Displays

Flooding can affect large regions leading to high economic and societal costs. Estimating regional flood risk is crucial for developing adaptation strategies, public awareness policies, and protection structures. Yet, estimating regional flood hazard is not trivial because of the few large flood events observed. Here, we derive regional flood hazard estimates for large river basins in the United States by using a stochastic streamflow generator. This allows us to increase the number of flood events available for the analysis and to investigate the simultaneous occurrence of flooding in different parts of a river basin.
We propose the continuous, stochastic simulation approach (PRSim.wave), which combines a non-parametric spatio-temporal model based on the wavelet transform with the parametric kappa distribution. The model reproduces the temporal and distributional characteristics of streamflow at individual sites and retains the spatial dependencies between sites even for spatial extremes. We use PRSim.wave to generate long and spatially consistent time series of daily discharge for a large set of catchments in the conterminous United States. For each catchment, we extract flood events from the simulated series using a peak-over-threshold approach to derive a spatial dataset of flood occurrences. Using this dataset, we estimate how probable it is that a certain percentage of stations within a specific river basin is jointly flooded. We show that: (1) there are strong regional differences in the likelihood of joint and potentially widespread flooding and (2) there are spatial differences in regional flood hazard estimates which could not be derived from observed data only. We deem our approach a valuable tool for water managers and policy makers to make informed decisions on the risk of widespread flooding.

How to cite: Brunner, M. I., Papalexiou, S., and Gilleland, E.: How likely are widespread floods in US river basins? Seeking answers using a stochastic, wavelet-based approach, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-1781, https://doi.org/10.5194/egusphere-egu2020-1781, 2020.

Climate change has led to changing flood synchrony scales (extents) and flood magnitudes across Europe. We discovered a tight alignment between extents and magnitudes and found the drivers of their joint trends. We analyzed the annual maximum floods of 3872 hydrometric stations across Europe from 1960-2010 and classified all floods in terms of their generating processes based on antecedent weather conditions. There is a positive correlation between flood extents and magnitudes for 95% of the stations. While both parameters increased in Central and Western Europe, they jointly decreased in the East. This widespread magnitude extent correlation is caused by similar correlations for precipitation, soil moisture and snowmelt. We found trends in the relevance of the different flood generation processes, which explain the regional flood trends. The aligned increases of flood extents and magnitudes emphasize the growing importance of transnational flood risk management.

How to cite: Kemter, M., Merz, B., Marwan, N., Vorogushyn, S., and Blöschl, G.: Mutual increases in flood extents and magnitudes intensify flood hazard in Central and Western Europe, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-4731, https://doi.org/10.5194/egusphere-egu2020-4731, 2020.

     For engineering practice, the determination of design storms mainly applies Intensity-duration-frequency method or revised ranking method with fixed given duration based on the long-term rainfall record. However, this kind of method does not address the influence of storm duration. Both long or short storms are combined together in performing frequency analysis. As a result, in some cases, the peak intensity of short duration design storm might be overestimated by traditional methods. Consequently, how to incorporate the concept of duration to improve current frequency analysis is essential to improve design storms are the two main purposes of this research. With this concern, this study tries to find better definitions of the rainfall characteristics based on different durations, and generate revised design storm in the consideration of more comprehensive rainfall patterns. For the purpose of flood control, this study tries to proposed the improved method based on a conceptual watershed with known drainage capacity. In the watershed, when the runoff exceeds this capacity, the accumulation of the surface runoff will become the flood water, and keep increasing with raining time. Considering this situation, the duration of storm plays an important role in the severity of flooding. For this concern, this study tries to incorporate two parameters, the peak intensity, and time that rainfall exceeds drainage capacity, to indicate the severity of storm. Following, to avoid the under/over-estimation of rainfall intensity without considering the influence of duration, we propose the data weight index to composed design storm by combining different duration rainfall events. More specifically, if the duration of rainfall event is similar to the design storm, the parameters of data will get a higher weight index to shape the rainfall characteristics of design. With this process, we can obtain the suitable parameters for our design storms and compare the risk of flood with any two storms with different duration in a same drainage capacity. Based on the equivalent risk of storms, we could transform the parameters of the rainfall with the same risk level even for different duration. By this way, this study will propose a method of design storms which offers a better application for flood protection plan and management.

