| 1. |
- Dalton, April S., et al.
(författare)
-
Deglaciation of the north American ice sheet complex in calendar years based on a comprehensive database of chronological data: NADI-1
- 2023
-
Ingår i: QUATERNARY SCIENCE REVIEWS. - 0277-3791 .- 1873-457X. ; 321
-
Tidskriftsartikel (refereegranskat)abstract
- The most recent deglaciation of the North American Ice Sheet Complex (NAISC: comprising the Innuitian, Cordilleran, and Laurentide ice sheets) offers a broad perspective from which to analyze the timing and rate of ice retreat, deglacial sea-level rise, and abrupt climate change events. Previous efforts to portray the retreat of the NAISC have been focused largely on minimum-limiting radiocarbon ages and ice margin location(s) tied to deglacial landforms that were not, for the most part, chronologically constrained. Here, we present the first version of North American Deglaciation Isochrones (NADI-1) spanning 25 to 1 ka in calendar years before present. Key new features of this work are (i) the incorporation of cosmogenic nuclide data, which offer a direct constraint on the timing of ice recession; (ii) presentation of all data and time-steps in calendar years; (iii) optimal, minimum, and maximum ice extents for each time-step that are designed to capture uncertainties in the ice margin position, and; (iv) extensive documentation and justification for the placement of each ice margin. Our data compilation includes 2229 measurements of Be-10, 459 measurements of Al-26 and 35 measurements of Cl-36 from a variety of settings, including boulders, bedrock surfaces, cobbles, pebbles, and sediments. We also updated a previous radiocarbon dataset (n = 4947), assembled luminescence ages (n = 397) and gathered uranium-series data (n = 2). After scrutiny of the geochronological dataset, we consider >90% of data to be reliable or likely reliable. Key findings include (i) a highly asynchronous maximum glacial extent in North America, occurring as early as 27 ka to as late as 17 ka, within and between ice sheets. In most marine realms, extension of the ice margin to the continental shelf break at 25 ka is somewhat speculative because it is based on undated and spatially scattered ice stream and geomorphic evidence; (ii) detachment of the Laurentide and Cordilleran ice sheets took place gradually via southerly and northerly 'unzipping' of the ice masses, starting at 17.5 ka and ending around 14 ka; (iii) the final deglaciation of Hudson Bay began at 8.5 ka, with the collapse completed by 8 ka. The maximum extent of ice during the last glaciation occurred at 22 ka and covered 15,470,000 km(2). All North American ice sheets merged at 22 ka for the first time in the Quaternary. The highly asynchronous Last Glacial Maximum in North America means that our isochrones (starting at 25 ka) capture ice advance across some areas, which is based on limited evidence and is therefore somewhat speculative. In the Supplementary Data, the complete NADI-1 chronology is available in PDF, GIF and shapefile format, together with additional visualizations and spreadsheets of geochronological data. The NADI-1 shapefiles are also available at https://doi.org/10.5281/zenodo.8161764.
|
|
| 2. |
- Dalton, April S., et al.
(författare)
-
The marine δ18O record overestimates continental ice volume during Marine Isotope Stage 3
- 2022
-
Ingår i: Global and Planetary Change. - Amsterdam : Elsevier. - 0921-8181 .- 1872-6364. ; 212, s. 103814-103814
-
Tidskriftsartikel (refereegranskat)abstract
- There is disagreement in the Quaternary research community in how much of the marine δ18O signal is driven by change in ice volume. Here, we examine this topic by bringing together empirical and modelling work for Marine Isotope Stage 3 (MIS 3; 57 ka to 29 ka), a time when the marine δ18O record indicates moderate continental glaciation and a global mean sea level between −60 m and −90 m. We compile and interpret geological data dating to MIS 3 to constrain the extent of major Northern Hemisphere ice sheets (Eurasian, Laurentide, Cordilleran). Many key data, especially published in the past ~15 years, argue for an ice-free core of the formerly glaciated regions that is inconsistent with inferences from the marine δ18O record. We compile results from prior studies of glacial isostatic adjustment to show the volume of ice inferred from the marine δ18O record is unable to fit within the plausible footprint of Northern Hemisphere ice sheets during MIS 3. Instead, a global mean sea level between −30 m and − 50 m is inferred from geological constraints and glacial isostatic modelling. Furthermore, limited North American ice volumes during MIS 3 are consistent with most sea-level bounds through that interval. We can find no concrete evidence of large-scale glaciation during MIS 3 that could account for the missing ~30 m of sea-level equivalent during that time, which suggests that changes in the marine δ18O record are driven by other variables, including water temperature. This work urges caution regarding the reliance of the marine δ18O record as a de facto indicator of continental ice when few geological constraints are available, which underpins many Quaternary studies.
|
|
| 3. |
- Nhu, Viet-Ha, et al.
