HS7.8
Orals: Fri, 28 Apr | Room 2.44
Rare to very rare floods (associated to return periods of 1'000–100'000 years) can cause extensive human and economic damage. Still, their estimation is limited by the comparatively short streamflow records available. Some of the limitations of commonly used estimation methods can be avoided by using continuous simulation (CS), which considers many simulated meteorological configurations and a conceptual representation of hydrological processes. CS also avoids assumptions about antecedent conditions and their spatial patterns.
We present an implementation of CS to estimate rare and very rare floods at multiple sites in a large river basin (19 locations in the Aare River basin, Switzerland; area: 17'700 km²), using exceedingly long simulations from a hydrometeorological model chain (Viviroli et al., 2022). The model chain consisted of three components: First, the multi-site stochastic weather generator GWEX provided 30 meteorological scenarios (precipitation and temperature) spanning 10'000 years each. Second, these weather generator simulations were used as input for the bucket-type hydrological model HBV, run at an hourly time step for 80 catchments covering the entire Aare River basin. Third, runoff simulations from the individual catchments were routed for a representation of the entire Aare River system using the routing system model RS Minerve, including a simplified representation of main river channels, major lakes and relevant floodplains. The final simulation outputs spanned about 300'000 years at hourly resolution and cover the Aare River outlet, critical points further upstream as well as the outlets of the hydrological catchments. The comprehensive evaluation over different temporal and spatial scales showed that the main features of the meteorological and hydrological observations were well represented. This implied that meaningful information on floods with low probability can be inferred. Although uncertainties were still considerable, the explicit consideration of important flood generating processes (snow accumulation, snowmelt, soil moisture storage) and routing (bank overflow, lake regulation, lake and floodplain retention) was a substantial advantage compared to common extrapolation of streamflow records.
The suggested approach allows for comprehensively exploring possible but unobserved spatial and temporal patterns of hydrometeorological behaviour. This is particularly valuable in a large river basin where the complex interaction of flows from individual tributaries and lake regulations are typically not well represented in the streamflow records. The framework is also suitable for estimating more common, i.e., more frequently occurring floods.
Reference
Viviroli D, Sikorska-Senoner AE, Evin G, Staudinger M, Kauzlaric M, Chardon J, Favre AC, Hingray B, Nicolet G, Raynaud D, Seibert J, Weingartner R, Whealton C, 2022. Comprehensive space-time hydrometeorological simulations for estimating very rare floods at multiple sites in a large river basin. Natural Hazards and Earth System Sciences, 22(9), 2891–2920, doi:10.5194/nhess-22-2891-2022
How to cite: Viviroli, D., Sikorska-Senoner, A. E., Evin, G., Staudinger, M., Kauzlaric, M., Chardon, J., Favre, A.-C., Hingray, B., Nicolet, G., Raynaud, D., Seibert, J., Weingartner, R., and Whealton, C.: Estimating very rare floods at multiple sites in a large river basin with comprehensive hydrometeorological simulations, EGU General Assembly 2023, Vienna, Austria, 24–28 Apr 2023, EGU23-1384, https://doi.org/10.5194/egusphere-egu23-1384, 2023.
Spatially co-occurring floods pose a great threat to the resilience and the recovery potential of the communities. A timely forecasting of such events plays a crucial role for increasing the preparedness of public and private sectors and for limiting the associated losses. In this study we investigated the potential of dilated Convolutional Neural Networks (CNN) conditioned on a set of large-scale climatic indexes and antecedent precipitation for monthly forecast of widespread flooding severity in Germany using 63 years of streamflow observations. The severity of widespread flooding (i.e., spatially co-occurring floods) was estimated as simultaneous (within a given month) exceedance of an at-site two-year return period for streamflow peaks across 172 mesoscale catchments. The model was trained for the whole country and for the three diverse hydroclimatic regions individually to provide insights on spatial heterogeneity of model performance and drivers of flooding. Evaluation of the model skill for floods generated by different processes revealed the largest bias for events generated during dry conditions. The bias for rain-on-snow flood events was the lowest despite their higher severity indicating higher predictability of these events from large scale climatic indexes. Model-based feature attribution and independent wavelet coherence analyses both indicated considerable difference in the major drivers of widespread flooding in different regions. While the flooding in the North-Eastern region is strongly affected by the Baltic Sea (e.g., East Atlantic pattern), the North-Western region is affected more by global patterns associated with the El-Niño activity (e.g., Pacific North American pattern). In the Southern region in addition to the effect of the global patterns we also detect the effect of the Mediterranean Sea (Mediterranean Oscillation Index), while antecedent precipitation seems to play less important role in this region compared to the rest of the country. Our results indicate a considerable potential for forecasting widespread flood severity using dilated CNN especially as the length of the available time series for training increases.
How to cite: Tarasova, L., Ahrens, B., Hoff, A., and Lall, U.: Forecasting the monthly severity of widespread flooding in Germany using dilated convolutional neural networks conditioned by large-scale climatic indexes, EGU General Assembly 2023, Vienna, Austria, 24–28 Apr 2023, EGU23-4419, https://doi.org/10.5194/egusphere-egu23-4419, 2023.
Statistical dependency measures such as Kendall’s Tau or Spearman’s Rho are frequently used to analyse the coherence between time series in environmental data analyses. Autocorrelation of the data can however result in spurious cross correlations if not accounted for. Here, we present the asymptotic distribution of the estimators of Spearman’s Rho and Kendall’s Tau, which can be used for statistical hypothesis testing of cross-correlations between autocorrelated observations. The results are derived using U-statistics under the assumption of absolutely regular (or β-mixing) processes. These comprise many short-range dependent processes, such as ARMA-, GARCH- and some copula-based models relevant in the environmental sciences. We show that while the assumption of absolute regularity is required, the specific type of model does not have to be specified for the hypothesis test. Simulations show the improved performance of the modified hypothesis test for some common stochastic models and small to moderate sample sizes under autocorrelation. The methodology is applied to observed time series of flood discharges and temperatures in Europe and yields results that are consistent with the literature on flood regime changes in Europe. While the standard test results in spurious correlations between floods and temperatures, this is not the case for the proposed test, which is more consistent with the literature on flood regime changes in Europe.
