Uncovering the physical controls of deep subduction zone slow slip using supervised classification of subducting plate features

2020 ◽  
Author(s):  
Morgan McLellan ◽  
Pascal Audet
Keyword(s):  
Subduction Zones ◽  
Weighted Average ◽  
Support Vector ◽  
Slow Slip ◽  
Sediment Thickness ◽  
Short Term ◽  
Linear Discriminant ◽  
Brittle Ductile Transition ◽  
Subducting Plate

Summary Deep slow slip events (SSEs) at subduction zones have significantly contributed to refining our understanding of the megathrust earthquake cycle at the brittle-ductile transition. However, the specific combination of factors that determine their occurrence has not yet been fully explored. Here we evaluate the contribution of several of these characteristics using globally mapped geophysical data that are used as proxies for physical properties of the subducting plate. This is performed by classifying 25 km-wide, trench-parallel segments into binary classes based on the observation (or lack thereof) of deep, short- or long-term SSEs. The five characteristics explored here include subducting plate age, sediment thickness, relative plate velocity, slab dip, and plate surface roughness. We use these characteristics to train six Machine Learning models based on different learning algorithms: Gaussian Naïve Bayes, Logistic Regression, Linear Discriminant Analysis, Random Forest, Support Vector Machine, and K-Nearest Neighbour. Short-term SSE models show that subducting plate age, relative velocity, and sediment thickness have the strongest predictive power with the first two characteristics negatively correlating and sediment thickness positively correlating with SSE occurrence, respectively. These results are consistent with a conceptual model where slow slip is controlled by conditions favoring the enduring release (and possible storage) of fluids near the source region. However, the relationship between these features and elevated pore fluid pressures is not established here and further evidence is needed to validate this hypothesis. We then use a final model constructed as a weighted average of the best performing models to make predictions on the probability of SSE occurrence, with predicted short-term SSE occurrence in South America, the Aleutians, Sumatra, Vanuatu and Solomon, as well as long-term SSE occurrence in the Aleutians, Izu-Bonin, Kuril-Kamchatka, Mariana, and Tonga-Kermadec. Overall, long-term SSE models do not perform as well as the short-term SSE models which may indicate that long-term SSEs are controlled by a different and/or extended set of physical characteristics than the short-term SSEs.

scholarly journals Dynamics of interplate domain in subduction zones: influence of rheological parameters and subducting plate age

2012 ◽  
Vol 4 (2) ◽  
pp. 943-992
Author(s):  
D. Arcay
Keyword(s):  
Subduction Zones ◽  
Activation Volume ◽  
Rheological Parameters ◽  
Strength Increase ◽  
Thermomechanical Model ◽  
Mantle Viscosity ◽  
Brittle Ductile Transition ◽  
Transition Depth ◽  
Subducting Plate

Abstract. The properties of the subduction interplate domain are likely to affect not only the seismogenic potential of the subduction area but also the overall subduction process, as it influences its viability. Numerical simulations are performed to model the long-term equilibrium state of the subduction interplate when the diving lithosphere interacts with both the overriding plate and the surrounding convective mantle. The thermomechanical model combines a non-Newtonian viscous rheology and a pseudo-brittle rheology. Rock strength here depends on depth, temperature and stress, for both oceanic crust and mantle rocks. I study the evolution through time of, on one hand, the kinematic decoupling depth, zdec and, on the other hand, of the brittle-ductile transition (BDT) depth, zBDT, simulated along the subduction interplate. The results reveal that zBDT mainly depends on the friction coefficient characterising the interplate channel and on the viscosity at the lithosphere-asthenosphere boundary. The influence of the weak material activation energy is of second order but not negligible. zBDT becomes dependent on the ductile strength increase with depth (activation volume) if the BDT occurs at the interplate deocupling depth. Regarding the interplate decoupling depth, it is basically a function of (1) mantle viscosity at asthenospheric wedge tip, (2) difference in mantle and interplate activation anergy, and (3) activation volume. Specific conditions yielding zBDT = zdec are discussed. I then present how the subducting lithosphere age affects the brittle-ductile transition depth and the kinematic decoupling depth in this model. Simulations show that a rheological model in which the respective activation energies of mantle and interplate material are too close impedes strain localization during incipient subduction of a young (20 Myr old) and soft lithosphere under a thick upper plate. Finally, both the BDT depth and the decoupling depth are a function of the subducting plate age, but are not influenced in the same fashion: cool and old subducting plates deepen the BDT but shallow the interplate decoupling depth. Even if BDT and kinematic decoupling are instrinsically related to different mechanisms of deformation, this work shows that they are able to interact closely.