(författare)
-
Landslide Susceptibility Mapping Using Machine Learning Algorithms and Remote Sensing Data in a Tropical Environment
- 2020
-
Ingår i: International Journal of Environmental Research and Public Health. - Switzerland : MDPI. - 1661-7827 .- 1660-4601. ; 17:14
-
Tidskriftsartikel (refereegranskat)abstract
- We used AdaBoost (AB), alternating decision tree (ADTree), and their combination as an ensemble model (AB-ADTree) to spatially predict landslides in the Cameron Highlands, Malaysia. The models were trained with a database of 152 landslides compiled using Synthetic Aperture Radar Interferometry, Google Earth images, and field surveys, and 17 conditioning factors (slope, aspect, elevation, distance to road, distance to river, proximity to fault, road density, river density, normalized difference vegetation index, rainfall, land cover, lithology, soil types, curvature, profile curvature, stream power index, and topographic wetness index). We carried out the validation process using the area under the receiver operating characteristic curve (AUC) and several parametric and non-parametric performance metrics, including positive predictive value, negative predictive value, sensitivity, specificity, accuracy, root mean square error, and the Friedman and Wilcoxon sign rank tests. The AB model (AUC = 0.96) performed better than the ensemble AB-ADTree model (AUC = 0.94) and successfully outperformed the ADTree model (AUC = 0.59) in predicting landslide susceptibility. Our findings provide insights into the development of more efficient and accurate landslide predictive models that can be used by decision makers and land-use managers to mitigate landslide hazards.
|
|
| 4. |
- Nhu, Viet-Ha, et al.
(författare)
-
Shallow Landslide Susceptibility Mapping : A Comparison between Logistic Model Tree, Logistic Regression, Naïve Bayes Tree, Artificial Neural Network, and Support Vector Machine Algorithms
- 2020
-
Ingår i: International Journal of Environmental Research and Public Health. - Switzerland : MDPI. - 1661-7827 .- 1660-4601. ; 17:8, s. 1-30
-
Tidskriftsartikel (refereegranskat)abstract
- Shallow landslides damage buildings and other infrastructure, disrupt agriculture practices, and can cause social upheaval and loss of life. As a result, many scientists study the phenomenon, and some of them have focused on producing landslide susceptibility maps that can be used by land-use managers to reduce injury and damage. This paper contributes to this effort by comparing the power and effectiveness of five machine learning, benchmark algorithms—Logistic Model Tree, Logistic Regression, Naïve Bayes Tree, Artificial Neural Network, and Support Vector Machine—in creating a reliable shallow landslide susceptibility map for Bijar City in Kurdistan province, Iran. Twenty conditioning factors were applied to 111 shallow landslides and tested using the One-R attribute evaluation (ORAE) technique for modeling and validation processes. The performance of the models was assessed by statistical-based indexes including sensitivity, specificity, accuracy, mean absolute error (MAE), root mean square error (RMSE), and area under the receiver operatic characteristic curve (AUC). Results indicate that all the five machine learning models performed well for shallow landslide susceptibility assessment, but the Logistic Model Tree model (AUC = 0.932) had the highest goodness-of-fit and prediction accuracy, followed by the Logistic Regression (AUC = 0.932), Naïve Bayes Tree (AUC = 0.864), ANN (AUC = 0.860), and Support Vector Machine (AUC = 0.834) models. Therefore, we recommend the use of the Logistic Model Tree model in shallow landslide mapping programs in semi-arid regions to help decision makers, planners, land-use managers, and government agencies mitigate the hazard and risk.
|
|
| 5. |
- Salvati, Aryan, et al.