How to cite: Lun, D., Fischer, S., Viglione, A., and Blöschl, G.: Attribution of flood changes with time series in the presence of autocorrelation: Modifications for Spearman‘s Rho and Kendall‘s Tau, EGU General Assembly 2023, Vienna, Austria, 24–28 Apr 2023, EGU23-7352, https://doi.org/10.5194/egusphere-egu23-7352, 2023.
Coastal flooding is among the world’s deadliest and costliest natural hazards. The impacts caused by coastal flooding can be particularly high when an event affects a large spatial area, as witnessed during Hurricane Katrina and Cyclone Xaver. Current large-scale flood risk studies assume that the probabilities of water levels during such events do not vary in space. This failure to capture flood spatial dependence can lead to large misestimates of the hazard and risk at large spatial scales, and therefore potentially misinform the risk management community. In this contribution, we assess the effects of spatial dependence on coastal flood risk estimation at the global scale. To this end, we compare the assessments using two spatial dependence scenarios: i) complete dependence and ii) modelled dependence of water level return periods. For the complete dependence scenario, we use the existing risk information calculated by the GLOFRIS global risk modelling framework. To estimate the spatially-dependent risks, we use an event-based multivariate statistical approach and consider 10,000-year extreme coastal flood events derived from the global synthetic dataset of spatially-dependent extreme sea levels. The associated spatially coherent return periods of each event are then combined with the GLOFRIS spatially-constant inundation layers to create the spatially-dependent inundation map. These hazard maps, overlaid with exposure layers and vulnerability information, are further used to assess the coastal flood impacts. The flood risk is estimated using Weibull’s plotting formula and presented in terms of expected annual population and expected annual damage. This study will improve our understanding of flood spatial dependence and will provide improved risk estimation at the global scale. Such reliable estimates could lead to improved large-scale flood risk management through better wide-area planning decisions, more accurate insurance coverage, and better emergency response.
How to cite: Li, H., Haer, T., Enríquez, A., and Ward, P.: The role of spatial dependence in global-scale coastal flood risk assessment, EGU General Assembly 2023, Vienna, Austria, 24–28 Apr 2023, EGU23-230, https://doi.org/10.5194/egusphere-egu23-230, 2023.
In locations where measured timeseries are not available or not sufficiently long, reliable predictions of the rainfall depth associated with a given duration and exceedance probability may be obtained through regional frequency analysis (RFA). The scientific literature reports on a large number of different approaches to RFA of rainfall extremes, each one characterized by specific advantages and disadvantages. One of the most common drawbacks is that regional models specifically refer to a single duration or a single exceedance probability. Second, several approaches require the definition of a homogeneous region where the model is trained; this leads to higher accuracy, but also the applicability of the model is limited to those locations that are hydrologically similar to the homogeneous group used in the training. Moreover, most models require filtering the available gauged stations based on the length of the measured timeseries to perform reliable frequency analysis. These aspects lead to discard a significant amount of data, which could turn out to be detrimental to the accuracy of the regional prediction in some cases.
We set up a few alternative models aiming to investigate and discuss a different and innovative approach for RFA of rainfall extremes. We want to address three main research questions: (1) Can a single model represent the frequency of extreme rainfall events over a large, climatically, and morphologically complex geographical area? (2) Can a single RFA model handle all sub-daily durations (i.e., from 1 to 24h)? (3) Is it possible to exploit all available annual maximum series, regardless of their length (i.e., very short ones too)? We select a large study area that is located in north-central Italy. We make use of more than 2300 Annual Maximum Series of rainfall depth for different time-aggregation intervals between 1 and 24 hours, that have been collected between 1928 and 2011 in the Italian Rainfall Extreme Dataset (I2-RED). For each gauged station, several morpho-climatic descriptors are retrieved (e.g., minimum distance to the coast, elevation of orographic barriers, aspect, terrain slope, etc.) and used as covariates for the prediction models. Our models are based on ensembles of unsupervised artificial neural networks (ANNs) and are able to predict parameters of a Gumbel distribution for any location and any duration in the 1-24 hours range based on the morphoclimatic descriptors. Through the analysis of results over 100 gauged validation stations, a profitable discussion is enabled on the potential and drawbacks of ensembles of unsupervised ANNs for regional frequency analysis of sub-daily rainfall extremes.
How to cite: Magnini, A., Lombardi, M., Ouarda, T. B. M. J., and Castellarin, A.: Improved data assimilation in regional frequency analysis of rainfall extremes across large and morphologically complex geographical areas , EGU General Assembly 2023, Vienna, Austria, 24–28 Apr 2023, EGU23-10538, https://doi.org/10.5194/egusphere-egu23-10538, 2023.
Snow depth is a significant component in the hydrological cycle and global energy and water balances, contributing to climate change impacts. Weather stations with gauges for snow depth are scarce, especially in complex terrain regions, and require high accuracy for measurements. Advances in observational systems offer unconventional solutions yet are expensive. To bridge these gaps, stochastic generation methods offer a cost-effective solution to reproduce time series of hydrological variables, preserving their stochastic properties. Stochastic generation methods are well-established for total precipitation but lack snow depth generation. Here, we introduce a stochastic method to exclusively generate snow depth time series that preserve their distinct statistical properties on different time scales. We use 450 observed snow depth time series and 470 CMIP6 simulations to detect Canada's observed and physical statistical properties. The results indicate that snow depth has a light tail, and the distribution might change daily. The probability of zero snow depth shows a clear seasonal pattern. The synthetic snow depth time series can be an alternative to climate models’ outputs, offering a computationally effective solution to investigate the snow depth variability. This method advances the generation of stochastic time series of snow depth and can be applied to investigate catastrophes from snowmelt processes and avalanches that lead to severe damage and fatalities.
