Extreme precipitation in Illinois contributes to impacts across diverse landscapes, posing threats to agriculture in rural areas and infrastructure in urban centers. Previous studies have reported an increase in the frequency of heavy precipitation in the region and projected its amplification under climate change. However, these findings are often characterized by inconsistent and/or inappropriate approaches for estimating historical trends and their significance and often lack process-based understanding regarding future changes in extreme event climatology. This study aims to obtain robust regional extreme precipitation trends and relate those trends to large-scale circulation and humidity. The climatology and trends of daily extreme precipitation are established by applying a peaks-over-threshold approach to the newly developed NOAA NCEI nClimGrid-D dataset which includes daily precipitation totals at 5-km resolution. For trend estimation, we use Theil-Sen estimation with three approaches designed to emphasize correction of inflation in the significance of the estimated trends: (1) a “naïve” approach in which we simply consider the direct output of the Theil-Sen method and assess significance using a traditional Mann-Kendall test, (2) an approach based on a modified Mann-Kendall test to account for serial autocorrelation in the assessment of significance, and (3) an approach that also controls for the false discovery rate associated with a large number of tests by considering field significance. To relate these trends to large scale drivers, a multivariate self-organizing map is constructed based on standardized 500 mb geopotential height and 850 mb specific humidity obtained from the ECMWF ERA-5 reanalysis dataset. We use a Monte Carlo experiment to identify weather types most associated with extreme precipitation in the area. Temporal and spatial characteristics of the identified weather types are then analyzed to better understanding their role in changes in the frequency of extreme precipitation events across the region. As expected, the results indicate a stark contrast between the naive and more complex approaches for significance testing, where controlling for autocorrelation and test multiplicity reduces the spatial extent of significant trends across all extreme precipitation thresholds. Extreme precipitation in Illinois is found to be associated with a small number of specific weather types characterized by distinct patterns of geopotential height and humidity. Furthermore, the weather types most frequently associated with extreme precipitation are increasing in frequency, suggesting that changes in atmospheric circulation related to moisture transport and convergence are a major contributor to changes in extreme precipitation in Illinois.

The purpose of this study was to evaluate the influence of horizontal moisture fluxes from Atmospheric Rivers (ARs) on extreme precipitation (EP) events in the continental United States (CONUS). Climatological results for both EP, objectively defined using a peaks-over-threshold and block maxima approach, and ARs were processed and analyzed for co-occurrence. EP analyses produced a positive linear trend in magnitude, determined through the block maxima approach, in the Central US and a positive linear trend in frequency, determined by the peaks-over-threshold approach, predominantly for the Northern half of the CONUS. AR results show over 70 AR days throughout the country, and a linear trend of 10 less days per decade in the Central US. Results of the co-occurrence analysis suggest an increasing trend of about one instance of co-occurrence per decade throughout much of the Eastern Coast, Midwest and Pacific Northwest, with a corresponding negative linear trend of about one instance of co-occurrence per decade for much of the Southwest US to Louisiana. Throughout the world, the study of EP, and the careful analysis of its behavior, and possible amplification sources such as ARs, at the national and regional scale is imperative to obtain a comprehensive understanding of hydrometeorological impacts.

京都大学 / 新制・課程博士 / 博士(工学) / 甲第23499号 / 工博第4911号 / 新制||工||1767(附属図書館) / 京都大学大学院工学研究科都市環境工学専攻 / (主査)教授 米田 稔, 教授 清水 芳久, 准教授 島田 洋子 / 学位規則第4条第1項該当 / Doctor of Philosophy (Engineering) / Kyoto University / DGAM

