Search

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. [ABSTRACT FROM AUTHOR]

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. [ABSTRACT FROM AUTHOR]

A comparison of moderate to extreme daily precipitation from the ERA‐5 reanalysis by the European Centre for Medium‐Range Weather Forecasts against two observational gridded data sets, EOBS and CMORPH, is presented. We assess the co‐occurrence of precipitation days and compare the full precipitation distributions. The co‐occurrence is quantified by the hit rate. An extended generalized Pareto distribution (EGPD) is fitted to the positive precipitation distribution at every grid point and confidence intervals of quantiles compared. The Kullback–Leibler divergence is used to quantify the distance between the entire EGPDs obtained from ERA‐5 and the observations. For days exceeding the local 90th percentile, the mean hit rate is 65% between ERA‐5 and EOBS (over Europe) and 60% between ERA‐5 and CMORPH (globally). Generally, we find a decrease of the co‐occurrence with increasing precipitation intensity. The agreement between ERA‐5 and EOBS is weaker over the southern Mediterranean region and Iceland compared to the rest of Europe. Differences between ERA‐5 and CMORPH are smallest over the oceans. Differences are largest over NW America, Central Asia, and land areas between 15°S and 15°N. The confidence intervals on quantiles are overlapping between ERA‐5 and the observational data sets for more than 80% of the grid points on average. The intensity comparisons indicate an excellent agreement between ERA‐5 and EOBS over Germany, Ireland, Sweden, and Finland, and a disagreement over areas where EOBS uses sparse input stations. ERA‐5 and CMORPH precipitation intensity agree well over the midlatitudes and disagree over the tropics. Key Points: The timing and the intensity of daily precipitation are assessed over Europe and globallyExtended generalized Pareto distributions are fitted to precipitation from ERA‐5 and station and satellite dataAgreement between data sets is highest in the midlatitudes [ABSTRACT FROM AUTHOR]

Low‐flow events can cause significant impacts to river ecosystems and water‐use sectors; as such, it is important to understand their variability and drivers. In this study, we characterise the variability and timing of annual total frequency of low‐streamflow days across a range of headwater streams within the continental United States. To quantify this, we use a metric that counts the annual number of low‐flow days below a given threshold, defined as the cumulative dry days occurrence (CDO). First, we identify three large clusters of stream gauge locations using a Partitioning Around Medoids (PAM) clustering algorithm. In terms of timing, results reveal that for most clusters, the majority of low‐streamflow days occur from the middle of summer until early fall, although several locations in Central and Western United States also experience low‐flow days in cold seasons. Further, we aim to identify the regional climate and larger scale drivers for these low‐streamflow days. Regionally, we find that precipitation deficits largely associate with low‐streamflow days in the Western United States, whereas within the Central and Eastern U.S. clusters, high temperature indicators are also linked to low‐streamflow days. In terms of larger scale, we examine sea surface temperature (SST) anomalies, finding that extreme dry years exhibit a high degree of co‐occurrence with different patterns of warmer SST anomalies across the Pacific and Northern Atlantic Oceans. The linkages identified with regional climate and SSTs offer promise towards regional prediction of changing conditions of low‐streamflow events. [ABSTRACT FROM AUTHOR]

The analysis of multiple extreme values aims to describe the stochastic behaviour of observations in the joint upper tail of a distribution function. For instance, being able to simulate multivariate extreme events is convenient for end users who need a large number of random replications of extremes as input of a given complex system to test its sensitivity. The simulation of multivariate extremes is often based on the assumption that the dependence structure, the so-called extremal dependence function, is described by a specific parametric model. We propose a simulation method for sampling bivariate extremes, under the assumption that the extremal dependence function is semi-parametric. This yields a flexible tool that can be broadly applied in real-data analyses. With the aim of estimating the probability of belonging to some extreme sets, our methodology is examined via simulation and illustrated by an analysis of strong wind gusts in France. Copyright © 2017 John Wiley & Sons, Ltd. [ABSTRACT FROM AUTHOR]

