Explore the vast range of titles available.

Invasive alien species are a growing threat to biodiversity, and identifying the mechanisms that enable these species to establish viable populations in their new environment is paramount for management of the problems they pose. Using an unusually large number of both failed and successful documented introductions of parakeets (Aves: Psittacidae) in Europe, we test two of the major hypotheses on the establishment success of invading species, namely the climate-matching and the human-activity hypothesis. European human population centres where ring-necked parakeet (Psittacula krameri) and/or monk parakeet (Myiopsitta monachus) introductions have occurred. Data on ring-necked and monk parakeet introductions in Europe were gathered from various sources, including published books and articles, but also from unpublished reports and local grey literature. Information was verified with experts from the region under consideration. In order to test the climate-matching hypothesis, we verified whether the climatic factors that determine the parakeets' native ranges also explain establishment success in Europe. Parakeet occurrence data from the native ranges were analysed using the presence-only modelling method M axent, and correlations between parakeet establishment and climatic and anthropogenic variables in Europe were assessed using both stepwise logistic regression and the information-theoretic model selection approach. The establishment success of ring-necked and monk parakeets was found to be positively associated with human population density, and, both in the native and in the introduced regions, parakeet occurrence was negatively correlated with the number of frost days. Thus, parakeets are more likely to establish in warmer and human-dominated areas. The large number of independent parakeet introductions in Europe allows us to test the often-used climate-matching and human-activity hypotheses at the species level. We show that both hypotheses offer insight into the invasion process of monk and ring-necked parakeets. Our results suggest that, in the future, parakeet establishment probability may increase even further because global warming is likely to cause a decrease in the number of frost days and because urbanization and human populations are still increasing.

Multiple cameras are used to resolve occlusion problem that often occur in single-view human activity recognition. Based on the success of learning representation with deep neural networks (DNNs), recent works have proposed DNNs models to estimate human activity from multi-view inputs. However, currently available datasets are inadequate in training DNNs model to obtain high accuracy rate. Against such an issue, this study presents a DNNs model, trained by employing transfer learning and shared-weight techniques, to classify human activity from multiple cameras. The model comprised pre-trained convolutional neural networks (CNNs), attention layers, long short-term memory networks with residual learning (LSTMRes), and Softmax layers. The experimental results suggested that the proposed model could achieve a promising performance on challenging MVHAR datasets: IXMAS (97.27%) and i3DPost (96.87%). A competitive recognition rate was also observed in online classification.

Atmospheric organic aerosol (OA) has important impacts on climate and human health but its sources remain poorly understood. Biogenic monoterpenes and sesquiterpenes are important precursors of secondary organic aerosol (SOA), but the amounts and pathways of SOA generation from these precursors are not well constrained by observations. We propose that the less-oxidized oxygenated organic aerosol (LO-OOA) factor resolved from positive matrix factorization (PMF) analysis on aerosol mass spectrometry (AMS) data can be used as a surrogate for fresh SOA from monoterpenes and sesquiterpenes in the southeastern US. This hypothesis is supported by multiple lines of evidence, including lab-in-the-field perturbation experiments, extensive ambient ground-level measurements, and state-of-the-art modeling. We performed lab-in-the-field experiments in which the ambient air is perturbed by the injection of selected monoterpenes and sesquiterpenes, and the subsequent SOA formation is investigated. PMF analysis on the perturbation experiments provides an objective link between LO-OOA and fresh SOA from monoterpenes and sesquiterpenes as well as insights into the sources of other OA factors. Further, we use an upgraded atmospheric model and show that modeled SOA concentrations from monoterpenes and sesquiterpenes could reproduce both the magnitude and diurnal variation of LO-OOA at multiple sites in the southeastern US, building confidence in our hypothesis. We estimate the annual average concentration of SOA from monoterpenes and sesquiterpenes in the southeastern US to be roughly 2 µg m−3.