Keywords: Design storm, Urban rainfall design, Hydrologic frequency analysis

How to cite: Yin-Huan, M. and Gene, J.-Y.: The consideration of storm durations in hydrological frequency analysis and design storm determination, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-6868, https://doi.org/10.5194/egusphere-egu2020-6868, 2020.

This research presents a methodology for the hydrological characterisation of the overtopping failure for an existing river levee. Conventional procedures usually consider only one variable for the hydrologic forcing, e.g. peak discharge Q. Such an approach might fail if also the volume of the hydrograph V is a significant factor.  The proposed procedure is based on the generation of a set of plausible hydrographs. Each hydrograph has a couple of Q and V resulting from an approximated bivariate distribution. The shapes of the hydrographs are classified according to their tendency to produce overtopping, introducing the new Overtopping Hydrograph Shape Index (OHSI). The levee of a river reach located in Tuscany Region, Italy, was tested. As a preliminary result, it was found that the hydrographs that produce overtopping failure lay in a zone in the Q-V space, delimited by the Critical Overtopping Flood Hydrograph (COFH) curve. The existence of the COFH demonstrates that overtopping failure is not determined by a unique variable Q but rather by the combination of both Q and V. The limiting case corresponds to a family of hydrographs with varying Q and V values. The COFH allows dividing the Q-V space into two zones: the no-failure zone and the failure zone, which simplify the evaluation of the failure probability for a river levee.  This procedure is based on tools that are widely known and it is replicable by the public administrations or public entities that are interesting in the hydrologic and hydraulic risk assessment.

How to cite: Isola, M., Caporali, E., and Garrote, L.: A methodology for the bivariate hydrological characterisation of the overtopping failure for river levees, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-19021, https://doi.org/10.5194/egusphere-egu2020-19021, 2020.

The Standardized Precipitation Index (SPI) is a widely accepted drought index. Its calculation algorithm normalizes the index via a distribution function. Which distribution function to use is still disputed within literature. This study illuminates the long-standing dispute and proposes a solution which ensures the normality of the index for all common accumulation periods in observations and simulations.

We compare the normality of SPI time-series derived with the gamma, Weibull, generalized gamma, and the exponentiated Weibull distribution. Our normality comparison evaluates actual against theoretical occurrence probabilities of SPI categories, and the quality of the fit of candidate distribution functions against their complexity with Akaike's Information Criterion. SPI time-series, spanning 1983–2013, are calculated from Global Precipitation Climatology Project's monthly precipitation data-set and seasonal precipitation hindcasts from the Max Planck Institute Earth System Model. We evaluate these SPI time-series over the global land area and for each continent individually during winter and summer. While focusing on an accumulation period of 3-months, we additionally test the drawn conclusions for other common accumulation periods (1-, 6-, 9-, and 12-months).

Our results suggest to exercise caution when using the gamma distribution to calculate SPI; especially in simulations or their evaluation. Further, our analysis shows a distinctly improved normality for SPI time-series derived with the exponentiated Weibull distribution relative to other distributions. The use of the exponentiated Weibull distribution maximizes the normality of SPI time-series in observations and simulations both individual as well as concurrent. Its use further maximizes the normality of SPI time-series over each continent and for every investigated accumulation period. We, therefore, advocate to derive SPI with the exponentiated Weibull distribution, irrespective of the heritage of the precipitation data or the length of analyzed accumulation periods.

How to cite: Pieper, P., Düsterhus, A., and Baehr, J.: Global and regional performances of SPI candidate distribution functions in observations and simulations, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-2634, https://doi.org/10.5194/egusphere-egu2020-2634, 2020.

Statistical analysis of hydrologic variables is of great importance for water resources systems. Design and operation of these systems is often based on the assumption of data stationarity. However, long-term average of variables such as rainfall as well as sea level is observed to shift over time, mostly attributed to the climate change. These changes, in turn, affect flood volume, peak value and frequency. In this study, a framework was proposed for bi- variate frequency analysis of extreme sea level and rainfall. The analysis was performed on rainfall for the coastal area of Charleston and Savannah, and sea level for the coastal area of Charleston and Fort Pulaski, South Carolina, USA. Extreme values were selected based on the peak over threshold method. To determine the most appropriate distribution, AIC and BIC goodness of fit tests were used. Frequency analysis was then carried out using nonstationary Generalized Extreme Value probability distribution function. Results showed an increase in the sea level long term average, significant trends and outliers (specifically in recent decades), while although the analysis of rainfall data confirms the presence of outliers in the time series, it does not indicate significant trends or heterogeneity. Therefore, in performing bi-variate frequency analysis of extreme rainfall and sea level, non-stationary approaches should be used to provide a more reliable prediction of the joint probability of these variables.