Determinants of rental strategy: short-term vs long-term rental strategy

2020 ◽  
Vol 13 (12) ◽  
pp. 3873-3894
Author(s):  
Sina Shokoohyar ◽  
Ahmad Sobhani ◽  
Anae Sobhani
Keyword(s):  
Nearest Neighbor ◽  
Performance Metrics ◽  
Sharing Economy ◽  
Real Estate Market ◽  
Support Vector ◽  
Attractive Alternative ◽  
K Nearest Neighbor ◽  
Short Term ◽  
Content Type

Purpose Short-term rental option enabled via accommodation sharing platforms is an attractive alternative to conventional long-term rental. The purpose of this study is to compare rental strategies (short-term vs long-term) and explore the main determinants for strategy selection. Design/methodology/approach Using logistic regression, this study predicts the rental strategy with the highest rate of return for a given property in the City of Philadelphia. The modeling result is then compared with the applied machine learning methods, including random forest, k-nearest neighbor, support vector machine, naïve Bayes and neural networks. The best model is finally selected based on different performance metrics that determine the prediction strength of underlying models. Findings By analyzing 2,163 properties, the results show that properties with more bedrooms, closer to the historic attractions, in neighborhoods with lower minority rates and higher nightlife vibe are more likely to have a higher return if they are rented out through short-term rental contract. Additionally, the property location is found out to have a significant impact on the selection of the rental strategy, which emphasizes the widely known term of “location, location, location” in the real estate market. Originality/value The findings of this study contribute to the literature by determining the neighborhood and property characteristics that make a property more suitable for the short-term rental vs the long-term one. This contribution is extremely important as it facilitates differentiating the short-term rentals from the long-term rentals and would help better understanding the supply-side in the sharing economy-based accommodation market.

scholarly journals Lithological control on the landscape form of the upper Rhône Basin, Central Swiss Alps

2016 ◽  
Vol 4 (1) ◽  
pp. 253-272 ◽  
Author(s):  
Laura Stutenbecker ◽  
Anna Costa ◽  
Fritz Schlunegger
Keyword(s):  
Swiss Alps ◽  
Short Term ◽  
Linear Discriminant ◽  
Erosion Rates ◽  
Digital Elevation ◽  
Discriminant Analyses ◽  
Controlling Variables ◽  
Uplift And Erosion ◽  
Landscape Morphology

Abstract. The development of topography depends mainly on the interplay between uplift and erosion. These processes are controlled by various factors including climate, glaciers, lithology, seismic activity and short-term variables, such as anthropogenic impact. Many studies in orogens all over the world have shown how these controlling variables may affect the landscape's topography. In particular, it has been hypothesized that lithology exerts a dominant control on erosion rates and landscape morphology. However, clear demonstrations of this influence are rare and difficult to disentangle from the overprint of other signals such as climate or tectonics. In this study we focus on the upper Rhône Basin situated in the Central Swiss Alps in order to explore the relation between topography, possible controlling variables and lithology in particular. The Rhône Basin has been affected by spatially variable uplift, high orographically driven rainfalls and multiple glaciations. Furthermore, lithology and erodibility vary substantially within the basin. Thanks to high-resolution geological, climatic and topographic data, the Rhône Basin is a suitable laboratory to explore these complexities. Elevation, relief, slope and hypsometric data as well as river profile information from digital elevation models are used to characterize the landscape's topography of around 50 tributary basins. Additionally, uplift over different timescales, glacial inheritance, precipitation patterns and erodibility of the underlying bedrock are quantified for each basin. Results show that the chosen topographic and controlling variables vary remarkably between different tributary basins. We investigate the link between observed topographic differences and the possible controlling variables through statistical analyses. Variations of elevation, slope and relief seem to be linked to differences in long-term uplift rate, whereas elevation distributions (hypsometry) and river profile shapes may be related to glacial imprint. This confirms that the landscape of the Rhône Basin has been highly preconditioned by (past) uplift and glaciation. Linear discriminant analyses (LDAs), however, suggest a stronger link between observed topographic variations and differences in erodibility. We therefore conclude that despite evident glacial and tectonic conditioning, a lithologic control is still preserved and measurable in the landscape of the Rhône tributary basins.

