Environmetrics最新文献

This paper provides a new modeling for air pollution, simultaneously taking into account the six main pollutants (PM10 and PM2.5, Sulphate Dioxide, Nitrogen Dioxide, Carbon Monoxide, ground level Ozone concentrations) and their key determinants, employing Structural Equation Models (SEMs). The model is able to estimate the complex links among air pollutants, often neglected in literature, and identifies specific drivers of air pollution. In literature, indexes of air pollution achieved using a fully statistical methodology have not been proposed yet. Indeed, an added value of this proposal is the statistical procedure itself, which can be applied also to obtain indexes modeling different phenomena. In particular, in this study, the new Air Pollution Index (API) is based on a modeling approach that allows to assess, through statistical criteria, the goodness of fit of the SEM in modeling pollutants and the significance of their determinants. The performance of the new index is assessed using air quality data for municipal European areas, which are characterized by different socioeconomic, geographical, and meteorological features. SEMs are estimated and evaluated in terms of best fit and model complexity. The index resulting by the best SEM is compared with the well-established Air Quality Index (AQI). The new API is validated by means of a sensitivity analysis, performed with a simulation study. Finally, to visualize the meaningfulness of the obtained results, a model-based cluster analysis is estimated on the municipal areas. The proposed SEM contributes to a better understanding of the relationships between air pollutants and their determinants, and this knowledge can inform policy decisions aimed at reducing air pollution and improving public health.

Geolocation error in spaceborne sampling light detection and ranging (LiDAR) measurements of forest structure can compromise forest attribute estimates and degrade integration with georeferenced field measurements or other remotely sensed data. Data integration is especially problematic when geolocation error is not well quantified. We propose a general model that uses airborne laser scanning data to quantify and correct geolocation error in spaceborne sampling LiDAR. To illustrate the model, LiDAR data from NASA Goddard's LiDAR Hyperspectral and Thermal Imager (G-LiHT) was used with a subset of LiDAR data from NASA's Global Ecosystem Dynamics Investigation (GEDI). The model accommodates multiple canopy height metrics derived from a simulated GEDI footprint kernel using spatially coincident G-LiHT, and incorporates both additive and multiplicative mapping between the canopy height metrics generated from both datasets. A Bayesian implementation provides probabilistic uncertainty quantification in both parameter and geolocation error estimates. Results show a systematic geolocation error of 9.62 m in the southwest direction. In addition, estimated geolocation errors within GEDI footprints were highly variable, with results showing a $��\sim �$0.45 probability the true footprint center is within 20 m. Estimating and correcting geolocation error via the model outlined here can help inform subsequent efforts to integrate spaceborne LiDAR data, like GEDI, with other georeferenced data.

We propose a non-stationary spatial model based on a normal-inverse-Wishart framework, conditioning on a set of nearest-neighbors. The model, called nearest-neighbor Gaussian process with random covariance matrices is developed for both univariate and multivariate spatial settings and allows for fully flexible covariance structures that impose no stationarity or isotropic restrictions. In addition, the model can handle duplicate observations and missing data. We consider an approach based on integrating out the spatial random effects that allows fast inference for the model parameters. We also consider a full hierarchical approach that leverages the sparse structures induced by the model to perform fast Monte Carlo computations. Strong computational efficiency is achieved by leveraging the adaptive localized structure of the model that allows for a high level of parallelization. We illustrate the performance of the model with univariate and bivariate simulations, as well as with observations from two stationary satellites consisting of albedo measurements.

The modeling and identification of time series data with a long memory are important in various fields. The streamflow discharge is one such example that can be reasonably described as an aggregated stochastic process of randomized affine processes where the probability measure, we call it reversion measure, for the randomization is not directly observable. Accurate identification of the reversion measure is critical because of its omnipresence in the aggregated stochastic process. However, the modeling accuracy is commonly limited by the available real-world data. We resolve this issue by proposing the novel upper and lower bounds of a statistic of interest subject to ambiguity of the reversion measure. Here, we use the Tsallis value-at-risk (TsVaR) as a convex risk functional to generalize the widely used entropic value-at-risk (EVaR) as a sharp statistical indicator. We demonstrate that the EVaR cannot be used for evaluating key statistics, such as mean and variance, of the streamflow discharge due to the blowup of some exponential integrand. We theoretically show that the TsVaR can avoid this issue because it requires only the existence of some polynomial moment, not exponential moment. As a demonstration, we apply the semi-implicit gradient descent method to calculate the TsVaR and corresponding Radon–Nikodym derivative for time series data of actual streamflow discharges in mountainous river environments.