Human Activity Recognition (HAR) has been studied extensively, yet current approaches are not capable of generalizing across different domains (i.e., subjects, devices, or datasets) with acceptable performance. This lack of generalization hinders the applicability of these models in real-world environments. As deep neural networks are becoming increasingly popular in recent work, there is a need for an explicit comparison between handcrafted and deep representations in Out-of-Distribution (OOD) settings. This paper compares both approaches in multiple domains using homogenized public datasets. First, we compare several metrics to validate three different OOD settings. In our main experiments, we then verify that even though deep learning initially outperforms models with handcrafted features, the situation is reversed as the distance from the training distribution increases. These findings support the hypothesis that handcrafted features may generalize better across specific domains.

Patterns of deliberate human activity and behavior are of utmost importance in areas as diverse as disease spread, resource allocation, and emergency response. Because of its widespread availability and use, e-mail correspondence provides an attractive proxy for studying human activity. Recently, it was reported that the probability density for the inter-event time τ between consecutively sent e-mails decays asymptotically as τ⁻α, with α [almost equal to] 1. The slower-than-exponential decay of the inter-event time distribution suggests that deliberate human activity is inherently non-Poissonian. Here, we demonstrate that the approximate power-law scaling of the inter-event time distribution is a consequence of circadian and weekly cycles of human activity. We propose a cascading nonhomogeneous Poisson process that explicitly integrates these periodic patterns in activity with an individual's tendency to continue participating in an activity. Using standard statistical techniques, we show that our model is consistent with the empirical data. Our findings may also provide insight into the origins of heavy-tailed distributions in other complex systems.

In this article, we study activity recognition in the context of sensor-rich environments. In these environments, many different constraints arise at various levels during the data generation process, such as the intrinsic characteristics of the sensing devices, their energy and computational constraints, and their collective (collaborative) dimension. These constraints have a fundamental impact on the final activity recognition models as the quality of the data, its availability, and its reliability, among other things, are not ensured during model deployment in real-world configurations. Current approaches for activity recognition rely on the activity recognition chain which defines several steps that the sensed data undergo: This is an inductive process that involves exploring a hypothesis space to find a theory able to explain the observations. For activity recognition to be effective and robust, this inductive process must consider the constraints at all levels and model them explicitly. Whether it is a bias related to sensor measurement, transmission protocol, sensor deployment topology, heterogeneity, dynamicity, or stochastic effects, it is essential to understand their substantial impact on the quality of the data and ultimately on activity recognition models. This study highlights the need to exhibit the different types of biases arising in real situations so that machine learning models, e.g., can adapt to the dynamicity of these environments, resist sensor failures, and follow the evolution of the sensors’ topology. We propose a metamodeling approach in which these biases are specified as hyperparameters that can control the structure of the activity recognition models. Via these hyperparameters, it becomes easier to optimize the inductive processes, reason about them, and incorporate additional knowledge. It also provides a principled strategy to adapt the models to the evolutions of the environment. We illustrate our approach on the SHL dataset, which features motion sensor data for a set of human activities collected in real conditions. The obtained results make a case for the proposed metamodeling approach; noticeably, the robustness gains achieved when the deployed models are confronted with the evolution of the initial sensing configurations. The trade-offs exhibited and the broader implications of the proposed approach are discussed with alternative techniques to encode and incorporate knowledge into activity recognition models.

Water resources in river systems have been changing under the impact of both climate variability and human activities. Assessing the respective impact on decadal streamflow variation is important for water resource management. By using an elasticity-based method and calibrated TOPMODEL and VIC hydrological models, we quantitatively isolated the relative contributions that human activities and climate variability made to decadal streamflow changes in the Jinghe basin, located in the northwest of China. This is an important watershed of the Shaanxi province that supplies drinking water for a population of over 6 million people. The results showed that the maximum value of the moisture index (E0∕P) was 1.91 and appeared in 1991–2000, and the decreased speed of streamflow was higher since 1990 compared with 1960–1990. The average annual streamflow from 1990 to 2010 was reduced by 26.96 % compared with the multiyear average value (from 1960 to 2010). The estimates of the impacts of climate variability and human activities on streamflow decreases from the hydrological models were similar to those from the elasticity-based method. The maximum contribution value of human activities was 99 % when averaged over the three methods, and appeared in 1981–1990 due to the effects of soil and water conservation measures and irrigation water withdrawal. Climate variability made the greatest contribution to streamflow reduction in 1991–2000, the values of which was 40.4 %. We emphasized various source of errors and uncertainties that may occur in the hydrological model (parameter and structural uncertainty) and elasticity-based method (model parameter) in climate change impact studies.

