Search

The Tangni landslide in Chamoli, India, has experienced several landslide incidents in the recent past. Due to the fatalities and injuries caused, it is essential to predict slope movements at this site. A recent approach to predicting slope movements is via machine-learning algorithms. In machine learning literature, recurrent neural networks (simple LSTMs, stacked LSTMs, bidirectional LSTMs, convolutional LSTMs, CNN-LSTMs, and encoder–decoder LSTMs) and non-recurrent neural networks (multilayer perceptrons) have been proposed. However, evaluating recurrent and non-recurrent neural networks for real-world slope movements prediction has been less explored. This research’s primary objective is to develop and evaluate novel recurrent and non-recurrent neural network algorithms in their ability to predict slope movements. We used two years’ weekly data of slope movements from the Tangni landslide site in Chamoli, India. Different recurrent and non-recurrent neural networks were calibrated on the training data and then predicted the test data. Different hyperparameters (epochs; packet shuffle; look-back period; the number of nodes per layer; and the number of layers) were calibrated to training data. Later, the developed models were evaluated on test data. Results revealed that, during training, the recurrent stacked LSTMs and bidirectional LSTMs performed the best and second-best, respectively, compared to other recurrent and non-recurrent neural networks. However, during the test, the recurrent CNN-LSTMs and simple LSTMs performed best and second best, respectively, compared to other recurrent and non-recurrent neural networks. We discuss the implications of our results for predicting slope movements at real-world landslide sites.

Air pollution causes a number of pulmonary and cardiovascular diseases. Recording of air pollution via real-time low-cost IoT-based monitoring systems and its subsequent forecasting are likely to help timely warn people about prevailing air pollution across a large number of sites. In this paper, we propose and compare a real-time low-cost IoT-based air pollution monitoring system against an existing, accurate, and expensive industry-grade system. Furthermore, we undertake the task of predicting the accurate values of the industry-grade system from values recorded by the low-cost system. For forecasting, a Vector Autoregressive (VAR) model, a Vector Autoregressive Moving Average (VARMA) model, a Seasonal Autoregressive Integrated Moving Average with Exogenous variable (SARIMAX) model, and a weighted ensemble model of VAR, VARMA, and SARIMAX models were trained and tested on particular matter data. Data for forecasting were collected from the low-cost monitoring system and the industry-grade system over a period of time. Results revealed that the low-cost monitoring system predicted the values of the industry-grade system accurately. Furthermore, the ensemble model performed the best among all models in forecasting of accurate particular matter values of the industry-grade system by using the output of the low-cost system. We highlight the implication of using low-cost systems for monitoring of air pollution.

Machine learning (ML) proposes an extensive range of techniques, which could be applied to forecasting soil movements using historical soil movements and other variables. For example, researchers have proposed recurrent ML techniques like the long short-term memory (LSTM) models for forecasting time series variables. However, the application of novel LSTM models for forecasting time series involving soil movements is yet to be fully explored. The primary objective of this research is to develop and test a new ensemble LSTM technique (called “Bidirectional-Stacked-LSTM” or “BS-LSTM”). In the BS-LSTM model, forecasts of soil movements are derived from a bidirectional LSTM for a period. These forecasts are then fed into a stacked LSTM to derive the next period’s forecast. For developing the BS-LSTM model, datasets from two real-world landslide sites in India were used: Tangni (Chamoli district) and Kumarhatti (Solan district). The initial 80% of soil movements in both datasets were used for model training and the last 20% of soil movements in both datasets were used for model testing. The BS-LSTM model’s performance was compared to other LSTM variants, including a simple LSTM, a bidirectional LSTM, a stacked LSTM, a CNN-LSTM, and a Conv-LSTM, on both datasets. Results showed that the BS-LSTM model outperformed all other LSTM model variants during training and test in both the Tangni and Kumarhatti datasets. This research highlights the utility of developing recurrent ensemble models for forecasting soil movements ahead of time.