How to cite: Abdelmoaty, H. and Papalexiou, S.: Stochastic Generation of Snow Depth in Canada, EGU General Assembly 2023, Vienna, Austria, 24–28 Apr 2023, EGU23-2974, https://doi.org/10.5194/egusphere-egu23-2974, 2023.
Extreme wildfires continue to be a significant cause of human death and biodiversity destruction across the globe, with recent worrying trends in their activity (i.e., occurrence and spread) suggesting that wildfires are likely to be highly impacted by climate change. In order to facilitate appropriate risk mitigation for extreme wildfires, it is imperative to identify their main drivers and assess their spatio-temporal trends, with a view to understanding the impacts of global warming on fire activity. To this end, we analyse monthly burnt area due to wildfires using a hybrid statistical deep-learning framework that exploits extreme value theory and quantile regression. Three study regions are considered: the contiguous U.S., Mediterranean Europe and Australia.
How to cite: Richards, J. and Huser, R.: Insights into the drivers and spatio-temporal trends of extreme wildfires with statistical deep-learning, EGU General Assembly 2023, Vienna, Austria, 24–28 Apr 2023, EGU23-2332, https://doi.org/10.5194/egusphere-egu23-2332, 2023.
Posters on site: Fri, 28 Apr, 08:30–10:15 | Hall A
The recurrent and destructive nature of floods causes enormous economic damage and loss of human lives, leaving people in flood-prone areas fearful and insecure. It is essential to have a thorough knowledge of the factors that contribute to it. However, most catchment characterization studies are limited to case studies or regional domains. A detailed global characterization is currently unavailable due to the limitation in the holistic dataset that it demands. This study aims to fill this gap by utilizing multiple global datasets describing physiographic explanatory variables to characterize streamflow extremes. The role of catchment features such as landcover, geomorphology, climatology, lithology, etc., on spatial patterns and temporal changes of high streamflow extremes, was investigated in detail. Moreover, the multidimensional correlations between streamflow extremes and catchment features were modeled using a Random Forest approach and integrated with an interpretable machine learning framework to find the most dominating elements in different climate classes. The interpretation reveals that climatological variables are the most influential across all climates. However, the variables and their influences fluctuate between climates. Furthermore, distinct geomorphological variables dominate throughout climatic classes (such as basin relief in warm temperate and drainage texture in arid climates). Overall, the insights of this study would play a vital role in estimating the unit peak discharge at ungauged stations based on known watershed features. In addition, these findings can also help assess the nature of extremes in future climate scenarios, consequently implicating risk management methods.
How to cite: Kuntla, S. K. and Saharia, M.: Embracing Large-sample Data to Characterize Streamflow Extremes at a Global-scale, EGU General Assembly 2023, Vienna, Austria, 24–28 Apr 2023, EGU23-90, https://doi.org/10.5194/egusphere-egu23-90, 2023.
The prevalence of heavy tailed (HT) populations in hydrology is becoming increasingly commonplace due in part to the increasing need and use of high frequency and high-resolution data. In addition to the impact of HT on extremes, HT populations can have a profound impact on a wide range of other hydrologic statistics and methods associated with planning, management and design for extremes. We review the known impacts of HT populations on the instability and bias in a wide range of commonly used hydrologic statistics. Experiments reveal that HT distributions result in the degradation of many commonly used statistical methods including the bootstrap, probability plots, the central limit theorem, and the law of large numbers. We document the gross instability of perhaps the best-behaved statistic of all, the sample mean (SM) when computed from HT distributions. The SM is ubiquitous because it is a component of and related to a myriad of statistical methods, thus its unstable behavior provides a window into future challenges faced by the hydrologic community. We outline many challenges associated with HT data, for example, upper product moments are often infinite for HT populations, yet upper L-moment always exist, so that the theory of L-moments is uniquely suited to HT distributions and data. We introduce a magnification factor for evaluating the impact of HT distributions on the behavior of extreme quantiles
How to cite: Vogel, R., Lamontagne, J., and Dolan, F.: The Prevalence and Impact of Heavy Tails on Hydrologic Extremes and Other Statistics, EGU General Assembly 2023, Vienna, Austria, 24–28 Apr 2023, EGU23-2129, https://doi.org/10.5194/egusphere-egu23-2129, 2023.
Drought is one of the world's major natural disasters. In order to monitor drought and reduce drought damage through preemptive response, it is important to understand the spatiotemporal evolutionary characteristics of drought. Droughts have a three-dimensional (3-D) space-time structure, typically spanning hundreds of kilometers and lasting months to years. In this study, a high-resolution(5 km) SPEI-HR(Standardized Precipitation Evaporation Index) dataset was used, considering climatic (typical temperate continental climate) and various geographic characteristics (mountainous terrain, lowland basin, desert, grassland, etc.). In addition, all large- and small-scale drought events that evolve spatiotemporally were extracted using the dynamic drought detection technique (DDDT) algorithm. These 3D-drought properties are important information to explain the spatiotemporal evolution of drought and are characterized by drought patches in dynamic drought maps. As a result, most of the trajectories of droughts in Central Asia during the period 1981 to 2018 tended to move laterally to the east and west (ENE, E, ESE, WNW, W, WSW). In addition, droughts in Central Asia are characterized by very strong correlations between indicators of duration, severity, area, and trajectory movement distance. These Central Asian drought characteristics are interpreted as meaning that there is consistency among various drought information in determining the most severe drought event. In addition, the dynamic drought map, which includes all 3D-drought properties, has the advantage of producing high-level drought information (temporal continuity of drought events and dynamic evolution characteristics, etc.) that are not found in general drought maps through various conditional drought monitoring.
