Environmetrics最新文献

The functional single index models are widely used to describe the nonlinear relationship between a scalar response and a functional predictor. The conventional functional single index model assumes that the coefficient function is nonzero in the entire time domain. In other words, the functional predictor always has a nonzero effect on the response all the time. We propose a new compact functional single index model, in which the coefficient function is only nonzero in a subregion. We also propose an efficient method that can simultaneously estimate the nonlinear link function, the coefficient function and also the nonzero region of the coefficient function. Hence, our method can identify the region in which the functional predictor is related to the response. Our method is illustrated by an application example in which the total number of daily bike rentals is predicted based on hourly temperature data. The finite sample performance of the proposed method is investigated by comparing it to the conventional functional single index model in a simulation study

Capture-recapture (CR) data and corresponding models have been used extensively to estimate the size of wildlife populations when detection probability is less than 1. When the locations of traps or cameras used to capture or detect individuals are known, spatially-explicit CR models are used to infer the spatial pattern of the individual locations and population density. Individual locations, referred to as activity centers (ACs), are defined as the locations around which the individuals move. These ACs are typically assumed to be independent, and their spatial pattern is modeled using homogeneous Poisson processes. However, this assumption is often unrealistic, since individuals can interact with each other, either within a species or between different species. In this article, we consider a vector of point processes from the general class of interaction point processes and develop a model for CR data that can account for interactions, in particular repulsions, between and within multiple species. Interaction point processes present a challenge from an inferential perspective because of the intractability of the normalizing constant of the likelihood function, and hence standard Markov chain Monte Carlo procedures to perform Bayesian inference cannot be applied. Therefore, we adopt an inference procedure based on the Monte Carlo Metropolis Hastings algorithm, which scales well when modeling more than one species. Finally, we adopt an inference method for jointly sampling the latent ACs and the population size based on birth and death processes. This approach also allows us to adaptively tune the proposal distribution of new points, which leads to better mixing especially in the case of non-uniformly distributed traps. We apply the model to a CR data-set on leopards and tigers collected at the Corbett Tiger Reserve in India. Our findings suggest that between species repulsion is stronger than within species, while tiger population density is higher than leopard population density at the park.

In recent years, immense amounts of data have been generated, from sensors to purchase transaction records, mobile GPS signals, digital satellite images, and social media. The raise of data collection has brought the need for quantitative minded professionals able to transform that data into information and decision making. In this opinion piece, we will share some of our views and experiences about the general role that data science plays nowadays, with a special interest in the field of environmetrics. We will include a limited number of examples that highlight the usefulness of data science in environmetrics, and a specific illustration of the behavior of the wildfires in Brazil between January and December of 2021.

There has been a great deal of recent interest in the development of spatial prediction algorithms for very large datasets and/or prediction domains. These methods have primarily been developed in the spatial statistics community, but there has been growing interest in the machine learning community for such methods, primarily driven by the success of deep Gaussian process regression approaches and deep convolutional neural networks. These methods are often computationally expensive to train and implement and consequently, there has been a resurgence of interest in random projections and deep learning models based on random weights—so called reservoir computing methods. Here, we combine several of these ideas to develop the random ensemble deep spatial (REDS) approach to predict spatial data. The procedure uses random Fourier features as inputs to an extreme learning machine (a deep neural model with random weights), and with calibrated ensembles of outputs from this model based on different random weights, it provides a simple uncertainty quantification. The REDS method is demonstrated on simulated data and on a classic large satellite data set.