In statistics, extreme events are often defined as excesses above a given large threshold. This definition allows hydrologists and flood planners to apply Extreme-Value Theory (EVT) to their time series of interest. Even in the stationary univariate context, this approach has at least two main drawbacks. First, working with excesses implies that a lot of observations (those below the chosen threshold) are completely disregarded. The range of precipitation is artificially shopped down into two pieces, namely large intensities and the rest, which necessarily imposes different statistical models for each piece. Second, this strategy raises a nontrivial and very practical difficultly: how to choose the optimal threshold which correctly discriminates between low and heavy rainfall intensities. To address these issues, we propose a statistical model in which EVT results apply not only to heavy, but also to low precipitation amounts (zeros excluded). Our model is in compliance with EVT on both ends of the spectrum and allows a smooth transition between the two tails, while keeping a low number of parameters. In terms of inference, we have implemented and tested two classical methods of estimation: likelihood maximization and probability weighed moments. Last but not least, there is no need to choose a threshold to define low and high excesses. The performance and flexibility of this approach are illustrated on simulated and hourly precipitation recorded in Lyon, France. [ABSTRACT FROM AUTHOR]

ABSTRACT We describe a fully automatized homogenization procedure and illustrate it on Argentinean weather station temperature series. The procedure relies on multiple pairwise comparisons between a candidate station and its surrounding stations. The main advantage of this approach is to get around the difficulty of defining a reliable reference series; its main drawback is to often require visual attribution and grouping of shifts resulting in too high a cost in human time for implementation on large datasets. Here, we fully automatize these two steps by using a probabilistic metric of similarity between shifts which is leveraged within two optimized clustering schemes. Simulation results show performance improvements versus both visual inspection and the automatized procedure of Menne MJ, Williams CN, Jr. 2009. Homogenization of temperature series via pairwise comparisons. J. Clim. 22: 1700-1717. Implementation on Argentinean temperature series results in the identification and removal of numerous inhomogeneities; corrected series reveal stronger and spatially smoother warming trends. [ABSTRACT FROM AUTHOR]

In atmospheric chemistry, daily maxima concentrations capture information about the variability among peak values. Statistically, they can often be modeled by a Gumbel distribution. This is the case for two very important greenhouse gases methane and nitrous oxide maxima when they are measured at our site of interest, Gif-sur-Yvette, a city south west of Paris. In practice, those two daily concentrations are not always recorded during the same period, and it would be of interest to reconstruct one from the other one. Such a type of inference can be handled within a state space modeling framework, but state space models are not tailored to represent the dynamics among Gumbel-distributed maxima. By building on our previous work, which made a link between linear autoregressive time series and Gumbel-distributed maxima, we propose and study such a state space model. It has the advantages of being linear and of preserving the Gumbel characteristic in both the state and observational equations. Concerning the inference of the hidden maxima at the state equation level, we derive the optimal weights of the auxiliary particle filtering approach of Pitt and Shephard. A simulation study indicates that our approach offers a gain over the Kalman filter, the bootstrap filter, and the nonmodified version of the Pitt and Shephard auxiliary filter. Copyright © 2012 John Wiley & Sons, Ltd. [ABSTRACT FROM AUTHOR]

The article presents a study on autoregressive (AR) models for maxima and their applications to methane (CH4) and nitrous oxide ( N2O). The authors impose that AR models have to follow a Gumbel distribution whose main theoretical properties are derived. The study also offers a simple way to model short-term dependencies among maxima stemming.

Currently, there is a tremendous scientific research effort in the area of climate change. In this paper, our motivation is to improve the understanding of historical climatic events such as the Little Ice Age (LIA), a period of relatively cold weather around 1450–1850 AD. Although the LIA is well documented in Europe, its extent and timing are not known in areas of the globe where climatological records were not kept during this period. To study the climate, which predates historical records, proxy climate records must be used. A proxy record for the timing of climatic cooling events are the ages of the moraines left behind by glacial advances. Unfortunately, to determine the ages of these moraines in alpine environments there is little material available but lichens. Hence, lichenometry was developed to determine the ages of glacial landforms by using lichen measurements. To our knowledge, this article provides the first attempt at deriving a comprehensive statistical model for lichenometry. Our model foundation is based on extreme value theory because only the largest lichens are measured in lichenometry studies. This application is novel to extreme value theory because the quantities of interest (the ages of climatic events) are not the measured quantities (lichen diameters), i.e., it is a inverse problem. We model the lichen measurements with the generalized extreme value (GEV) distribution, upon which a Bayesian hierarchical model is built. The hierarchical model enables estimation of the hidden covariate ages of the moraines. The model also allows for pooling of data from different locations and evaluation of spatial differences in lichen growth. Parameter inference is obtained using a straightforward Markov Chain Monte Carlo method. Our procedure is applied to data gathered from the Cordillera Real region in Bolivia. [ABSTRACT FROM AUTHOR]