Acknowledgements: This work was supported by the National Research Foundation of Korea (No. NRF-2020R1C1C1014636) and Korea Environment Industry & Technology Institute (KEITI) (No.2022003610001) funded by the Korean government (MSIT and MOE).
How to cite: Yoo, J., Kim, J., Kwon, H.-H., and Kim, T.-W.: Spatial and Temporal Evolution of Drought Events Using High-Resolution SPEI and Dynamic Drought Detection Algorithm, EGU General Assembly 2023, Vienna, Austria, 24–28 Apr 2023, EGU23-3709, https://doi.org/10.5194/egusphere-egu23-3709, 2023.
Heavy precipitation events are strongly affected by climate change and there is a high confidence that these extremes will become more frequent and more severe in the future. Moreover, potential changes in the seasonality of these events are important in terms of planning and preparation against these events. While efforts have been focused on changes in the magnitude and seasonality of extreme precipitation events, these studies have treated these two quantities separately.
In order to overcome to this limit, a different perspective is here used by modeling the seasonality and magnitude of extreme precipitation events together through circular-linear copulae. We perform analyses at the global scale and develop bivariate models for an historical dataset. The outputs provided from several global climate models from the sixth phase of the Coupled Model Intercomparison Project (CMIP6) and Shared Socioeconomic Pathways (SSPs) from 1-2.6 to 5-8.5 are then used to examine the joint projected changes in the seasonality and magnitude of extreme precipitation at the global scale.
How to cite: Treppiedi, D., Villarini, G., Bender, J., and Noto, L.: On the Projected Changes in the Seasonality and Magnitude of Precipitation Extremes, EGU General Assembly 2023, Vienna, Austria, 24–28 Apr 2023, EGU23-3851, https://doi.org/10.5194/egusphere-egu23-3851, 2023.
Storm surges and accompanying extreme water levels pose a major threat to coastal structures, urban and industrial areas and human life in general. In order to develop effective risk mitigation strategies, it is crucial to improve the understanding of these extreme events as well as their occurrence probabilities and quantiles, respectively.
The standard procedure to estimate extreme quantiles (return-levels) is to fit a suitable distribution to the observed extreme values on a site-by-site basis. However, this approach exhibits some disadvantages: 1) Estimates of extreme quantiles are only available at gauged locations. 2) The small amount of extreme events in tide gauge records makes these estimates highly uncertain.
We tackle both issues by pooling all available tide gauge records together through a covariates model that allows for smoothly varying distribution parameters in space. Using this approach, the model is not only able to reduce the uncertainty in quantile estimates, but also enables the interpolation of the distribution parameters at arbitrary ungauged locations, e.g. in between tide gauge locations.
Deploying our model for the German North Sea coast, we generate a probabilistic reanalysis of extreme water levels as well as associated probabilities for the period 2000 – 2019.
How to cite: Ditzinger, G., Rust, H., Möller, J., Kruschke, T., Schaffer, L., and Hinrichs, C.: A spatial covariates model for storm surge extremes in the German Bight, EGU General Assembly 2023, Vienna, Austria, 24–28 Apr 2023, EGU23-5298, https://doi.org/10.5194/egusphere-egu23-5298, 2023.
Using the German weather radar data for the last 20 years with a high spatial and temporal resolution, the occurrence of rainfall extremes was analysed. By extracting and examining connected rainfall areas, several research questions were investigated: (1) How many extremes occur in a given area independent of their location? (2) To what extent is their occurrence in space a random and to what extent a structured process? (3) How are the connected volumes behaving in space and time? (4) How does the areal extent relate to event duration, rainfall volume, and discharge volume? The first two research questions were investigated for all of Germany, the last two by analysing rainfall and run-off data in several small and medium size headwater catchments in southern and western Germany.
The results show that the occurrence of events in space is related to their areal extent; there are regions where the frequency of occurrence of large spatially distributed events is greater than that of smaller ones. Moreover, there are interesting relationships between the spatial extent of an event, the event duration, and the event rainfall volumes. For high discharge values, not only does the rainfall intensity matter but also the event duration and spatial distribution of rainfall within a catchment. Many discharge peaks are not necessarily caused by high-intensity events (hourly or daily maxima) but by the accumulation of rainfall cells in space and time.
How to cite: El Hachem, A., Seidel, J., and Bárdossy, A.: Areal extremes from a different perspective: rainfall as 2D and 3D connected objects., EGU General Assembly 2023, Vienna, Austria, 24–28 Apr 2023, EGU23-7564, https://doi.org/10.5194/egusphere-egu23-7564, 2023.
Quantifying design rainfall events at varying durations is crucial for assessing flood risk and mitigating losses and damages. Yet, in a changing climate, they are fundamental tool for a reliable design of water related infrastructures, such as flood retention reservoirs, spillways, and urban drainage systems. Usually, design rainfall is quantified where rain gauges are located, and regionalization methods are used to provide estimates in ungauged locations. During the last years, convection-permitting climate models (CPM) are receiving increasing attention because, thanks to their high spatial resolution (~3km) and ability of explicitly resolving atmospheric convection, they allow for better estimating precipitation spatial patterns and extreme rainfall at multiple durations compared to coarser models.
In this work, we combine at-site rain gauge measurements with CPM simulations, within a non-asymptotic statistical framework for the analysis of extreme rainfall. We aim at quantifying the added value of the physics-based information provided by CPM simulations for the estimation of high quantiles of rainfall in ungauged locations.
The performance of the new regionalization approach is compared with traditional interpolation methods (i.e. interpolation of distribution function parameters) using leave- one-out cross-validation as well as considering different rain gauge densities.
Preliminary results show that the proposed methodology based on CPM simulation provides: i) similar performances compared to traditional gauge-based regionalization methods for high station density scenarios and ii) improved performances for low station density scenarios.
