ridge relief of its precursors. The resultant morphology is fan‐shaped, convex‐up with a surface area of approximately 500 km2 and a maximum observed thickness of 3 m. The results show inter‐avulsion periods of the river of up to 1.28 ± 0.34 ka. A paucity in fluvial activity around 2 ka BP, and at present, is interpreted as the result of low river discharge related to long‐term dry periodicity in the El Niño Southern Oscillation circulation system. Each river path started as a low sinuous, single‐thread channel in a narrow belt, and in time increased its width and sinuosity by point‐bar expansion and rotation.
- 1 September 2020
- Published by Elsevier in Marine and Petroleum Geology
- Vol. 119
Long-range, high-resolution stratigraphic correlation of Rotliegend fluvial-fan deposits in the central Dutch offshore
- 1 September 2020
- Published by Elsevier in Marine and Petroleum Geology
- Vol. 119
Long-range, high-resolution stratigraphic correlation of Rotliegend fluvial-fan deposits in the central Dutch offshore
- Open Access
- 19 March 2020
- Published by Netherlands Journal of Geosciences Foundation in Netherlands Journal of Geosciences
- Vol. 99
Geology of the Upper Jurassic to Lower Cretaceous geothermal aquifers in the West Netherlands Basin – an overview
In the past 10 years the mature hydrocarbon province the West Netherlands Basin has hosted rapidly expanding geothermal development. Upper Jurassic to Lower Cretaceous strata from which gas and oil had been produced since the 1950s became targets for geothermal exploitation. The extensive publicly available subsurface data including seismic surveys, several cores and logs from hundreds of hydrocarbon wells, combined with understanding of the geology after decades of hydrocarbon exploitation, facilitated the offtake of geothermal exploitation. Whilst the first geothermal projects proved the suitability of the permeable Upper Jurassic to Lower Cretaceous sandstones for geothermal heat production, they also made clear that much detail of the aquifer geology is not yet fully understood. The aquifer architecture varies significantly across the basin because of the syn-tectonic sedimentation. The graben fault blocks that contain the geothermal targets experienced a different tectonic history compared to the horst and pop-up structures that host the hydrocarbon fields from which most subsurface data are derived. Accurate prediction of the continuity and thickness of aquifers is a prerequisite for efficient geothermal well deployment that aims at increasing heat recovery while avoiding the risk of early cold-water breakthrough. The potential recoverable heat and the current challenges to enhance further expansion of heat exploitation from this basin are evident. This paper presents an overview of the current understanding and uncertainties of the aquifer geology of the Upper Jurassic to Lower Cretaceous strata and discusses new sequence-stratigraphic updates of the regional sedimentary aquifer architecture.
- 19 March 2020
- Published by Netherlands Journal of Geosciences Foundation in Netherlands Journal of Geosciences
- Vol. 99
Open AccessGeology of the Upper Jurassic to Lower Cretaceous geothermal aquifers in the West Netherlands Basin – an overview
In the past 10 years the mature hydrocarbon province the West Netherlands Basin has hosted rapidly expanding geothermal development. Upper Jurassic to Lower Cretaceous strata from which gas and oil had been produced since the 1950s became targets for geothermal exploitation. The extensive publicly available subsurface data including seismic surveys, several cores and logs from hundreds of hydrocarbon wells, combined with understanding of the geology after decades of hydrocarbon exploitation, facilitated the offtake of geothermal exploitation. Whilst the first geothermal projects proved the suitability of the permeable Upper Jurassic to Lower Cretaceous sandstones for geothermal heat production, they also made clear that much detail of the aquifer geology is not yet fully understood. The aquifer architecture varies significantly across the basin because of the syn-tectonic sedimentation. The graben fault blocks that contain the geothermal targets experienced a different tectonic history compared to the horst and pop-up structures that host the hydrocarbon fields from which most subsurface data are derived. Accurate prediction of the continuity and thickness of aquifers is a prerequisite for efficient geothermal well deployment that aims at increasing heat recovery while avoiding the risk of early cold-water breakthrough. The potential recoverable heat and the current challenges to enhance further expansion of heat exploitation from this basin are evident. This paper presents an overview of the current understanding and uncertainties of the aquifer geology of the Upper Jurassic to Lower Cretaceous strata and discusses new sequence-stratigraphic updates of the regional sedimentary aquifer architecture.