Convective snow bands develop commonly over the open water surface of lakes or seas when cold airgets advected from a continent. Enhanced heat and moisture fluxes from the comparatively warm waterbody trigger shallow convection and an unstable boundary layer builds up. Relatively strong wind canorganize this convection into wind-parallel quasi-stationary cloud bands with moving individual cells.Depending on various factors like the horizontal wind, the vertical shear or the shape of the coast, thosecloud bands can form of different strength and structure. When the air mass meets the coast orographicforcing causes horizontal convergence and vertical lifting intensifies the precipitation at the coast. If thewind direction stays constant for several days a single snow band would accumulate its precipitation ina very restricted region and cause locally a significant increase in snow depth. This process leads in thecold season repeatedly to severe precipitation events at the Swedish east coast. Large amounts of snowalong with strong wind speeds can cause serious problems for traffic and infrastructure.Two different cases of convective snow bands in the Baltic Sea area were selected to simulate theassociated atmospheric conditions with a total of five different model systems. The atmosphere climatemodel RCA has been used independently at default settings as well as with increased resolution on avertical and a horizontal scale and furthermore coupled either to the ice-ocean model NEMO or the wavemodel component WAM.Comparing all models the crucial parameters like wind, temperature, heat fluxes, and precipitationvary generally in a reasonable range. However, the model systems show systematical differences amongthemselves. The strongest 10 meter wind speeds can be observed for both RCA models with increasedresolution. The RCA-WAM simulation shows its wind enhancement during the snow band event witha time shift to the other models by several hours. The mean directional wind shear above the Gulf ofBothnia, the snow band’s region of origin, is for all models small. The warmest sea surface temperaturesare reached by the RCA-NEMO simulation, which as a result also stands out for its most intense heatfluxes in both sensible and latent heat. Both high resolution RCA models as well as RCA-NEMO givethe most remarkable local precipitation rates. The original RCA and RCA-WAM simulate significantlyless snowfall. Local comparison with SMHI station measurements show that the models represent thetrend of wind, temperature and precipitation evolution well. However, all models decelerate the air masstoo rapidly when meeting the coast. Moreover, it remains a challenge to simulate the exact time andlocation of the extreme precipitation.The coupling of the atmosphere model with the ice-ocean model as well as the increased resolution ofthe atmospheric component have been observed to show great improvements in the model performanceand are suggested for future research work to be used in combination with each other for the regionalmodelling of convective snow bands in the Baltic Sea area.

Classically, we look at annual maximum precipitation series from the perspective of extreme value statistics, which provides a useful statistical distribution, but does not allow much flexibility in the context of climate change. Such distributions are usually assumed to be static, or else require some assumed information about possible trends within the data. For this study, we treat the maximum rainfall series as sums of underlying signals, upon which we perform a decomposition technique, Empirical Mode Decomposition. This not only allows the study of non-linear trends in the data, but could give us some idea of the periodic forces that have an effect on our series. To this end, data was taken from stations in the New England area, from different climatological regions, with the hopes of seeing temporal and spacial effects of climate change. Although results vary among the chosen stations the results show some weak signals and in many cases a trend-like residual function is determined.

The current thesis is structured in four parts. Vector smoothing methods are used to study environmental data, in particular records of extreme precipitation, the models utilized belong to the vector generalized additive class. In the statistical analysis of observational studies the identification and adjustment for prognostic factors is an important component of the analysis; employing flexible statistical methods to identify and characterize the effect of potential prognostic factors in a clinical trial, namely "generalized additive models", presents an alternative to the traditional linear statistical model. The classes of models for which the methodology gives generalized additive extensions include grouped survival data from the Surveillance, Epidemiology, and End Results tumors of the brain and the central nervous system database; we are employing piecewise linear functions of the covariates to characterize the survival experienced by the population. Finally, both descriptive and analytical methods are utilized to study incidence rates and tumor sizes associated with the disease.