How to cite: Formetta, G., Marra, F., Dallan, E., and Borga, M.: Interpolation of design rainfall at ungauged locations exploiting the potential of convection-permitting climate models., EGU General Assembly 2023, Vienna, Austria, 24–28 Apr 2023, EGU23-11332, https://doi.org/10.5194/egusphere-egu23-11332, 2023.
The multivariate return period is a measure of the frequency with which simultaneous sets of variables are expected to occur in a given area. So far, most approaches to calculate the multivariate return period of various hydrological variables have used copulas in two and three dimensions. (Salvadori et al., 2011) proposed a methodology for calculating the return period based on Archimedean copulas and the Kendall measure in 2 and 3 dimensions. (Gräler et al., 2013) proposed the calculation of the trivariate return period based on Vine copulas and Kendall distribution functions to describe the characteristics of the design hydrogram, considering the annual maximum peak discharge, its volume and duration. (Tosunoglu et al., 2020) applied three-dimensional Archimedean, Elliptical and Vine copulas to study the characteristics of floods. These studies have shown that the use of copulas can improve the accuracy of the risk measure of extreme events compared to univariate approaches, that only consider one variable at a time.
One of the limitations in describing the occurrence of multivariate extreme involving more than three simultaneous variables is the complex mathematical model to be solved (highest probability density point of a hypersurface) and the high computational cost that this imposes. However, in some areas of hydrology, developing more robust analyses that consider more than three variables can further improve risk assessments. For example, considering multiple rainfall stations in a watershed may help to properly capture the spatial structure of extremes -instead of relying on other spatial distribution procedures-. This improvement can provide a more accurate measure of the return period in the design of critical infrastructure, flood prediction, risk plans, etc.
In this context, we present an application where an extreme characterization of 5 rain gauges is considered simultaneously, using vine copulas based on Kendall distribution functions. More specifically, we analyze which measures are suitable for explaining the spatial and temporal correlation of rain events in different locations within a network of stations; which families and structures of vine copulas can optimally capture the spatial dependence structure within a region; how to solve the complex mathematics that is imposed when expanding the dimensionality; what is a computationally reasonable alternative to improve the computational cost involved.; and how multivariate analysis can improve the precision of the extreme event risk measure compared to univariate approaches.
These questions are answered by applying the proposed methods to a pilot case, which is developed in a basin located in northern Spain. Multivariate modeling is becoming increasingly relevant in the field of hydrology due to its ability to model extreme stochastic events, which are key to mitigating the risk and damages caused by floods.
References
Gräler, B., Berg, M. J. van den, Vandenberghe, S., Petroselli, A., Grimaldi, S., De Baets, B. & Verhoest, N. E. C., 2013. Multivariate return periods in hydrology: a critical and practical review focusing on synthetic design hydrograph estimation. Hydrol. Earth Syst. Sci., 17(4), 1281–1296.
Salvadori, G., De Michele, C. & Durante, F., 2011. On the return period and design in a multivariate framework. Hydrol. Earth Syst. Sci., 15(11), 3293–3305.
How to cite: Urrea Méndez, D. A., Gómez Rave, D. V., and Del Jesus Peñil, M.: The Future of Extreme Event Risk Assessment: A Look at Multivariate Return Periods in More than Three Dimensions, EGU General Assembly 2023, Vienna, Austria, 24–28 Apr 2023, EGU23-11828, https://doi.org/10.5194/egusphere-egu23-11828, 2023.
Accurate rainfall modeling is crucial to understand the way water is intercepted, infiltrates and flows through surfaces and rivers. In particular, it is paramount for the study of the influence of rainfall spatio-temporal distribution on basin hydrologic response and the structure of floods. Current weather radar products allow capturing the variability of rainfall extremes mainly at 1 km spatial resolution. In France, radar measurements are performed at a 5-minute time step, while gauge-based reanalysis are computed at hourly resolutions. During short-duration high-intensity precipitations, pluvial floods (PF, or flash floods) can occur outside the hydrographic network in runoff-prone areas, leading to various types of damages such as soil erosion, mud and debris flows, landslides, vegetation uprooting or sediment load deposits. Contrary to fluvial floods, PF are highly correlated to local rainfall. Depending on generic susceptibility linked to topography, soil texture and land use, specific precipitation patterns can trigger intense overland flow. Hence, after extreme weather events, precise reports on PF locations provide key information for rainfall reanalysis and downscaling at fine spatial resolution.
This work focuses on two extreme Mediterranean events (more than 300 mm of rainfall in 24 hours) that took place in the South of France between 2018 and 2020. Time series of hourly rainfall intensities from Comephore radar reanalysis data at 1 km resolution (Météo-France) are confronted to (i) maps of PF that occurred during the events and (ii) generic susceptibility maps to surface runoff. For (i), runoff-related impact maps of the events are produced using the remote sensing-based FuSVIPR algorithm (Cerbelaud et al., 2023) based on Sentinel-2 temporal change images and Pléiades satellite or airborne post event acquisitions. For (ii), the IRIP© method (Dehotin and Breil, 2011; Cerbelaud et al., 2022) is used to generate PF susceptibility maps. The model is run with the RGE Alti® 5 m DEM, the OSO French land cover dataset, and soil type susceptibility characteristics derived from both climatological information and the ESDAC database.
We primarily show that areas with higher IRIP levels are more likely to be impacted by PFs, and even more so where short-term precipitation was heavier. Additionally, rainfall intensities are negatively correlated with IRIP susceptibility scores in PF impacted areas. This corroborates that somewhat higher rainfall intensities are required for flash floods to occur in less susceptible areas. Similarly, comparatively smaller rainfall amounts can trigger PFs in locations where susceptibility is high. Then, the Comephore products are downscaled at 50 m resolution on both events using the SAMPO stochastic simulator (Leblois and Creutin, 2013). Among multiple scenarios, optimal ones are chosen based on the assumption that the negative correlation with the IRIP susceptibility levels in the affected areas should be equally or even more present in the downscaled rainfall time series. This study hence suggests an original way of selecting disaggregated extreme rainfall scenarios that are consistent with the observed consequences of intense runoff on the land surface using various tools such as a stochastic simulator, a hydrological risk mapping method and earth observation data.