How to cite: Golian, S., Razmi, A., Mardani, H. A., and Zahmatkesh, Z.: Nonstationary Bi-Variate Frequency Analysis of Extreme Sea Level and Rainfall Under Climate Change Impacts: South Carolina Coastal Area, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-3098, https://doi.org/10.5194/egusphere-egu2020-3098, 2020.

Land conversion from semi-natural grassland to intensively managed pasture (for sheep, beef and dairy production) has altered the near-surface, soil moisture regime across much of the uplands of Europe. This widespread conversion has modified both the temporal distributions and spatial structure of surface volumetric wetness, thus affecting the incidence of flood-producing overland flow and resilience of the grasslands to drought stresses. In order to investigate these spatiotemporal dynamics, an intensive fieldwork campaign captured high-resolution (1m²) surface volumetric wetness from a 1536m² paired-plot monitored over a year including both drought and fully saturated conditions. The measurements and combined statistical and geostatistical analyses form part of integrated studies into the hydrological effects of agricultural interventions to mitigate floods in the Cumbrian mountains of the UK.

The intensive monitoring highlighted significant temporal variations between land-uses. The pasture dried faster than the semi-natural grassland with the onset of a severe drought, but these effects were more than offset by the application of livestock slurry. This artificial wetting did however produce a more rapid build-up of moisture in the pasture with autumn storms. The large rain-event of Storm Diana (28-29 Nov 2019) did, however fully saturate both the pasture and semi-natural grassland to generate visible saturation-excess overland flow. Seasonal changes in the spatial patterns of volumetric wetness were equally evident. The semi-natural grassland contained significantly larger variation within soil moisture statistical distributions and substantially larger coefficients of variation compared to the pasture throughout the study. Very weak spatial structure was observed within the semi-natural grassland. Conversely, a relatively strong spatial structure was observed within the pasture plot, which intensified with saturation, suggesting farming practices (ploughing, reseeding, artificial inputs, etc.) have removed natural soil moisture variability and encouraged moisture redistribution. A geostatistical model showed that the weak semi-natural grassland spatial structure remained relatively stationary, whereas the pasture showed extreme non-stationarity, with increasing saturation causing a gradual transition from an exponential to a gaussian geostatistical relationship.

The work highlights the complexity of spatiotemporal soil moisture dynamics taking place at the metre- to decimetre-scales through wetting-and-drying cycles and the strong impact of pasture management upon this. It justifies the need for both intensive soil moisture sampling at experimental sites, sampling across seasons, and the need for combined statistical and geostatistical analyses. Further such analyses in the uplands of Europe are needed if we are to better understand the effects of pastureland management on flood and drought hydrology, and to use this knowledge to mitigate our impacts on floods and droughts.

How to cite: Wallace, E. and Chappell, N.: The spatiotemporal dynamics of surface soil moisture within upland grassland ecosystems, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-8798, https://doi.org/10.5194/egusphere-egu2020-8798, 2020.

In the traditional flood frequency analysis, researchers typically assume the flood events result from a homogeneous flood population. However, actually flood events are likely to be generated by distinct flood generation mechanisms (FGMs), such as snowmelt-induced floods and rainfall-induced floods. To address this problem in flood frequency analysis, currently, the most popular practice for mixture modeling of flood events is to use two-component mixture distributions (TCMD) without a priori classification of distict FGMs, which could result in component distributions without physical reality or lead to a larger standard error of the estimated quantiles. To improve the mixture distribution modeling in Norway, we firstly classify the flood series of 34 watersheds into snowmelt-induced long-duration floods and rainfall-induced short-duration floods based on an index named flood timescale (FT), defined as the ratio of the flood volume to peak value. A total of ten types of mixture distributions are considered in the application of FT-based TCMD to model the flood events in Norway. The results indicate that the FT-based TCMD model can reduce the uncertainty in the estimation of design floods. The improved predictive ability of the FT-based TCMD model is largely due to its explicit recognition of distinct FGMs, enabling the determination of the weighting coefficient without optimization.

