Environmental and Ecological Statistics最新文献

Abstract

Environmental signals, acquired, e.g., by remote sensing, often present large gaps of missing observations in space and time. In this work, we present an innovative approach to identify the main variability patterns, in space–time data, when data may be affected by complex missing data structures. We formalize the problem in the framework of functional data analysis, proposing an innovative method of functional principal component analysis (fPCA) for incomplete space–time data. The functional nature of the proposed method permits to borrow information from measurements observed at nearby spatio-temporal locations. The resulting functional principal components are smooth fields over the considered spatio-temporal domain, and can lead to interesting insights in the spatio-temporal dynamic of the phenomenon under study. Moreover, they can be used to provide a reconstruction of the missing entries, also under severe missing data patterns. The proposed model combines a weighted rank-one approximation of the data matrix with a roughness penalty. We show that the estimation problem can be solved using a majorize–minimization approach, and provide a numerically efficient algorithm for its solution. Thanks to a discretization based on finite elements in space and B-splines in time, the proposed method can handle multidimensional spatial domains with complex shapes, such as water bodies with complicated shorelines, or curved spatial regions with complex orography. As shown by simulation studies, the proposed space–time fPCA is superior to alternative techniques for Principal Component Analysis with missing data. We further highlight the potentiality of the proposed method for environmental problems, by applying space–time fPCA to the study of the lake water surface temperature (LWST) of Lake Victoria, in Central Africa, starting from satellite measurements with large gaps. LWST is considered one of the fundamental indicators of how climate change is affecting the environment, and is recognized as an essential climate variable.

Retrospective assessment of individual exposure in occupational settings is often based on the association of individual work histories with quantitative and semi-quantitative exposure information. In the absence of exposure information, researchers have commonly used proxy variables, but with strong assumptions and some limitations. In the present work, we estimate the time-varying exposure-risk function associated with the outcomes of interest, taking into account functional regression models and individual work periods. The work was motivated by the analysis of a cohort of dock workers occupationally exposed to asbestos in Italy. We evaluated the potential of our proposal through a series of simulations. We then compared our approach with traditional methods that use exposure proxy variables.

We present ARPALData, an R package that can help international users retrieve, handle, and analyze air quality and weather data in the Lombardy region (Northern Italy). The software provides a user-friendly tool that directly inquires into the platform of the regional environmental protection agency and ensures real-time updating of information using standardized syntax. The software provides data in standard statistical formats. Eventually, all measurements, metadata, and subsequent analytical tools are provided to users in English, facilitating accessibility to international and domestic users. Data are collected from the open database of the Regional Agency for Environmental Protection of Lombardy, namely ARPA Lombardia. ARPALData returns measurements at several temporal frequencies (infra-hourly to yearly) collected through air quality and weather ground monitoring networks managed by ARPA Lombardia, as well as estimates of several pollutants at the municipal level. In addition to data download functions, ARPALData provides functions to explore, describe, analyze, and graphically represent air quality and weather data. In particular, users are provided with functions to compute key descriptive statistics and input data maps, temporally aggregate measurements, detect outliers, and study missing-value (gap length) patterns. Herein, we discuss purposes, goals, and functioning of the package, and present three guided examples and case studies in which the software is used to characterize air quality and meteorology in different settings. The examples are designed to provide a step-by-step guide for accomplished analyses using the most relevant tools included in ARPALData.

Chile is one of the most seismic countries in the world especially due to the subduction of the Nazca plate under the South America plate along the Chilean cost. Normally, the spatial distribution of seismic events tends to form spatial and temporal clusters around the main event including both precursor and aftershock events. However, it is very difficult to identify whether an event is a precursor, a main event or an aftershock. In the literature, only some large earthquakes are well described but it does not exist an automatic method to classify them. In this work, we propose a new density based clustering method, called ST-DBSCAN-EV (Space-time DBSCAN with Epsilon Variable), which allows the Epsilon parameter (the radius) to vary depending on the density of the points. The results of the ST-DBSCAN-EV are validated on three important earthquakes with magnitude greater than 8.0 Mw occurred in Chile in the last 20 years, by carrying out a series of experiments considering different combinations of parameters. A comparison with some traditional clustering techniques such as the DBSCAN, ST-DBSCAN, and the K-means has been implemented for assessing the performance of the proposed method. Almost in all cases ST-DBSCAN-EV outperformed traditional ones by providing an F1-Score metric higher than 0.8. Finally, the results of classification are compared with a declustering method.