How to cite: Cerbelaud, A., Leblois, E., Breil, P., Roupioz, L., Rodriguez-Suquet, R., Blanchet, G., and Briottet, X.: Space-time downscaling of extreme rainfall using stochastic simulations, intense runoff susceptibility modeling and remote sensing-based pluvial flood mapping, EGU General Assembly 2023, Vienna, Austria, 24–28 Apr 2023, EGU23-12249, https://doi.org/10.5194/egusphere-egu23-12249, 2023.
Coastal cities are increasingly prone to compound flooding events. Particularly in estuaries, interactions between both freshwater fluxes (rainfall or discharge) and coastal water levels (tide, surge, waves, or combinations thereof) can strongly modulate flood hazard. These separate but physically connected processes can often occur simultaneously (but not necessarily in extreme conditions), resulting in compound events that may eventually have significant economic, environmental and social impacts. Conventional risk assessment mainly considers univariate-flooding drivers and does not include multivariate approaches; nevertheless, ignoring compound analysis may lead to a significant misestimation of flood risk.
In this respect, the complex interactions between coastal flooding drivers imply multidimensionality, nonlinearity and nonstationarity issues, and consequently, more relevant uncertainties. Copula-based frameworks are flexible alternatives to overcome limitations of traditional univariate approaches, and can incorporate the joint boundary conditions in riverine and coastal interactions in a statistically sound way (Harrison et al., 2021; Bevacqua et al., 2019; Couasnon et al., 2018, Moftakhari et al., 2017). However, incorporations are often limited to the bivariate joint case. Trivariate (or higher dimensional) joint distribution are scarce, due to the convoluted and computationally expensive composition (Latif & Sinonovic, 2022). Notably, a need for robust and efficient approaches that help to characterize the nature of compound hazard remains (Moftakhari et al., 2021).
This study aims to improve copula-based methodologies that can adequately estimate the compound flood probability in estuarine regions, considering more than two variables, including more sources of uncertainty into the stochastic dependence analysis, raising the degree of accuracy to risk inference. This work develops a vine copula framework for the analysis of estuarine compound flooding risk, considering interactions and dependency structures between several oceanographic, hydrological, and meteorological processes and variables (rainfall, river discharge, waves, and storm tides). We show the potential of the framework in Santoña, a strategic estuarine ecosystem in Northern Spain. In order to yield proper design events, we focus here on estimating the multivariate joint and conditional joint return periods statistics, using the best-fitted model in the assessment of the extreme regime, based on Archimedean and Elliptical copula families. We also present the complexities of determining the ensemble of compound events corresponding to a given return period and compare these ensembles to the results of univariate extreme value analysis, to remark the importance of multivariate characterization of extremes.
References
Bevacqua, E., Maraun, D., Vousdoukas, M. I., Voukouvalas, E., Vrac, M., Mentaschi, L., & Widmann, M. (2019). Higher probability of compound flooding from precipitation and storm surge in Europe under anthropogenic climate change. Science advances, 5(9), eaaw5531.
Couasnon, A., Sebastian, A., & Morales-Nápoles, O. (2018). A copula-based Bayesian network for modeling compound flood hazard from riverine and coastal interactions at the catchment scale: An application to the Houston Ship Channel, Texas. Water, 10(9), 1190.
How to cite: Gomez Rave, D. V., Urrea Méndez, D. A., and Del Jesus Peñil, M.: Multivariate Probability Analysis of Compound Flooding Dynamics., EGU General Assembly 2023, Vienna, Austria, 24–28 Apr 2023, EGU23-12736, https://doi.org/10.5194/egusphere-egu23-12736, 2023.
Daily precipitation extremes are crucial in the hydrological design of major water control structures. The extremes are usually present in the upper part of the probability distribution of daily precipitation data, known as the tail. The distributions are bifurcated into heavy or light-tailed distributions depending on the tail. Heavy tails signify a higher frequency of occurrences of extreme precipitation events. Prediction of extreme precipitation depends on reliable modelling of the tail. Tail behaviour can be studied by graphical as well as threshold-based fitting approaches; however, each approach has associated shortcomings. In this work, we utilize a versatile and simple empirical index known as the “Obesity Index” (OB) to assess the tail of probability distributions of daily gridded precipitation data for India. This comprehensive regional analysis has been undertaken to quantify the tail heaviness of 4801 daily precipitation records over India for historical (1970–2019) and future (2020–2100) time periods. Future projections of daily precipitation are downscaled from the latest generation of climate models knowns as Coupled Model Intercomparison Project Phase 6 (CMIP6) under different emission scenarios. Finally, the application of the OB-based approach is extended to characterize daily precipitation in Indian Meteorological Subdivisions. Results indicate the applicability of heavy-tailed distributions in representing daily precipitation over India and establish the utility of the OB-based approach in diagnosing tail behaviour. Also, the spatial patterns of the tail heaviness are found to be matching with the Köppen–Geiger climate classification of India. The findings from this can be an input for the policymakers to develop adaptation strategies in response to the projected climate change impact.
Keywords: Extreme precipitation, Climate Change, India, Obesity index, Tail heaviness
How to cite: Gupta, N. and Chavan, S.: Assessing daily precipitation tails over India under changing climate, EGU General Assembly 2023, Vienna, Austria, 24–28 Apr 2023, EGU23-13328, https://doi.org/10.5194/egusphere-egu23-13328, 2023.