How to cite: Yan, L., Xiong, L., Li, L., Ruan, G., Xu, C.-Y., and Liu, P.: Estimating design floods in Norway considering distinct flood generation mechanisms, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-1838, https://doi.org/10.5194/egusphere-egu2020-1838, 2020.

To assess the safety of dams, design floods are typically used as a basis. Of particular interest are events with a return period of 1’000 years and even rarer events derived from that with help of simple return period conversion factors given by design codes. However, both the peaks and even more the flood volumes of such rare events are subject to large uncertainties due to limited length and spatial coverage of gauge records. Bivariate approaches can help reduce the uncertainty related to the flood volumes. Nevertheless, both univariate and bivariate approaches require long-term observations on which the return periods of flood events can be calculated.

In this study, we make use of very long simulated hydrographs in hourly resolution for Swiss catchments (scale range: ~300–18’000 km²). The hydrographs span about 300’000 years each and stem from a hydro-meteorological modelling chain starting with a stochastic multi-site weather generator. With these hydrographs, we develop a framework to characterize design floods through a realistic hydrograph using functional data analysis as well as hydrographs that envelope 50%, say, of the most central observations (corresponding to the 25% and 75% quantiles in a univariate setting).

In a first step, we assigned the simulated annual maximum flood events to return period classes of 100, 1000 and 10000 years. We then built clusters of similar events within each class using functional clustering. Here we explore some of the possibilities of the approach and in particular show how sensitive the functional clustering is to the choice 1) of event characterization (peak only, flood peak and volume, flood volume given a minimum flood peak), and 2) to the separation of event and baseflow of the selected events in the bivariate case and 3) to the different functional latent mixture models that are applied within the functional clustering.

How to cite: Staudinger, M., Furrer, R., and Viviroli, D.: A framework to characterize flood events of defined return period ranges using functional boxplots, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-9950, https://doi.org/10.5194/egusphere-egu2020-9950, 2020.

Phase randomization and its variants such as the Amplitude-adjusted (AAFT) and the Iterative amplitude adjusted (IAAFT) Fourier transform are used to check statistical significance of a given hypothesis and/or to generate time series that are similar to a reference in some statistical sense. These methods have the drawback of producing incorrect dependence structures e.g. empirical copula density, asymmetries and entropies. Recently, another form of such methods, “Phase Annealing”, was introduced, giving a possibility to generate n-dimensional realizations of a process under given constraint(s). The main concern using this method is the selection of correct objective function(s).

Here we show discharge time series generation using Phase Annealing with new objective functions. This allowed us to generate time series that are much longer than the reference, which in turn was helpful in establishing better distributions of floods.

We also show the generation of discharge time series at multiple locations that have the correct spatio-temporal dependences among all the series. Using the results, we generated full distributions of simultaneous extremes at observation locations.

Further uses may include clustering catchments that are likely to bring floods together and reliability analysis i.e. simulating distributions of failures for a system with many dependent/independent components. Drawbacks using this method are also shown.

How to cite: Anwar, F. and Bárdossy, A.: Using Phase Annealing to generate surrogate discharge time series, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-7955, https://doi.org/10.5194/egusphere-egu2020-7955, 2020.

In November 2017, a storm event of substantial but unknown local intensity caused a flash flood in Western Attica, Greece, which was responsible for 24 human fatalities and large-scale economical losses. Our focus is to the neighbouring catchment of Sarantapotamos, which has been equipped with an automatic stage recorder that was destroyed during the rising of the flood. Our overall objective is the estimation of the rainfall over the broader area of interest, through a reverse rainfall-runoff modelling approach at this specific catchment. Several sources of information are accounted for in order to reproduce the “observed” flood hydrograph, including photos and videos. We then employ Monte Carlo simulations to evaluate the uncertainty induced from limited and even missing data. Utilising the outcome of these analyses, we provide probabilistic estimations of the modelled rainfall, as well as risk evaluations, by estimating the maximum intensities and associated return periods of the storm event across multiple time scales.

How to cite: Ntigkakis, C., Nezi, M., and Efstratiadis, A.: Post-extraction of flood hydrographs under limited and heterogeneous information: Case study of Western Attica event, November 2017, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-18262, https://doi.org/10.5194/egusphere-egu2020-18262, 2020.