scholarly journals Cost Index Predictions for Construction Engineering Based on LSTM Neural Networks

2020 ◽  
Vol 2020 ◽  
pp. 1-14 ◽  
Author(s):  
Jiacheng Dong ◽  
Yuan Chen ◽  
Gang Guan
Keyword(s):  
Neural Network ◽  
Support Vector ◽  
Short Term ◽  
Cost Index ◽  
Long Distance ◽  
Research Knowledge ◽  
Construction Engineering ◽  
Engineering Cost ◽  
Optimization Mechanism

In recent years, the cost index predictions of construction engineering projects are becoming important research topics in the field of construction management. Previous methods have limitations in reasonably reflecting the timeliness of engineering cost indexes. The recurrent neural network (RNN) belongs to a time series network, and the purpose of timeliness transfer calculation is achieved through the weight sharing of time steps. The long-term and short-term memory neural network (LSTM NN) solves the RNN limitations of the gradient vanishing and the inability to address long-term dependence under the premise of having the above advantages. The present study proposed a new framework based on LSTM, so as to explore the applicability and optimization mechanism of the algorithm in the field of cost indexes prediction. A survey was conducted in Shenzhen, China, where a total of 143 data samples were collected based on the index set for the corresponding time interval from May 2007 to March 2019. A prediction framework based on the LSTM model, which was trained by using these collected data, was established for the purpose of cost index predictions and test. The testing results showed that the proposed LSTM framework had obvious advantages in prediction because of the ability of processing high-dimensional feature vectors and the capability of selectively recording historical information. Compared with other advanced cost prediction methods, such as Support Vector Machine (SVM), this framework has advantages such as being able to capture long-distance dependent information and can provide short-term predictions of engineering cost indexes both effectively and accurately. This research extended current algorithm tools that can be used to forecast cost indexes and evaluated the optimization mechanism of the algorithm in order to improve the efficiency and accuracy of prediction, which have not been explored in current research knowledge.

Intraindividual variation of glycohemoglobin: implications for interpretation and analytical goals

1993 ◽  
Vol 39 (11) ◽  
pp. 2305-2308 ◽  
Author(s):  
G Phillipou ◽  
P J Phillips
Keyword(s):  
Weighted Average ◽  
Affinity Column ◽  
Time Interval ◽  
Short Term ◽  
Intraindividual Variation ◽  
Clinical Control ◽  
Significant Difference ◽  
Patients With Diabetes ◽  
Analytical Imprecision

Abstract Intraindividual variation (CVi) for glycohemoglobin (GHb) was estimated from serial measurements in patients with diabetes in either stable or variable clinical control. GHb determinations were performed by an affinity column procedure with an analytical imprecision of 4.9% (weighted average; GHb 8.2-14.7%). Within the groups of patients, both a short- (28-32 days) and long-term (approximately 85 days) sampling protocol was used. The derived CVi for each category was 4.2% (n = 16, stable, short-term), 7.1% (n = 23, stable, long-term), 5.1% (n = 13, variable, short-term), and 9.8% (n = 21, variable, long-term). The mean GHb within each category was similar (approximately 11%), and there was no statistically significant difference in GHb values between categories. The results establish that the CVi for GHb is affected by both clinical control and the sampling time interval. These findings have important implications for the estimation of significant differences between serial GHb measurements and the setting of appropriate analytical precision goals.

scholarly journals Research on Regional Short-Term Power Load Forecasting Model and Case Analysis

Processes ◽  
2021 ◽  
Vol 9 (9) ◽  
pp. 1617
Author(s):  
Kang Qian ◽  
Xinyi Wang ◽  
Yue Yuan
Keyword(s):  
Meteorological Factors ◽  
Load Forecasting ◽  
Weighted Average ◽  
Support Vector ◽  
Forecasting Model ◽  
Electric Load ◽  
Short Term ◽  
Power Load ◽  
Power Load Forecasting ◽  
Electric Load Forecasting