Landslides plague the Himalayan region, and landslide occurrence is widespread in hilly areas. Thus, it is important to predict soil movements and associated landslide events in advance of their occurrence. A recent approach to predicting soil movements is to use machine-learning techniques. In machine-learning literature, both moving-average-based methods (Seasonal Autoregressive Integrated Moving Average (SARIMA) model and Autoregressive (AR) model) and support-vector-based methods (Sequential Minimal Optimization; SMO) have been proposed. However, an evaluation of these methods on real-world landslide prediction has been little explored. The primary objective of this paper is to compare SARIMA, AR, and SMO methods in their ability to predict soil movements recorded at a real-world landslide site. A SARIMA model, an AR model, and a SMO model were compared in their ability to predict soil movements (in degrees) at the Tangni landslide in Chamoli, India. Time-series data about soil movements from five-sensors placed on the Tangni landslide hill were collected daily over a 78-week period from July 2012 to July 2014. Different model parameters were calibrated to the training data (first 62-weeks) and then made to predict the test data (the last 16-weeks). Results revealed that the moving-average models (SARIMA and AR) performed better compared to the support-vector models (SMO) during both training and test. Specifically, the SARIMA model possessed the smallest error compared to the AR and SMO models during test. We discuss the implications of using moving-average methods in predicting soil movements and associated landslides at real-world landslide locations.

A landslide, that is, collapse of a mass of earth or rock from a mountain or cliff, is a common phenomenon in hills. Landslides pose a large threat to life and infrastructure and there is a need to develop low‐cost sensing frameworks that could help in monitoring landslides and alert people before they occur. Certain existing technologies have been used for monitoring landslides (e.g., the use of unmanned aerial vehicle‐based remote sensing). The I nternet of T hings ( IoT ) technologies could provide alternate solutions for monitoring landslides. However, existing IoT technologies are diverse and expensive to use. Thus, there is a need for developing low‐cost IoT frameworks for monitoring landslides, especially in developing countries. This chapter proposes a microelectromechanical system ( MEMS )‐based IoT framework for sensing landslides at the lab‐scale. The proposed IoT framework offers a promising low‐cost solution for monitoring landslides with a large deployment potential in landslide‐prone areas.

Landslide is a natural calamity which has devastating effects that can lead to multiple deaths and loss of property. Thus, landslide monitoring is of substantial importance in reducing the catastrophic effects of landslides. In this paper, we propose a landslide monitoring system based on Raspberry Pi implementing IOT using a video camera. It performs real-time analysis of the region, based on the video stream acquired by the camera and applies computer vision algorithms to detect landslide and notify stakeholders via Android app. Raspberry Pi being a low-cost device with low power demands can be installed in any region.

Landslide experts have developed very detailed landslide hazard maps for different parts of Himalayas in India. These maps indicate the types of damage possible and the probabilities of adverse events. Six categories of risk severity are defined on the maps, ranging from severe risk to very low risk. Based on the existing maps, we selected respondents for a survey, some from areas high in risk and others from low-risk regions. Respondents answered several questions related to landslide risk perception and preparedness. Survey results showed a lack of awareness about the scientific causes of landslides among Mandi residents. Most of the respondents were of the belief that they lived at a safe place. Survey results suggested that many inhabitants did not know that landslide hazard maps existed for their region and most of them were not able to understand them. People overestimated the risks associated with landslides. Consequently, some people were more worried of landslides than was justified by the facts. Another important finding was that since catastrophic landslides are rare events, most of the people were risk averse. These people did not show prevention behaviour, and they were not well prepared for an adverse event. Furthermore, results suggest that respondents’ experiences with landslides were positively related to their perceptions of landslide risk. Thus, findings of the present study comply with the concept of availability heuristic.