A new generalization of the extreme value distribution is presented with its density function, having a wide variety of density and tail shapes for modeling extreme value data. This generalized extreme value distribution will be referred to as the odd generalized extreme value distribution. It is derived by considering the distributions of the odds of the generalized extreme value distribution. Consequently, the new distribution is enlightened by not only having all six families of extreme value distributions; Gumbel, Fréchet, Weibull, reverse-Gumbel, reverse-Fréchet, and reverse-Weibull as submodels but also convenient for modeling bimodal extreme value data that are frequently found in environmental sciences. Basic properties of the distribution, including tail behavior and tail heaviness, are studied. Also, quantile-based aliases of the new distribution are illustrated using Galton's skewness and Moor's kurtosis plane. The adequacy of the new distribution is illustrated using well-known goodness-of-fit measures. A simulation is performed to validate the estimated risk measures due to repeated data points frequently found in temperature data. The Grand Rapids and well-known Wooster temperature data sets are analyzed and compared to nine different extreme value distributions to illustrate the new distribution's bimodality, flexibility, and overall fitness.

Regression-based “fingerprinting” methods in climate science employ total least squares (TLS) or orthogonal regression to remedy attenuation bias arising from measurement error due to reliance on climate model-generated explanatory variables. Proving the consistency of multivariate TLS requires assuming noise variances are equal across all variables in the model. This assumption has been challenged empirically in the climate context but little is known about TLS biases when the assumption is violated. Monte Carlo analysis is used herein to examine coefficient biases when the noise variances are not equal. The analysis allows the explanatory variables to be negatively correlated which is typical in climate applications. Ordinary least squares (OLS) exhibits the expected attenuation bias which vanishes as the noise variances on the explanatory variables disappear. In some cases, TLS corrects attenuation bias but more typically imparts large and generally positive biases. OLS performs well when the true value of $��\beta �=�0�$ whereas TLS performs quite poorly. This implies that TLS is not well suited for tests of the null. When $��\beta �=�1�$ TLS tends to exhibit opposite biases to OLS. Diagnostic information specific to each data sample should be consulted before using TLS to avoid spurious inferences and replacing OLS attenuation bias with other, potentially larger biases.

We develop two models for the temporal evolution of extreme events of multivariate $��k�$th order Markov processes. The foundation of our methodology lies in the conditional extremes model of Heffernan and Tawn (Journal of the Royal Statistical Society: Series B (Methodology), 2014, 66, 497–546), and it naturally extends the work of Winter and Tawn (Journal of the Royal Statistical Society: Series C (Applied Statistics), 2016, 65, 345–365; Extremes, 2017, 20, 393–415) and Tendijck et al. (Environmetrics 2019, 30, e2541) to include multivariate random variables. We use cross-validation-type techniques to develop a model order selection procedure, and we test our models on two-dimensional meteorological-oceanographic data with directional covariates for a location in the northern North Sea. We conclude that the newly-developed models perform better than the widely used historical matching methodology for these data.

Among the most relevant processes in the Earth system for human habitability are quasi-periodic, ocean-driven multi-year events whose dynamics are currently incompletely characterized by physical models, and hence poorly predictable. This work aims at showing how (1) data-driven, stochastic machine learning approaches provide an affordable yet flexible means to forecast these processes; (2) the associated uncertainty can be properly calibrated with fast ensemble-based approaches. While the methodology introduced and discussed in this work pertains to synoptic scale events, the principle of augmenting incomplete or highly sensitive physical systems with data-driven models to improve predictability is far more general and can be extended to environmental problems of any scale in time or space.

Designs that produces spatially balanced, or well-spread, samples are desirable as they increase the probability of obtaining a sample highly representative of the population. Spatially correlated Poisson sampling (SCPS) is a method for selecting well-spread samples. In the SCPS method, the sampling outcomes (inclusion or exclusion of units) are decided sequentially. After each decision, the inclusion probabilities of surrounding units are updated. A specific order for deciding the sampling outcomes is not enforced for SCPS, that is, the order can be chosen randomly or be fixed. A new modified method called locally correlated Poisson sampling (LCPS) is suggested. In this new method, the order of the decisions makes sure the inclusion probabilities are updated (more) locally. As a result, a stronger negative correlation between inclusion indicators of nearby units is achieved. Simulations on various data sets show that the resulting samples from LCPS, in general, are more spatially balanced and produce lower variance than samples from SCPS and the local pivotal method.