- 1 December 2019
- Published by Elsevier in Marine and Petroleum Geology
- Vol. 110, 450-462
Quantitative integration of sedimentological core descriptions and petrophysical data using high-resolution XRF core scans
- 1 December 2019
- Published by Elsevier in Marine and Petroleum Geology
- Vol. 110, 450-462
Quantitative integration of sedimentological core descriptions and petrophysical data using high-resolution XRF core scans
- Open Access
- 23 August 2019
- Published by Taylor & Francis
- p. 636-637
Small-scale piped water supply: end-user inclusive water research in arsenic affected areas in India and Bangladesh (DELTAP)
- 23 August 2019
- Published by Taylor & Francis
- p. 636-637
Open AccessSmall-scale piped water supply: end-user inclusive water research in arsenic affected areas in India and Bangladesh (DELTAP)
- Open Access
- 23 August 2019
- Published by Taylor & Francis
- p. 52-53
Arsenic volume estimates in Holocene clay plug sediments in Bihar, India
- 23 August 2019
- Published by Taylor & Francis
- p. 52-53
Open AccessArsenic volume estimates in Holocene clay plug sediments in Bihar, India
- 18 May 2018
- Published by Wiley in Earth Surface Processes and Landforms
- Vol. 43 (11) , 2409-2420
The life cycle of crevasse splays as a key mechanism in the aggradation of alluvial ridges and river avulsion
Accommodation space in the unconfined distal part of low‐gradient fluvial fans facilitates abundant floodplain deposition. Here, the development of crevasse splays plays a key role in the aggradation of alluvial ridges and subsequent river avulsion. This study presents an analysis of different stages in the evolution of crevasse splays based on observations made in the modern‐day Río Colorado dryland fluvial fan fringing the endorheic Altiplano Basin in Bolivia. A generic life cycle is proposed in which crevasse‐splay channels adjust towards a graded equilibrium profile with their lower‐lying distal termini acting as a local base level. Initial development is dominantly controlled by the outflow of floodwater, promoting erosion near the crevasse apex and deposition towards the splay fringes. When proximal incision advances to below the maximum level of floodplain inundation, return flow occurs during the waning stage of flooding. This floodwater reflux leads to a temporary repositioning of the local base level to the deeper trunk‐channel thalweg at the apex of the crevasse‐splay channels. The resultant decrease in the floodplainward gradient of these channels ultimately leads to backfilling and abandonment of the crevasse splay, leaving a subtle local elevation of the floodplain. Consecutive splays form an alluvial ridge through lateral amalgamation and subsequent vertical stacking, which is mirrored by the aggradation of their parent channel floor. As this alluvial ridge becomes increasingly perched above the surrounding floodplain, splay equilibration may cause incision of the levee crevasse down to or below its trunk channel thalweg, leading to an avulsion. The mechanisms proposed in this study are relevant to fluvial settings promoting progradational avulsions. The relatively rapid accumulation rate and high preservation potential of crevasse splays in this setting makes them an important constituent of the resultant fluvial stratigraphy, amongst which are hydrocarbon‐bearing successions. Copyright © 2018 John Wiley & Sons, Ltd.
- 18 May 2018
- Published by Wiley in Earth Surface Processes and Landforms
- Vol. 43 (11) , 2409-2420
The life cycle of crevasse splays as a key mechanism in the aggradation of alluvial ridges and river avulsion
Accommodation space in the unconfined distal part of low‐gradient fluvial fans facilitates abundant floodplain deposition. Here, the development of crevasse splays plays a key role in the aggradation of alluvial ridges and subsequent river avulsion. This study presents an analysis of different stages in the evolution of crevasse splays based on observations made in the modern‐day Río Colorado dryland fluvial fan fringing the endorheic Altiplano Basin in Bolivia. A generic life cycle is proposed in which crevasse‐splay channels adjust towards a graded equilibrium profile with their lower‐lying distal termini acting as a local base level. Initial development is dominantly controlled by the outflow of floodwater, promoting erosion near the crevasse apex and deposition towards the splay fringes. When proximal incision advances to below the maximum level of floodplain inundation, return flow occurs during the waning stage of flooding. This floodwater reflux leads to a temporary repositioning of the local base level to the deeper trunk‐channel thalweg at the apex of the crevasse‐splay channels. The resultant decrease in the floodplainward gradient of these channels ultimately leads to backfilling and abandonment of the crevasse splay, leaving a subtle local elevation of the floodplain. Consecutive splays form an alluvial ridge through lateral amalgamation and subsequent vertical stacking, which is mirrored by the aggradation of their parent channel floor. As this alluvial ridge becomes increasingly perched above the surrounding floodplain, splay equilibration may cause incision of the levee crevasse down to or below its trunk channel thalweg, leading to an avulsion. The mechanisms proposed in this study are relevant to fluvial settings promoting progradational avulsions. The relatively rapid accumulation rate and high preservation potential of crevasse splays in this setting makes them an important constituent of the resultant fluvial stratigraphy, amongst which are hydrocarbon‐bearing successions. Copyright © 2018 John Wiley & Sons, Ltd.
- Open Access
- 1 February 2018
- Published by Elsevier in Applied Energy
- Vol. 212, 1433-1447
Synergy potential for oil and geothermal energy exploitation
- 1 February 2018
- Published by Elsevier in Applied Energy
- Vol. 212, 1433-1447
Open AccessSynergy potential for oil and geothermal energy exploitation