Landslides cause lots of damage to life and property world over. There has been research in machine-learning that aims to predict landslides based on the statistical analysis of historical landslide events and its triggering factors. However, prediction of landslides suffers from a class-imbalance problem as landslides and land-movement are very rare events. In this paper, we apply state-of-the-art techniques to correct the class imbalance in landslide datasets. More specifically, to overcome the class-imbalance problem, we use different synthetic and oversampling techniques to a real-world landslide data collected from the Chandigarh - Manali highway. Also, we apply several machine-learning algorithms to the landslide data set for predicting landslides and evaluating our algorithms. Different algorithms have been assessed using techniques like the area under the ROC curve (AUC) and sensitivity index (d'). Results suggested that random forest algorithm performed better compared to other classification techniques like neural networks, logistic regression, support vector machines, and decision trees. Furthermore, among class-imbalance methods, the Synthetic Minority Oversampling Technique with iterative partitioning filter (SMOTE-IPF) performed better than other techniques. We highlight the implications of our results and methods for predicting landslides in the real world.

To investigate how differing amounts of experiential feedback and feedback’s availability in an interactive simulation tool influences people’s decision-making against landslide risks. Feedback via simulation tools is likely to help people improve their decisions against disasters; however, currently little is known on how differing amounts of experiential feedback and feedback’s availability in simulation tools influences people’s decisions against landslides. We tested the influence of differing amounts of experiential feedback and feedback’s availability on people’s decisions against landslide risks in an Interactive Landslide Simulation (ILS) tool. In an experiment, in high-damage conditions, the probabilities of damages to life and property due to landslides were 10-times higher than those in the low-damage conditions. In feedback-present condition, experiential feedback was provided in numeric, text, and graphical formats in ILS. In feedback-absent conditions, the probabilities of damages were described; however, there was no experiential feedback present. Investments were greater in conditions where experiential feedback was present and damages were high compared to conditions where experiential feedback was absent and damages were low. Furthermore, only high-damage feedback produced learning in ILS. Experience gained in ILS enables people to improve their decision-making against landslide risks. Simulation tools seem appropriate for landslide risk communication and for performing what-if analyses.

Understanding landslide risks is important for people living in hilly areas in India. A promising way of communicating landslide risks is via simulation tools, where these tools integrate both human factors (e.g., public investments to mitigate landslides) and environmental factors (e.g., spatial geology and rainfall). In this paper, we develop an interactive simulation model on landslide risks and use it to design a web-based Interactive Landslide Simulator (ILS) microworld. The ILS microworld is based on the assumption that landslides occur due to both environmental factors (spatial geology and rainfall) as well as human factors (lack of monetary investments to mitigate landslides). We run a lab-based experiment involving human participants performing in ILS and we show that the ILS performance helps improve public understanding of landslide risks. Overall, we propose ILS to be an effective tool for doing what-if analyses by policymakers and for educating public about landslide risks.

We investigated a far-field superlens operating at mid-infrared wavelength that allows resolving subwavelength features in the far-field. By utilizing evanescent enhancement provided by surface plasmon excitation of silver nanorods and Moire effect, we numerically demonstrated that subwavelength information of an object can be converted to propagating information. This information can then be captured by conventional optical components. A simple image reconstruction algorithm can restore the subwavelength object. A sub-diffraction-limited resolution of 2.5 μm at 6-μm wavelength is demonstrated.

Cathodoluminescence spectroscopy has been performed on silver nanoparticles in a scanning electron microscopy setup. Peaks appearing in the visible range for particles fabricated on silicon substrate are shown to arrive from excitation of out-of-plane eigenmodes by the electron beam. Monochromatic emission maps have been shown to resolve spatial field variation of resonant plasmon mode on length scale smaller than 25 nm. Finite-difference time-domain numerical simulations are performed for both the cases of light excitation and electron excitation. The results of radiative emission under electron excitation show an excellent agreement with experiments. A complete vectorial description of induced field is given, which complements the information obtained from experiments.

