A stochastic contact network model for assessing outbreak risk of COVID-19 in workplaces

The COVID-19 pandemic has drastically shifted the way people work. While many businesses can operate remotely, a large number of jobs can only be performed on-site. Moreover as businesses create plans for bringing workers back on-site, they are in need of tools to assess the risk of COVID-19 for their employees in the workplaces. This study aims to fill the gap in risk modeling of COVID-19 outbreaks in facilities like offices and warehouses. We propose a simulation-based stochastic contact network model to assess the cumulative incidence in workplaces. First-generation cases are introduced as a Bernoulli random variable using the local daily new case rate as the success rate. Contact networks are established through randomly sampled daily contacts for each of the first-generation cases and successful transmissions are established based on a randomized secondary attack rate (SAR). Modification factors are provided for SAR based on changes in airflow, speaking volume, and speaking activity within a facility. Control measures such as mask wearing are incorporated through modifications in SAR. We validated the model by comparing the distribution of cumulative incidence in model simulations against real-world outbreaks in workplaces and nursing homes. The comparisons support the model’s validity for estimating cumulative incidences for short forecasting periods of up to 15 days. We believe that the current study presents an effective tool for providing short-term forecasts of COVID-19 cases for workplaces and for quantifying the effectiveness of various control measures. The open source model code is made available at github.com/abhineetgupta/covid-workplace-risk.

Unsupervised Domain Adaptation by Statistics Alignment for Deep Sleep Staging Networks

Deep sleep staging networks have reached top performance on large-scale datasets. However, these models perform poorer when training and testing on small sleep cohorts due to data inefficiency. Transferring well-trained models from large-scale datasets (source domain) to small sleep cohorts (target domain) is a promising solution but still remains challenging due to the domain-shift issue. In this work, an unsupervised domain adaptation approach, domain statistics alignment (DSA), is developed to bridge the gap between the data distribution of source and target domains. DSA adapts the source models on the target domain by modulating the domain-specific statistics of deep features stored in the Batch Normalization (BN) layers. Furthermore, we have extended DSA by introducing cross-domain statistics in each BN layer to perform DSA adaptively (AdaDSA). The proposed methods merely need the well-trained source model without access to the source data, which may be proprietary and inaccessible. DSA and AdaDSA are universally applicable to various deep sleep staging networks that have BN layers. We have validated the proposed methods by extensive experiments on two state-of-the-art deep sleep staging networks, DeepSleepNet+ and U-time. The performance was evaluated by conducting various transfer tasks on six sleep databases, including two large-scale databases, MASS and SHHS, as the source domain, four small sleep databases as the target domain. Thereinto, clinical sleep records acquired in Huashan Hospital, Shanghai, were used. The results show that both DSA and AdaDSA could significantly improve the performance of source models on target domains, providing novel insights into the domain generalization problem in sleep staging tasks.<br>

Modelling of the Ability of a Mixed Renewable Generation Electricity System with Storage to Meet Consumer Demand

In this paper, we explore how effectively renewable generation can be used to meet a country’s electricity demands. We consider a range of different generation mixes and capacities, as well as the use of energy storage. First, we introduce a new open-source model that uses hourly wind speed and solar irradiance data to estimate the output of a renewable electricity generator at a specific location. Then, we construct a case study of the Great Britain (GB) electricity system as an example using historic hourly demand and weather data. Three specific sources of renewable generation are considered: offshore wind, onshore wind, and solar PV. Li-ion batteries are considered as the form of electricity storage. We demonstrate that the ability of a renewables-based electricity system to meet expected demand profiles can be increased by optimising the ratio of onshore wind, offshore wind and solar PV. Additionally, we show how including Li-ion battery storage can reduce overall generation needs, therefore lowering system costs. For the GB system, we explore how the residual load that would need to be met with other forms of flexibility, such as dispatchable generation sources or demand-side response, varies for different ratios of renewable generation and storage.

The Source Characteristics of the Mw6.4, 2016 Meinong Taiwan Earthquake from Teleseismic Data Using the Hybrid Homomorphic Deconvolution Method

The kinematic source rupture process of the 2016 Meinong earthquake (Mw = 6.4) in Taiwan was derived from apparent source time functions retrieved from teleseismic S-waves by using a refined homomorphic deconvolution method. The total duration of the rupture process was approximately 15 s, and one slip-concentrated area can be represented as the source model based on images representing static slip distribution. The rupture process began in a down-dip direction from the fault toward Tainan City, strongly suggesting that the rupture had a unilateral northwestern direction. The asperity with an area of approximately 15 × 15 km2 and the maximum slip of approximately 2 m were centered 12.8 km northwest of the hypocenter. Coseismic vertical deformation was calculated based on the source model. Compared with the results derived from InSAR (Interferometric Synthetic Aperture Radar) data, our results demonstrated that the location with maximum uplift was accurately well detected, but our maximum value was just approximately 0.4 times of the InSAR-derived value. It reveals that there are the other mechanisms to affect the vertical deformation, rather than only depending on the source model. At different depths, areas west, east, and north of the hypocenter maintained high values of Coulomb stress changes. This explains the mechanism behind aftershocks being triggered and provides a reference for predicting aftershock locations after a large earthquake. The estimated seismic spectral intensities, including spectral acceleration and velocity intensity (SIa and SIv), were derived. Source directivity effects caused damage to buildings, and we concluded that all damaged buildings were located within a SIa value of 400 gal. Destroyed buildings taller than seven floors were located in an area with a SIv value of 30 cm/s. These observations agree with those on damages caused by the 2010 Jiasian earthquake (ML 6.4) in Tainan, Taiwan.