Walls of water (WoW) are a subset of flash floods characterised by an extremely fast increase in the discharge rate of rivers. In the UK, WoWs, events where an almost instantaneous increase in river flow happens, are responsible for several deaths, even when the maximum peak flow of the said event is not as noticeable. Using a national 15-minute continuous dataset, this study identified WoWs for catchments in the UK. Next, the antecedent atmospheric conditions for these WoWs were extracted from gridded datasets. Furthermore, catchment descriptors such as catchment area, elevation, slope, land use, and permeability of every catchment were downloaded from the National River Flow Archive. Finally, with the use of machine learning algorithms, that is, tree regressions and neural networks, this study identified vulnerable catchments and key conditions for WoWs to occur. Early results indicate that WoWs are not solely driven by rainfall intensity and that larger catchments (>500km) with low permeability are the most vulnerable to these hazards. Further studies using additional atmospheric conditions, i.e., temperature and windspeed will allow a better understanding of the drivers of these events.
How to cite: Fileni, F., Archer, D., Fowler, H., McLay, F., Lewis, E., and Yang, L.: Use of high temporal resolution data to identify the key drivers and locations where walls of water occur in the UK, EGU General Assembly 2023, Vienna, Austria, 24–28 Apr 2023, EGU23-13386, https://doi.org/10.5194/egusphere-egu23-13386, 2023.
South-western France has witnessed some of the most devastating extreme precipitation events that eventually lead to record-breaking severe flash flooding in the region and cause losses to human lives, urban transportation, agriculture, and infrastructure. In this study, two cases of deadly flash floods that occurred/reported in the Aude watershed in south-western France during 12-13 November 1999 and 14-15 October 2018 are studied using the WRF4.3.1 model simulations, with a particular emphasis on the model ability to capture these heavy precipitation events. We performed two simulations one with parameterized convection and one without the use of convection parameterizations for each case at gray-zone resolution (~9 km horizontal grid spacing) using the ERA5 reanalysis as the lateral boundary condition. In addition, attempts have been made to investigate the role of large-scale atmospheric circulation and atmospheric rivers in the production of these heavy precipitation events. The results from model simulations are compared quantitatively with available observations and reanalysis and found that the simulations at ~9 km gray-zone resolution capture the observed spatio-temporal distribution of precipitation characteristics during both extreme cases. The added value of gray-zone resolution simulations over driving coarse-scale ERA5 reanalysis datasets is observed in the representation of the precipitation characteristics. It has also been observed that the model simulation without the use of convection parameterization yields a reasonable and realistic representation of the precipitation characteristics during both extreme cases, and this suggests that at this “gray-zone” resolution the organized mesoscale convective systems/processes can be resolved explicitly by the model dynamics. The contribution of the large-scale atmospheric circulation and the atmospheric river (i.e., moisture transport) in the production of these flood events has also been observed.
How to cite: Shahi, N. K., Zolina, O., Gulev, S. K., Gavrikov, A., and Jomaa, F.: Simulation of extreme precipitation events over south-west France: the role of large-scale atmospheric circulation and atmospheric rivers, EGU General Assembly 2023, Vienna, Austria, 24–28 Apr 2023, EGU23-13732, https://doi.org/10.5194/egusphere-egu23-13732, 2023.
The growing number of extreme hydrological events observed has raised the level of attention toward the impact of climate change on rainfall process, which is difficult to quantify given its strong spatial and temporal heterogeneity. Therefore, the impact of the climate cannot be determined on the individual hydrological series but must be assessed on a regional and/or district scale. With this objective, the present work aims at identifying the trends and dynamics of extreme sub-daily rainfall in southern Italy in the period 1970-2020. The database of annual maxima was assembled using all available rainfall data (provided by the National Hydrographic and Mareographic Service - SIMN, and the Regional Civil Protection). However, due to the numerous changes (location, type of sensor, managing agencies) experienced by the national monitoring network, the time-series were found to be extremely uneven and fragmented. Since the spatio-temporal discontinuity could invalidate any statistical analysis, gap-filling techniques (deterministic and/or geostatistical [Teegavarapu, 2009]) were applied to reconstruct the missing data. In particular, the “Spatially-Constrained Ordinary Kriging” (SC-OK) method [Avino et al., 2021] was used, namely a mixed procedure that adopts the Ordinary Kriging (OK) approach with the spatial constraints of the Inverse Distance Weighting (IDW) method. The SC-OK method allows to reconstruct only missing data for stations selected by the IDW method (those with a sufficient number of functioning neighbouring rain gauges). Then, the reconstructed dataset has been used to explore trends and regional patterns in annual maxima highlighting, how rainfall are evolving in the most recent years.
REFERENCES
Avino, A., Manfreda, S., Cimorelli, L., and Pianese, D. (2021). Trend of annual maximum rainfall in Campania region (Southern Italy). Hydrological Processes, 35.
Teegavarapu, R.S.V. (2009). Estimation of Missing Precipitation Records Integrating Surface Interpolation Techniques and Spatio-temporal Association Rules. Journal of Hydroinformatics, 11(2).
How to cite: Avino, A., Cimorelli, L., Pianese, D., and Manfreda, S.: Updating annual rainfall maxima statistics in a data-scarce region, EGU General Assembly 2023, Vienna, Austria, 24–28 Apr 2023, EGU23-14934, https://doi.org/10.5194/egusphere-egu23-14934, 2023.
For many flood risk assessments at large spatial scales, long-term meteorological data (e.g. precipitation, temperature) with spatially coherent representation are needed. This is where a regional weather generator comes into play. Meteorological fields for a specific region are strongly dependent on weather circulation patterns (CP) at larger scales. Additionally, there is evidence that these fields covariate with the average regional surface temperature (ART). With future climate change, such changes in both CP and ART should be included in weather generators.