We demonstrate an effective method for depositing smooth silver (Ag) films on SiO(2)/Si(100) substrates using a thin seed layer of evaporated germanium (Ge). The deposited Ag films exhibit smaller root-mean-square surface roughness, narrower peak-to-valley surface topological height distribution, smaller grain-size distribution, and smaller sheet resistance in comparison to those of Ag films directly deposited on SiO(2)/Si(100) substrates. Optically thin ( approximately 10-20 nm) Ag films deposited with approximately 1-2 nm Ge nucleation layers show more than an order of magnitude improvement in the surface roughness. The presence of the thin layer of Ge changes the growth kinetics (nucleation and evolution) of the electron-beam-evaporated Ag, leading to Ag films with smooth surface morphology and high electrical conductivity. The demonstrated Ag thin films are very promising for large-scale applications as molecular anchors, optical metamaterials, plasmonic devices, and several areas of nanophotonics.

Artificially engineered metamaterials have emerged with properties and functionalities previously unattainable in natural materials. The scientific breakthroughs made in this new class of electromagnetic materials are closely linked with progress in developing physics-driven design and novel parallel fabrication methods. For example, a smooth superlens has been demonstrated with 30-nm imaging resolution, or 1/12 of the corresponding wavelength, far below the diffraction limit. Similarly, a photoswitchable optical negative-index material has been printed, showing a remarkable tuning range of refractive index in the communication wavelength. New frontiers are being explored as intrinsic limitations challenge the scaling of microwave metamaterial designs to optical frequencies. These novel metamaterials promise an entire new generation of passive and active optical elements, such as paper-thin superlenses and modulators.

This study demonstrates application of analysis of geotechnical sensors based data for landslide monitoring and prediction at Tangni landslide site, 12 km from Pipalkoti along Rishikesh — Badrinath highway (NH-58) in India. Prediction of slope deformations is carried out using back propagation neural network (BPNN) model. Training of the model is done using 60% of the data received from the landslide site while 20% of data is used for testing and finally 20% data is used for validation of the model. The application of the BPNN based model is to predict the slope deformations using daily and antecedent rainfall as input variables. It is found using hit and trial method that a back propagation neural network having four layers with two hidden layers and twelve neurons is performing in an optimal manner on data received from Tangni landslide. The results of the study suggest a promising BPNN model that produces an accuracy of approximately 95% in predicting the slope deformation.

Landslide is a common geological disaster, which seriously threatens people's safety and property. The landslide monitoring has great significance in reducing loss due to landslide. This paper introduces a real time landslide monitoring method using computer vision technologies. Our method uses a fixed video camera and a processor which is inexpensive yet efficient and is semi automatic. The method is based on background subtraction based moving object detection technique and entropy thresholding. Entropy is calculated for automatic landslide detection. We have also introduced the measurement of landslide movement's velocity. Velocity of landslide movement is an important factor in risk assessment. An alarm message is generated when a landslide event is detected which can transmitted as a real time SMS text massage.

We describe in this chapter development of plasmonic optical antennas for light concentration and near-field enhancement. A set of bow tie nanoantennas are fabricated and characterized with optical and electron excitation methods. Optical spectroscopy of these subwavelength antennas displays pronounced extinction peaks at resonant wavelength, showing total extinction cross sections as much as 10 times of their physical dimensions. On the other hand, coherent excitation of the bow tie antennas allows tuning the peak wavelength of the scattered light by changing the periodicity. Under dark-field microscopy, we observed the scattered waves from arrays of different bow tie antennas in complete visible spectrum. The local resonant modes of the bow tie antennas are also probed by focused electrons. Such cathodoluminescence spectroscopy reveals the fine details of enhanced field on the optical nanoantennas at resolution down to 20 nm. Finally, we show examples of surface-enhanced Raman spectroscopy on the nanoantennas. Effective designs based on local enhancement and radiation engineering of the plasmonic optical antennas would promise revolutionary changes in highly compact and integrated photonics for photon energy conversion, adaptive sensing, and image processing.