Recent research has revealed the significance of Hurst-Kolmogorov dynamics and inherent uncertainties in flood inundation and flood mapping. However, classic risk estimation for flood insurance practices is formulated under the assumption of independence between the frequency and the severity of extreme flood events, which is unlikely to be tenable in real-world hydrometeorological processes exhibiting long range dependence. Furthermore, insurable flood losses are considered as ideally independent and non-catastrophic due to the widely spread perception of limited risk regarding catastrophically large flood losses. As the accurate risk assessment is a fundamental process on flood insurance and reinsurance practices, this study investigates the effects of lack of fulfillment of these assumptions, paving the way for a deeper understanding of the underlying clustering mechanisms of stream flow extremes. For this purpose, we present a spatiotemporal analysis of the daily stream flow series from the US-CAMELS dataset, comprising the impacts of clustering mechanisms on return intervals, duration and severity of the over-threshold events which are treated as proxies for collective risk. Moreover, an exploratory analysis is introduced regarding the stochastic aspects of the correlation between  the properties of the extreme events and the actual claim records of the FEMA National Flood Insurance Program which are recently published.

How to cite: Papoulakos, K., Iliopoulou, T., Dimitriadis, P., Tsaknias, D., and Koutsoyiannis, D.: Investigating the impacts of clustering of floods on insurance practices; a spatiotemporal analysis in the USA, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-8667, https://doi.org/10.5194/egusphere-egu2020-8667, 2020.

Both mean and extreme precipitation are highly relevant and a probability distribution that models the entire precipitation distribution therefore provides important information. Very low and extremely high precipitation amounts have traditionally been modeled separately. Gamma distributions are often used to model low and moderate precipitation amounts and extreme value theory allows to model the upper tail of the distribution. However, difficulties arise when making a link between upper and lower tail. One solution is to define a threshold that separates the distribution into extreme and non-extreme values, but the assignment of such a threshold for many locations is not trivial. 

Here we apply the Extended Generalized Pareto Distribution (EGPD) used by Tencaliec & al. 2019. This method overcomes the problem of finding a threshold between upper and lower tails thanks to a transition function (G) that describes the transition between the empirical distribution of precipitation and a Pareto distribution. The transition cumulative distribution function G has to be constrained by the upper tail and lower tail behavior. G can be estimated using Bernstein polynomials.

EGPD is used here to characterize ERA-5 precipitation. ERA-5 is a new ECMWF climate re-analysis dataset that provides a numerical description of the recent climate by combining a numerical weather model with observations. The data set is global with a spatial resolution of 0.25° and currently covers the period from 1979 to present.

ERA-5 daily precipitation is compared to EOBS, a gridded dataset spatially interpolated from observations over Europe, and to CMORPH, a satellite-based global precipitation product. Simultaneous occurrence of extreme events is assessed with a hit rate. An intensity comparison is conducted with return levels confidence intervals and a Kullback Leibler divergence test, both derived from the EGPD.

Overall, extreme event occurrences between ERA5 and EOBS over Europe appear to agree. The presence of overlap between 95% confidence intervals on return levels highly depends on the season and the probability of occurrence.

How to cite: Rivoire, P., Martius, O., and Naveau, P.: Characterization of ERA5 daily precipitation using the extended generalized Pareto distribution, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-7198, https://doi.org/10.5194/egusphere-egu2020-7198, 2020.