This study presents the development of such a non-stationary gridded weather generator conditioned on large-scale weather circulation patterns for Central Europe. The reanalysis dataset ERA5 (1o x 1o) is used for weather type classification. The E-OBS gridded observational dataset (0.25ox 0.25o) is used to parameterize the meteorological fields, such as precipitation and temperature (minimum, maximum, average). The spatial and temporal dependence is represented by the multivariate auto-regressive model. Daily precipitation amount is modelled by the extended generalized Pareto distribution and daily temperature is modelled by the transformed normal distribution. Both fields are conditioned on CP and allow to covariate with ART. In this way, the regional weather generator is capable of capturing “between-type” and “within-type” climate variability and can be used to generate long synthetic data for flood risk assessment in present and future periods.
How to cite: Nguyen, V. D., Vorogushyn, S., Nissen, K., and Merz, B.: A non-stationary gridded weather generator conditioned on large-scale weather circulation patterns for Central Europe, EGU General Assembly 2023, Vienna, Austria, 24–28 Apr 2023, EGU23-15475, https://doi.org/10.5194/egusphere-egu23-15475, 2023.
The recent AR6 report of the Intergovernmental Panel on Climate Change (IPCC) explicitly shows that the observed change in hot extremes (including heatwaves) with high confidence in human contribution to the observed changes has highly increased in the South Asian (SAS) domain which comprises the Indian subcontinent. Extreme heat events are more frequent and intense across the globe since the 1950s and have adverse societal and economic impacts. Considering current warming trends and projections, heatwaves are becoming a serious problem in India. Exposure to extreme heat in the population is increasing due to climate change. Also, observed temperatures are increasing globally as well as regionally as an effect of global warming. As heat stress occurs when the human body cannot get rid of the excess heat, it can be considered a good proxy for the heatwave hazard. Heat stress results in heat stroke, exhaustion, cramps, or rashes. Exposure to extreme heat can result in occupational illnesses and injuries. An agrarian country like India will have large economic damage when climate-related heat stress increases the occurrence of droughts and exacerbate water scarcity for irrigation. Hence the impact of the heat stress hazard is spotted and largely discussed both in the academic and political domains. In this study, Universal Thermal Climate Index (UTCI) based hazard map is developed for India with a non-parametric multivariate approach. The prominent heat stress hazard areas are identified and mapped with reference to the UTCI assessment scale which is categorized based on thermal stress. The probability of occurrence is also mapped using the exceedance probability with the UTCI reference. Heat stress hazard map provides the basis for a wide range of applications in public and individual precautionary planning such as heatwave action plans, urban and regional planning, the tourism industry, and climate research. Hence a country-level extreme temperature hazard map is of dire necessity.
Keywords: Exceedance probability, hazard map, heat stress, multivariate approach, non-parametric method
How to cite: Shilin, A., Sudharsan, N., Mondal, A., Kalbar, P., and Karmakar, S.: Mapping Hazard to Extreme Temperature Events Over the Indian Subcontinent, EGU General Assembly 2023, Vienna, Austria, 24–28 Apr 2023, EGU23-16623, https://doi.org/10.5194/egusphere-egu23-16623, 2023.
Posters virtual: Fri, 28 Apr, 08:30–10:15 | vHall HS
Alaba Boluwade*
School of Climate Change & Adaptation, University of Prince Edward Island, Charlottetown, Canada; aboluwade@upei.ca; abolu2013@gmail.com
*Correspondence: aboluwade@upei.ca
Abstract
Hydrological risk assessment, such as flood protection, requires estimates of variables (e.g., precipitation) measured from several weather stations. The spatial modeling of average rainfall estimates differs from extreme precipitation analysis. This is because extremes are focused on the tail of the probability distribution and the assumption of Gaussianity may not be suitable. Extreme Value Theory (EVT) application to univariate weather variables measured at weather stations has been well documented; however, extreme precipitation at closer stations tend to show trends and dependencies (similar values). It is, therefore, crucial to quantify the dependency structure and trend surface of weather stations in space. The max-stable process has been well documented to model spatial extremes. The objective of this study is to quantify the spatial dependency and trend of an annual maxima precipitation (annual highest daily precipitation, from 1970-2020) across selected weather stations in the Northern Great Plains (i.e., Nelson Churchill River Basin (NCRB)) of North America. The annual maxima data were extracted from the Global Historical Climatology Network Daily (GHCNd) and Environment and Climate Change Canada (ECCC). NCRB covers four states and four provinces in the United States and Canada. A heterogenous rainfall pattern characterizes NCRB. This is due to enormous quantities of orographic rainfall in the west and the convective precipitation in the Prairies (which is dominated by short-duration, sporadic, extreme rain), causing millions of dollars in damages. This study uses max-stable processes to examine spatial extremes of annual maxima precipitation.
The results show that topography, time, and geographical coordinates were important covariates in reproducing the stochastic extreme precipitation field using the spatial generalized extreme value (SPEV). Takeuchi’s information criterion (TIC) shows that the SPEV model with all the covariates above superseded the one without the covariates. The inclusion of time as a covariate further confirms the impacts of climate change on extreme precipitation in the NCRB. The fitted Extremal-t max-stable model captured the spatial dependency and equally predicted the 50-year return period levels. Furthermore, ten realizations (equal probable) were simulated from the max-stable model. The study is relevant in quantifying the spatial trend and dependency of extreme precipitation in the Northern Great Plains. The result will help as a decision-support system in climate adaptation strategies in the United States and Canada.
Keywords: extreme events; Max-Stable processes; flood protection; maxima annual rainfall; flash flood protection; Canada, United States
How to cite: Boluwade, A.: Application of Max-Stable Process Model in Estimating the Spatial Trend & Dependency of Extremes in the Northern Great Plains, EGU General Assembly 2023, Vienna, Austria, 24–28 Apr 2023, EGU23-9758, https://doi.org/10.5194/egusphere-egu23-9758, 2023.