Nanoimprint lithography is used to fabricate a metamaterial with the "fishnet" structure composed of Ag/a-Si/Ag layers that exhibits negative refractive index in the near-infrared. We have carried out a femtosecond pump-probe experiment to measure the transient photo-induced response of this structure. With a pump fluence of 330μJ/cm 2 at 800nm, the transmission at the magnetic resonance is increased by ~15.4%. The induced change originated from carrier excitation in the a-Si layer has a fast decay constant of 1.1ps.

We show by pump-probe spectroscopy that the optical response of a fishnet metamaterial can be modulated on the femtosecond time scale. The modulation dynamics is dominated by pump-induced changes in the constituting dielectric medium, but the strength of modulation is dramatically enhanced through the plasmon resonance. The pump-induced spectral responses of the metamaterial provide understanding on how the resonance is modified by pump excitation. Our study suggests that metamaterials can be used as high-speed amplitude/phase modulators with terahertz-bandwidth.

The pressing need for energy conversion devices with higher efficiency and the pursuit for nano-scale sensing have called for intense research in the use of plasmonic nanostructures. In this work, we discuss fabrication and characterization of bowtie antennas with gaps down to 10 nm for optical frequencies. With structures as simple as equilateral triangles, we observed an extinction cross section of ten times the geometrical cross section, which is twice the value previously reported. Finally, we measured the scattering profile of a series of bowtie antennas with increased size using confocal microscopy.

We discuss the electrical resistivity and thermal stability of smooth thin silver (Ag) films characterized by significantly lower root mean square (RMS) surface roughness, peak-to-valley height distribution, and grain-size distribution. The smooth film was prepared by using a very thin (~1-2 nm) nucleation layer of evaporated germanium (Ge) prior to Ag evaporation. The sheet resistance of the rough silver (10 nm Ag) film is RAg-rough ~51 Omega/sq compared to the smooth silver (10 nm Ag/2 nm Ge) film with RAg.smooth ~22 Omega/sq. The smooth Ag is also shown to be thermally more stable for temperatures less than 120degC.

We investigate a novel design of negative index metamaterial modulator integrated on optical fiber. Numerical studies indicate strong modulation of fiber-guided optical signal, when the resonance frequency of the metamaterial is detuned with optical excitation.

It has been experimentally demonstrated that a single layer of silver functions as a “superlens” [Fang et al, Science 308, 534 (2005)], providing image resolution much better than the diffraction limit. Resolution as high as 60 nanometer (λ/6) half-pitch was achieved. In this paper, we explore the possibility of further refining the image resolution using a “multilayer superlens” design. With optimized design of silver-alumina multilayer superlens, our numerical simulations show a feasibility of resolving 15nm features, about 1/26th of the illumination wavelength. We present preliminary experimental results targeted towards achieving the molecular scale imaging resolution. The development of potential low-loss and high resolution superlens opens the door to exciting applications in nanoscale optical metrology and nanomanufacturing.

We demonstrate a smooth and low loss silver (Ag) optical superlens capable of resolving features at 1/12th of the illumination wavelength with high fidelity. This is made possible by utilizing state-of-the-art nanoimprint technology and intermediate wetting layer of germanium (Ge) for the growth of flat silver films with surface roughness at subnanometer scales. Our measurement of the resolved lines of 30 nm half-pitch shows a full-width at half-maximum better than 37 nm, in excellent agreement with theoretical predictions. The development of this unique optical superlens leads promise to parallel imaging and nanofabrication in a single snapshot.

We study the relation between near-field enhancement and far-field extinction spectra of three-fold rotationally symmetric nanopatterns consisting of metal nano-triangles. Our symmetry analysis benefits the understanding of the plasmon resonance phenomena in these nanopatterns.