The estimation of design rainfalls and design floods are required by engineers and hydrologists to design and quantify the risk of failure of hydraulic structures. Extreme design rainfall quantities such as high-return period rainfalls and the probable maximum precipitation (PMP) are needed to design high-hazard hydraulic structures. In South Africa, previous design rainfall estimates have been produced up to the 200 year return period. PMP estimates were last determined nearly 50 years ago based on only 30 years of data. Most studies on extreme rainfall reported are based on frequency analysis assuming stationary conditions. Previous studies in South Africa have assumed a stationary climate. However, the assumption of a stationary climate in rainfall and flood frequency analysis has been challenged owing to evidence of climate change. Recent literature indicates that the magnitude and frequency of extreme rainfall events has been changing and this is likely to continue in future. Hence, methods to account for trends in extreme rainfall events in a changing environment need to be developed. In addition, the concept of PMP, particularly as used for the design and safety evaluation of large dams in South Africa, is being challenged with the recommendation that high-return period design rainfalls be used in these assessments. The aims of this study are: (i) to estimate extreme design rainfall values, with a focus on return periods greater than 200 years, (ii) to update PMP estimates using updated data and modernised methods, and (iii) to account for non-stationary climate data in the estimation of these extreme rainfall events in South Africa. Frequency analysis using LH-moments, which more accurately fit the upper tail of distributions, have been used to estimate high-return period design rainfalls. Regular L-moments are shown to overestimate the extreme rainfall quantities when compared to LH-moments by giving undue favour to outliers. PMP estimates have been determined using a storm maximisation and transposition approach. Radial Basis Functions (RFBs) have been used to transpose PMP estimates to ungauged locations, producing PMPs for the entire country. Approximately 80 % of the new PMPs are greater than the previous estimates. This is probably due to the many limitations of the old approach and differences used in the new approach, indicating that the new approach undertaken in this study may provide improved estimates. The PMP represents the upper limit of extreme rainfall, however, comparisons of high-return period rainfalls to the PMP show that the PMP is sometimes exceeded by the high-return period rainfalls. To develop methods to estimate extreme design rainfall events in a non-stationary climate, this study explores the impacts of climate drivers, such as the Southern Oscillation Index (SOI), and changes in atmospheric variables, such as dew point temperature, on high-return period rainfalls and the PMP.

How to cite: Johnson, K. and Smithers, J.: Estimation of Extreme Rainfall in South Africa and Development of Methods to Account for Non-stationary Climate Data, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-3326, https://doi.org/10.5194/egusphere-egu2020-3326, 2020.

The hydrological regime of rivers in Luxembourg (Central Western Europe) is characterised by summer low flows and winter high flows. In winter, large-scale floods are typically triggered by long-lasting sequences of precipitation events, related to westerly atmospheric fluxes that carry wet and temperate air masses from the Atlantic Ocean. In recent years, several flash flood events have been observed in Luxembourg. While being a common feature of Mediterranean river basins, this type of flooding events is uncommon at higher latitudes. The design of the hydro-meteorological monitoring and forecasting systems operated in Luxembourg is not adapted to this type of extreme events and there is a pressing need for a better mechanistic understanding of flash flood triggering mechanisms.

Here, we explore two lines of research – focusing on (i) the spatio-temporal variability of flash flood generation across a set of 41 nested catchments covering a wide range of physiographic settings (with mixed land use, soil types and bedrock geology) and (ii) the responsivity (resistance) and elasticity (resilience) of these catchments to global change.

Our area of interest is the Sûre River basin (4,240 km²), characterised by a homogenous climate (temperate oceanic), as well as various bedrock (e.g. sandstone, marls, shale) and land use (e.g. forests, grassland, crops, urban areas) types. Based on 8 years’ worth of daily hydrological data (precipitation, discharge and potential evapotranspiration) we computed the increments of the water balance to determine the maximum storage capacity and pre-event wetness state (expressed as storage deficit). Based on the relationship between storage deficit and discharge we first estimated total storage at nearly zero flow conditions. Second, we compared event runoff ratios (Q/P) to pre-hydrological states (as expressed to storage deficit prior to a rainfall-runoff event) in order to assess each catchment’s sensitivity to antecedent wetness conditions. Third, we assessed the responsivity (resistance) and elasticity (resilience) to climate variations – as expressed through the PET/P and AET/P deviations from the Budyko curve – for each individual catchment. Finally, we investigated potential physiographic controls on catchment responsivity and elasticity across our set of 41 nested catchments.

How to cite: Tamez-Melendez, C., Meyer, J., Douinot, A., Blöschl, G., and Pfister, L.: Physiographic controls on pre-event hydrological states and hydrological response to extreme precipitation in the Alzette River Basin, Luxembourg, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-8954, https://doi.org/10.5194/egusphere-egu2020-8954, 2020.

With a global concern about climate nonstationary and predictions of more extreme weather events, considering new rainfall distribution patterns is necessary using the most current and complete data available at any location. In this study, extreme rainfall frequency is analyzed using daily precipitation data in Kansas located in the central United States and Tuscany in the central Italy. From Kansas, 39 stations with data from 1920-2009 are selected, while for from Tuscany Region, 472 stations with daily time series of at least 15 years in the period 1916-2017 are used in the analysis. Initial analysis showed an increase in extreme precipitation events in Kansas with extreme event values tending to increase in magnitude from the northwest to southeast part of the state. Comparing results of the first period (1920-1949) to the last of three study periods (1980–2009) showed that approximately 90% of the state had an increase in short-term rainfall event magnitudes. Long-term event magnitudes were predicted to be higher in 66% of the state. Tuscany analysis is being conducted. Generally, results show a shift in rainfall distribution patterns in Kansas and Tuscany spatially and temporally. This shift changes the design criteria for hydraulic infrastructures, both in runoff control and storage structures.