(författare)
-
Flood susceptibility mapping using support vector regression and hyper-parameter optimization
- 2023
-
Ingår i: Journal of Flood Risk Management. - : John Wiley and Sons Inc. - 1753-318X. ; 16:4
-
Tidskriftsartikel (refereegranskat)abstract
- Floods are both complex and destructive, and in most parts of the world cause injury, death, loss of agricultural land, and social disruption. Flood susceptibility (FS) maps are used by land-use managers and land owners to identify areas that are at risk from flooding and to plan accordingly. This study uses machine learning ensembles to produce objective and reliable FS maps for the Haraz watershed in northern Iran. Specifically, we test the ability of the support vector regression (SVR), together with linear kernel (LK), base classifier (BC), and hyper-parameter optimization (HPO), to identify flood-prone areas in this watershed. We prepared a map of 201 past floods to predict future floods. Of the 201 flood events, 151 (75%) were used for modeling and 50 (25%) were used for validation. Based on the relevant literature and our field survey of the study area, 10 effective factors were selected and prepared for flood zoning. The results show that three of the 10 factors are most important for predicting flood-sensitive areas, specifically and in order of importance, slope, distance to the river and river. Additionally, the SVR-HPO model, with area under the curve values of 0.986 and 0.951 for the training and testing phases, outperformed the other two tested models.
|
|
| 6. |
- Shahabi, Himan, et al.
(författare)
-
Flood Detection and Susceptibility Mapping Using Sentinel-1 Remote Sensing Data and a Machine Learning Approach : Hybrid Intelligence of Bagging Ensemble Based on K-Nearest Neighbor Classifier
- 2020
-
Ingår i: Remote Sensing. - Switzerland : MDPI. - 2072-4292. ; 12:2, s. 1-30
-
Tidskriftsartikel (refereegranskat)abstract
- Mapping flood-prone areas is a key activity in flood disaster management. In this paper, we propose a new flood susceptibility mapping technique. We employ new ensemble models based on bagging as a meta-classifier and K-Nearest Neighbor (KNN) coarse, cosine, cubic, and weighted base classifiers to spatially forecast flooding in the Haraz watershed in northern Iran. We identified flood-prone areas using data from Sentinel-1 sensor. We then selected 10 conditioning factors to spatially predict floods and assess their predictive power using the Relief Attribute Evaluation (RFAE) method. Model validation was performed using two statistical error indices and the area under the curve (AUC). Our results show that the Bagging–Cubic–KNN ensemble model outperformed other ensemble models. It decreased the overfitting and variance problems in the training dataset and enhanced the prediction accuracy of the Cubic–KNN model (AUC=0.660). We therefore recommend that the Bagging–Cubic–KNN model be more widely applied for the sustainable management of flood-prone areas.
|
|
| 7. |
- Sharifipour, Behzad, et al.
(författare)
-
Rangeland species potential mapping using machine learning algorithms
- 2023
-
Ingår i: Ecological Engineering. - : Elsevier. - 0925-8574 .- 1872-6992. ; 189
-
Tidskriftsartikel (refereegranskat)abstract
- Documenting habitats of rangeland plant species is required to properly manage rangelands and to understand ecosystem processes. A reliable rangeland species potential map can help managers and policy makers design a sustainable grazing system on rangelands. The aim of this study is to map the plant species in the Qurveh City rangelands, Kurdistan Province, Iran, using state-of-the-art machine learning algorithms, including Support Vector Machine (SVM), Artificial Neural Network (ANN), Naïve Bayes (NB), Bayes Net (BN) and Classification and Regression Tree (CART). A total of 185 rangeland species were used in the study, together with 20 conditioning factors, to build and validate models. The One-R feature section technique and multicollinearity test were used, respectively, to determine the most important factors and correlations between them. Model validation was performed using sensitivity, specificity, accuracy, F1-measure, Matthews correlation coefficient (MCC), Kappa, root mean square error (RMSE), and area under the receiver operating characteristic curve (AUC). Results showed that topographic wetness index (TWI), slope angle, elevation, soil phosphorus and soil potassium were the five most important factors to increase the rangeland plants habitat suitability. The Naïve Bayes algorithm (AUC = 0.782) had the highest performance and prediction accuracy and best consistency across the species in the investigated rangeland, followed by the SVM (AUC = 0.763), ANN (AUC = 0.762), CART (AUC = 0.627), and BN (AUC = 0.617) models.
|
|