Integrated energy services will have multiple values and far-reaching significance in promoting energy transformation and serving “carbon peak and carbon neutralization”. In order to balance the supply and demand of power system in integrated energy, it is necessary to establish a scientific model for power load forecasting. Different algorithms for short-term electric load forecasting considering meteorological factors are presented in this paper. The correlation between electric load and meteorological factors is first analyzed. After the principal component analysis (PCA) of meteorological factors and autocorrelation analysis of the electric load, the daily load forecasting model is established by optimal support vector machine (OPT-SVM), Elman neural network (ENN), as well as their combinations through linear weighted average, geometric weighted average, and harmonic weighted average method, respectively. Based on the actual data of an industrial park of Nantong in China, the prediction performance in the four seasons with the different models is evaluated. The main contribution of this paper is to compare the effectiveness of different models for short-term electric load forecasting and to give a guideline to build the proper methods for load forecasting.

scholarly journals A Comparative Analysis of Machine Learning Approaches for Short-/Long-Term Electricity Load Forecasting in Cyprus

2020 ◽  
Vol 12 (9) ◽  
pp. 3612 ◽  
Author(s):  
Davut Solyali
Keyword(s):  
Machine Learning ◽  
Power Generation ◽  
Historical Data ◽  
Sustainable Growth ◽  
Solar Irradiation ◽  
Support Vector ◽  
Short Term ◽  
Inference System ◽  
Electricity Load

Estimating the electricity load is a crucial task in the planning of power generation systems and the efficient operation and sustainable growth of modern electricity supply networks. Especially with the advent of smart grids, the need for fairly precise and highly reliable estimation of electricity load is greater than ever. It is a challenging task to estimate the electricity load with high precision. Many energy demand management methods are used to estimate future energy demands correctly. Machine learning methods are well adapted to the nature of the electrical load, as they can model complicated nonlinear connections through a learning process containing historical data patterns. Many scientists have used machine learning (ML) to anticipate failure before it occurs as well as predict the outcome. ML is an artificial intelligence (AI) subdomain that involves studying and developing mathematical algorithms to understand data or obtain data directly without relying on a prearranged model algorithm. ML is applied in all industries. In this paper, machine learning strategies including artificial neural network (ANN), multiple linear regression (MLR), adaptive neuro-fuzzy inference system (ANFIS), and support vector machine (SVM) were used to estimate electricity demand and propose criteria for power generation in Cyprus. The simulations were adapted to real historical data explaining the electricity usage in 2016 and 2107 with long-term and short-term analysis. It was observed that electricity load is a result of temperature, humidity, solar irradiation, population, gross national income (GNI) per capita, and the electricity price per kilowatt-hour, which provide input parameters for the ML algorithms. Using electricity load data from Cyprus, the performance of the ML algorithms was thoroughly evaluated. The results of long-term and short-term studies show that SVM and ANN are comparatively superior to other ML methods, providing more reliable and precise outcomes in terms of fewer estimation errors for Cyprus’s time series forecasting criteria for power generation.

scholarly journals Abiotic and biotic processes that drive carboxylation and decarboxylation reactions

2020 ◽  
Vol 105 (5) ◽  
pp. 609-615
Author(s):  
Cody S. Sheik ◽  
H. James Cleaves ◽  
Kristin Johnson-Finn ◽  
Donato Giovannelli ◽  
Thomas L. Kieft ◽  
...  
Keyword(s):  
Metabolic Pathways ◽  
Subduction Zones ◽  
Carbon Fixation ◽  
Prebiotic Chemistry ◽  
Tricarboxylic Acid Cycle ◽  
Tricarboxylic Acid ◽  
Short Term ◽  
Planetary Evolution ◽  
Broad Overview

Abstract Carboxylation and decarboxylation are two fundamental classes of reactions that impact the cycling of carbon in and on Earth’s crust. These reactions play important roles in both long-term (primarily abiotic) and short-term (primarily biotic) carbon cycling. Long-term cycling is important in the subsurface and at subduction zones where organic carbon is decomposed and outgassed or recycled back to the mantle. Short-term reactions are driven by biology and have the ability to rapidly convert CO2 to biomass and vice versa. For instance, carboxylation is a critical reaction in primary production and metabolic pathways like photosynthesis in which sunlight provides energy to drive carbon fixation, whereas decarboxylation is a critical reaction in metabolic pathways like respiration and the tricarboxylic acid cycle. Early life and prebiotic chemistry on Earth likely relied heavily upon the abiotic synthesis of carboxylic acids. Over time, life has diversified (de)carboxylation reactions and incorporated them into many facets of cellular metabolism. Here we present a broad overview of the importance of carboxylation and decarboxylation reactions from both abiotic and biotic perspectives to highlight the importance of these reactions and compounds to planetary evolution.

Sign in / Sign up

Export Citation Format

Share Document