The poleward range shift of the red fox (Vulpes vulpes) > 1,700 km into the Arctic is one of the most remarkable distribution changes of the early twentieth century. While this expansion threatens a smaller arctic ecological equivalent, the arctic fox (Vulpes lagopus), the case became a textbook example of climate‐driven range shifts. We tested this classical climate change hypothesis linked to an important range shift which has attracted little research thus far. We analysed Canadian fur harvest data from the Hudson's Bay Company Archives (14 trading posts; 1926–1950), testing hypotheses based on changes in summer and winter climates. Summer warming might have triggered a bottom‐up increase in ecosystem productivity, while winter warming might have lowered thermal stress, both favouring red fox expansion. Additionally, we evaluated the hypothesis that red fox expansion was driven by the appearance of human sedentary sites (n = 110) likely bringing food subsidies into the unproductive tundra. Analysis of red fox expansion chronologies showed that expansion speed was higher during warmer winters. However, the expansions occurred under both cooling and warming trends, being faster during cooler summers in the Baffin Island region. The increasing proportion of red fox in fox fur harvests was best explained by human activity, while generalized linear mixed models also revealed a marginal effect of warmer winters. Generalized additive models confirmed human presence as the most important factor explaining rates of change in the proportion of red fox in fox fur harvests. Using historical ecology, we disentangled the relative influences of climate change and anthropogenic habitat change, two global drivers that transformed arctic biodiversity during the last century and will likely continue to do so during this century. Anthropogenic food subsidies, which constitute stable food sources, facilitated the invasion of the tundra biome by a new mammalian predator and competitor, with long‐term consequences that still remain to be understood. This research is a critical investigation of a widely accepted climate change hypothesis explaining species range expansions. Its results show the importance of considering a wider scope of global drivers to explain observed biological change.

The ability to count finger and wrist movements throughout the day with a nonobtrusive, wearable sensor could be useful for hand-related healthcare applications, including rehabilitation after a stroke, carpal tunnel syndrome, or hand surgery. Previous approaches have required the user to wear a ring with an embedded magnet or inertial measurement unit (IMU). Here, we demonstrate that it is possible to identify the occurrence of finger and wrist flexion/extension movements based on vibrations detected by a wrist-worn IMU. We developed an approach we call “Hand Activity Recognition through using a Convolutional neural network with Spectrograms” (HARCS) that trains a CNN based on the velocity/acceleration spectrograms that finger/wrist movements create. We validated HARCS with the wrist-worn IMU recordings obtained from twenty stroke survivors during their daily life, where the occurrence of finger/wrist movements was labeled using a previously validated algorithm called HAND using magnetic sensing. The daily number of finger/wrist movements identified by HARCS had a strong positive correlation to the daily number identified by HAND (R2 = 0.76, p < 0.001). HARCS was also 75% accurate when we labeled the finger/wrist movements performed by unimpaired participants using optical motion capture. Overall, the ringless sensing of finger/wrist movement occurrence is feasible, although real-world applications may require further accuracy improvements.

Quantifying river ecohydrological conditions and their drivers is essential for protecting river ecosystems. Using runoff data from the Shimen Hydrological Station, we quantified changes in the basin's ecohydrological situation with the RVA method. Based on the ABCD model and Budyko's hypothesis, we quantified the differences in runoff drivers on time scales, such as yearly, quarterly, and monthly. The results showed that after the sudden change in the river basin in 1983, the runoff depth was reduced by 51.31 mm, and all five groups of ecohydrological indicators reached more than moderate alteration. The number of reversals was altered by 90.85%, resulting in drought impacting the bottom-mobile streamside organisms. Differences in the contribution of drivers at different time scales were more pronounced. At the annual scale, human activity was the dominant factor in runoff change; at the seasonal scale, human activity was more heavily weighted in winter, accounting for 30.5% of the total. On the monthly scale, human activities contributed more significantly in April, June, October, and December, with 82.91, 78.83, 58.01, and 97.09%, respectively, and climate change was the main driver in the rest of the months (50.26–89.64%).