Heavy rainfall distributional modeling is essential in any impact studies linked to the water cycle, for example, flood risks. Still, statistical analyses that both take into account the temporal and multivariate nature of extreme rainfall are rare, and often, a complex de-clustering step is needed to make extreme rainfall temporally independent. A natural question is how to bypass this de-clustering in a multivariate context. To address this issue, we introduce the stable sums method. Our goal is to incorporate time and space extreme dependencies in the analysis of heavy tails. To reach our goal, we build on large deviations of regularly varying stationary time series. Numerical experiments demonstrate that our novel approach enhances return levels inference in two ways. First, it is robust concerning time dependencies. We implement it alike on independent and dependent observations. In the univariate setting, it improves the accuracy of confidence intervals compared to the main estimators requiring temporal de-clustering. Second, it thoughtfully integrates the spatial dependencies. In simulation, the multivariate stable sums method has a smaller mean squared error than its component-wise implementation. We apply our method to infer high return levels of daily fall precipitation amounts from a national network of weather stations in France.

Climate change stands to have a profound impact on human society, and on political and other conflicts in particular. However, the existing literature on understanding the relation between climate change and societal conflicts has often been criticized for using data that suffer from sampling and other biases, often resulting from being too narrowly focused on a small region of space or a small set of events. These studies have likewise been critiqued for not using suitable statistical tools that ($��i�$) address spatio-temporal dependencies, ($��i�i�$) obtain probabilistic uncertainty quantification, and ($��i�i�i�$) lead to consistent statistical inferences. In this article, we propose a Bayesian framework to address these challenges. We find that there is a strong and substantial association between temperature anomalies on aggregated material conflicts and verbal conflicts globally. Going deeper, we also find significant evidence to suggest that positive temperature anomalies are associated with social conflict primarily through government-civilian and government-rebel material conflicts, as in civilian protests, rebel attacks against government resources, or acts of state repression. We find that majority of the conflicts associated with climate anomalies are triggered by rebel actors, and others react to such acts of conflict. Our results exhibit considerably nuanced relationships between temperature deviations and social conflicts that have not been noticed in previous studies. Methodologically, the proposed Bayesian framework can help social scientists explore similar domains involving large-scale spatial and temporal dependencies. Our code and a synthetic dataset has been made publicly available.

The Kalman filter is a workhorse of dynamical modeling. But there are challenges when using the Kalman filter in environmental science: the complexity of environmental processes, the complicated and irregular nature of many environmental datasets, and the scale of environmental datasets, which may comprise many thousands of observations per time-step. We show how these challenges can be met within the Kalman filter, identifying some situations which are relatively easy to handle, such as datasets which are high-resolution in time, and some which are hard, like areal observations on small contiguous polygons. Overall, we conclude that many applications in environmental science are within the scope of the Kalman filter, or its generalizations.

Within the statistical climatology literature, inferring the contributions of potential causes with regard to climate change has become a recurrent research theme during this last decade. In particular, disentangling human induced (anthropogenic) forcings from natural causes represents a nontrivial statistical task, especially when the focal point moves away from mean behaviors and goes towards extreme events with high societal impacts. Most studies found in the field of extreme event attributions (EEA) rely on extreme value theory. Under this theoretical umbrella, it is often assumed that, for a given location, temporal changes in extremes can be detected in both location and scale parameters of an extreme value distribution, while its shape parameter remains unchanged over time. This assumption of constant tail shape parameters between a so-called factual world (all forcings) and a counterfactual one (without anthropogenic forcing) can be challenged due to the fact that important forcing changes could impact large scale atmospheric and oceanic circulation patterns, and consequently, the latter can reshape the full distribution, including its shape parameter that drives extremal behavior. In this article, we study how allowing different tail shape parameters between the factual and counterfactual worlds can affect the analysis of records. In particular, we extend the work of Naveau et al. in which this case was not treated. We also add properties and theoretical inferential results about records in EEA and propose a procedure for model validation. A simulation study of our approach is detailed. Our method is applied to records of yearly maxima of daily maxima of near-surface air temperature issued from the numerical climate model CNRM-CM6-1 of Météo-France.

Sequential sampling occurs when an entire population is unknown in advance and data are received one by one or in groups of units. This article proposes a new algorithm to sequentially select a balanced sample. The algorithm respects equal and unequal inclusion probabilities. The method can also be used to select a spatially balanced sample if the population of interest contains spatial coordinates. A simulation study is proposed, and the results show that the proposed method outperforms other methods.