Atmospheric rivers (ARs) are important contributors to cool season precipitation in the Southwestern US, and in some cases can lead to extreme hydrometeorological events in the region. We performed a climatological analysis and identified two predominant types of ARs that affect the central mountainous region in Arizona: Type 1 ARs originate in the tropics near Hawaii (central Pacific) and enhance their moisture in the midlatitudes, with maximum moisture transport over the ocean at low-levels of the troposphere. On the other hand, moisture in Type 2 ARs has a more direct tropical origin and meridional orientation with maximum moisture transfer at mid-levels. We then analyze future projections of Southwest ARs in a suite of global and regional climate models used in the North American Regional Climate Change Assessment Program (NARCCAP), to evaluate projected future changes in the frequency and intensity of ARs under warmer global climate conditions. We find a consistent and clear intensification of the water vapor transport associated with the ARs that impinge upon Arizona and adjacent regions, however, the response of AR-related precipitation intensity to increased moisture flux and column-integrated water vapor is weak and no robust variations are projected either by the global or the regional NARCCAP models. To evaluate the effect of horizontal resolution and improve our physical understanding of these results, we numerically simulated a historical AR event using the Weather Research and Forecasting (WRF) model at a 3-km resolution. We then performed a pseudo-global warming experiment by modifying the lateral and lower boundary conditions to reflect possible changes in future ARs (as projected by the ensemble of global model simulations used for NARCCAP). Interestingly we find that despite higher specific humidity, some regions still receive less rainfall in the warming climate experiments - partially due to changes in thermodynamics, but primarily due to AR dynamics. Therefore, we conclude from this analysis that overall future increase in atmospheric temperature and water content as projected by global climate models will not necessarily translate into generalized heavier AR-related precipitation in the Southwestern US.

京都大学 / 新制・課程博士 / 博士(工学) / 甲第23164号 / 工博第4808号 / 新制||工||1752(附属図書館) / 京都大学大学院工学研究科社会基盤工学専攻 / (主査)教授 立川 康人, 准教授 KIM SUNMIN, 教授 中北 英一 / 学位規則第4条第1項該当 / Doctor of Philosophy (Engineering) / Kyoto University / DFAM

No description available.

Les précipitations extrêmes en France sont responsables de phénomènes d'inondations entraînant la perte de vies humaines et des millions d'euros en dégâts matériels. Mesurer le risque associé à ces événements météorologiques rares fait appel à la théorie statistique des valeurs extrêmes, qui propose plusieurs approches permettant d'évaluer des scénarios catastrophes. Cette thèse s'intéresse en particulier à trois mesures de risque faisant intervenir à la fois des lois de probabilité jointes et des méthodes de prédiction spatiale liées à la géostatistique.Dans un premier temps, plusieurs modèles spatiaux de valeurs extrêmes construits sur des données de maxima annuels sont évalués dans une étude comparative sous la forme d'un article. La comparaison des méthodes est menée en se servant de simulations construites à partir de données réelles de maxima annuels de précipitations en France et porte sur des critères liés aux deux mesures de risque que sont le niveau de retour centennal et le coefficient extrémal.Un modèle en particulier, le processus max-stable et hiérarchique de Reich et Shaby (2012) est étudié en détail et fait l'objet d'une implémentation sous la forme d'un package R dédié à la simulation et à l'estimation par cette méthode.Dans un second temps, les données journalières dépassant un seuil élevé sont modélisées dans un cadre spatial dans le but d'estimer une probabilité d'échec conditionnelle. Plusieurs estimateurs de cette mesure sont proposés en se concentrant d'une part sur des méthodes paramétriques liées aux processus Pareto et d'autre part sur deux approches non paramétriques. Les méthodes sont construites de sorte que la dépendance temporelle observable dans les valeurs journalière soit prise en compte lors de l'estimation.Tout au long de la thèse, les méthodes développées sont appliquées sur des données journalières de précipitations en France / Extreme precipitation in France are responsible for flooding events that cause people's deaths and billions of euros in material damage. Measuring the risk associated to these rare meteorological events is possible thanks to the extreme value theory which allows the estimation of such catastrophic scenarios. This thesis focus on three risk measures involving joint probabilities and spatial prediction methods related to geostatistics.In a first time, several spatial models for extreme values built on annual maxima are evaluated in a comparative study in the form of an article. This comparison is performed using simulated data from real annual maxima of precipitation in France. It is also based on two criteria linked to risk measures: the hundred years return level and the extremal coefficient. One particular model is presented in details: the one of Reich and Shaby (2012). This model is implemented under a R package entirely dedicated to its estimation and simulation procedures.In a second time, exceedances of spatial daily data are modelled in order to estimate a conditional failure probability. Several estimators of this measure are proposed, based on the one hand on parametric methods involving Pareto processes and on the other hand on non parametric approaches. The temporal dependence in extremes is also considered with care when estimating this probability.Along this thesis, the methods are applied on daily data of precipitation in France