This article provides, for the first time, direct information on the levels and trends of nitrogen oxides and particulate matter measured by a recently installed air-quality monitoring station in the city of Venice (Italy). High levels of air pollution affect human health and built cultural heritage with corrosion, loss of material due to chemical attack, and soiling: this is particularly dangerous in a World Heritage city like Venice. The pollution levels measured in the historical city are compared to those of a background station in the city of Venice and of urban and background stations in the mainland, also investigating climate factors which might affect pollution in all stations. The first results of the investigation are that the NO₂, as well as the PM₁₀, annual average levels in Venice definitely exceeded the limit values set by EU directives. This is an astonishing and unexpected result in a car free city. To contrast the poor air quality, the Venice Municipality decreed in spring 2019 to limit traffic in one of the most overcrowded Venice canals. To investigate the usefulness of the implemented policy we performed a comparative study in which Generalized Additive Models are employed to model the potential reduction in measured nitrogen dioxide in the urban station as compared to the background station. This is done for stations in the historical city of Venice and in the mainland, to give a stronger indication of whether detected changes can be attributable to the traffic policy and no other exogenous factors. The policy is found to have a minor impact in the reduction of measured nitrogen dioxide.

Scale mixtures of skew normal distributions are flexible models well-suited to handle departures from multivariate normality. This paper is concerned with the stochastic comparison of vectors that belong to the family of scale mixtures of skew normal distributions. The paper revisits some of their properties with a proposal that allows to carry out tail weight stochastic comparisons. The connections of the proposed stochastic orders with the non-normality parameters of the multivariate model are also studied for some popular distributions within the family. The role played by these parameters to tackle the non-normality of multivariate data is enhanced as a result. This work is motivated by the analysis of multivariate data in environmental studies which usually collect maximum or minimum values exhibiting departures from normality. The implications of our theoretical results in addressing the stochastic comparison of extreme environmental records is illustrated with an application to a real data study on maximum temperatures in the Iberian Peninsula throughout the last century. The resulting findings may elucidate whether extreme temperatures are evolving for such a long period.

Abstract

The partially linear varying coefficient spatial autoregressive model is a semi-parametric spatial autoregressive model in which the coefficients of some explanatory variables are variable, while the coefficients of the remaining explanatory variables are constant. For the nonparametric part, a local linear smoothing method is used to estimate the vector of coefficient functions in the model, and, to investigate its variable selection problem, this paper proposes a penalized robust regression estimation based on exponential squared loss, which can estimate the parameters while selecting important explanatory variables. A unique solution algorithm is composed using the block coordinate descent (BCD) algorithm and the concave-convex process (CCCP). Robustness of the proposed variable selection method is demonstrated by numerical simulations and illustrated by some housing data from Airbnb.

Several models have been proposed as an extension to the classical Holling’s disc equation to evaluate the predator and prey interactions and their applied aspect in biological control and population regulation of the target organisms. In a one prey and two predator dynamic system with mutual interference (m) as a quadratic parameter of predator density, an evaluation was made to the resultant impact on the prey. A simulation was carried out to see the extinction of prey and the stability of the system at origin, i.e., when all the three species are extinct. We assumed the data obtained for the interactions between the mosquito and the water bug predators that are common in the freshwater wetlands and involved in the population regulation. Despite the benefits to prey population due to interference competition, the expected extinction of prey is still observed. With varying magnitudes of m the declining growth curve of prey population, shifted. The equation proposed was also compared with Crowley–Martin functional response, and considerable differences were observed in selected instances when compared for the growth rate of the predators, in a species-specific manner. The stability of the system was deduced with eigenvalues of Jacobian matrix at origin to prove the extinction is stable. Our assessment supports the possible cooccurrence of the predators and mosquito prey in the wetlands with the mutual interference being one of the major factors.

Abstract

The paper focuses on the evaluation of hailstorms’ and thunderstorms winds’ events in the United States of America, in the period from 1996 to 2022, under the marked spatio-temporal self-exciting point processes point of view. The aim of the present article is the assessment and description of the spatio-temporal spontaneous and reproducing activity of severe hailstorms’ and thunderstorms winds’ processes. The present application shows how the spatio-temporal pattern is well-fitted and clearly explainable, according to the flexible semi-parametric ETAS model fitting.

Understanding the causality between biological variables or their related variables is beneficial in environmental or biological policy making. The usual approaches revealing the relations between them are traditional ANOVA or regression models. These models normally resort to a plethora of assumptions regarding the population, the covariance or the error distributions. Checking the validity of these assumptions might in turn rely on other batches of assumptions. This shall cause a huge burden on the interpretation and calculation. Even if all the assumptions are taken for granted or validly checked, the traditional approaches reveal more on the correlation or association properties and less on the causality, because of the fundamental reasoning is based on distance functions or the least squared methods, which are symmetric indicators. We devise a method which directly measures the causality between vectors, which in turn measures the causal relation between agriculture-related variables. The measure takes monotonicity, temporal properties, asymmetry and additivity into consideration. It is then implemented by a set of simulated data and two sets of agriculture-related data. This method could validate or invalidate the existence of positive or negative causal relations between agriculture-related variables. In the end, we analyze the advantages and disadvantages of this method.