How to cite: Rahmani, V. and Caporali, E.: How changes in extreme precipitation impacts hydraulic structure design storms, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-12321, https://doi.org/10.5194/egusphere-egu2020-12321, 2020.

Runoff simulation using hydrological models has a key role in water resources
management. Thus, there is a need to investigate how rainfall-runoff models preserve
the stochastic characteristics of real-world streamflow data. It is also useful to compare
the stochastic properties of output with those of input processes (rainfall) and internal
state variables (soil moisture), with focus on marginal distribution tails and long-term
persistence. To this aim, we perform a case study using the ENNS rainfall-runoff model
with real and synthetic rainfall time series, and for all processes we study the marginal
distributions and the dependence structures. In the analyses we use recently developed
stochastic tools such as K-moments and climacograms.

How to cite: Kourakos, V., Iliopoulou, T., Dimitriadis, P., Koutsoyiannis, D., Kaleris, V., and Langousis, A.: Investigation of marginal distribution and dependence structure of simulated streamflow by a rainfall-runoff model, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-11402, https://doi.org/10.5194/egusphere-egu2020-11402, 2020.

Floods are one of the most recurrent and damaging natural disasters in cities in developing countries today. The city of Abidjan (economic capital of Côte d'Ivoire) is not immune to these disasters. Indeed, according to the ONPC (2014), floods have killed an average of 13 people each year since 2009. One of the causes of the worsening and recurrence of these disasters is the urbanization experienced by the city of Abidjan in recent years. This urbanization has resulted in an increase in waterproofed areas and at the same time an increase in the volume of runoff water. The main objective of this research is to study the dynamics of floods and to see the impact of human activities on the hydrological functioning of the Gourou and Bonoumin watersheds in order to anticipate the risks of flooding. To do this, several specific objectives have been assigned in this study. These are: (1) the morphometric characterization of each watershed ; (2) the cartographic and diagnostic of stormwater management works and the dynamics of land use as well as the study of the variability of rainfall relative to the resurgence of floods in each basin ; (3) the establishment of a geographic information system for the study of floods; and (4) the prospective study of the evolution of land use and the future hydrological functioning of the watersheds studied using scenarios in order to take measures to fight against floods. The equipment used is composed of rain gauges (for measuring the spatial variation of rainfall in each sub-basin), limnimetric scales (to measure the heights of water in rivers), a double-ring infiltrometer (to measure the infiltration capacity on the basins) and software (Goldsim for simulating the behavior of watersheds). The data used consist of climatic data (temperature, rain), historical data on the floods in Abidjan, land use data (satellite images), physical characterization data of the watersheds (slope, hydrographic network, sanitation network , rainwater management works, morphometric parameters of the basins. The methodology adopted consisted of (1) collecting historical data on past floods in Abidjan; (2) description of the environment during rainy events; (3) studying the morphometric parameters of the watersheds studied and also (4) the study of the evolution of rainfall. Thus, the results obtained made it possible to show that the Gourou and Bonoumin basins cover respectively an area of 27.42 km² and 46.37 km² and the Gravelius indices of the two basins are respectively 4.89 and 5.51. Each year, the risk of a flood occurring is 75% with an average of more than 500 million property damage. The level reached in flooded areas of about 1 meter and according to historical data about inundations in Abidjan, on average 16 people lose their lives each year.

How to cite: Aka, K., Corzo P, G. A., and Yao, K.: Dynamic assessment of flood inundation based on a spatiotemporal hydrological model feed by a dynamic representation of human activities , EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-11928, https://doi.org/10.5194/egusphere-egu2020-11928, 2020.

Statistical analysis of rainfall and runoff extremes plays a crucial role in hydrological design and flood risk management. Usually this analysis is performed separately for the two processes of interest, thus ignoring their dependencies, which appear at multiple temporal scales. Actually, the generation of a flood strongly depends on soil moisture conditions, which in turn depends on past rainfall. Using daily rainfall and runoff data from about 400 catchments in USA, retrieved from the MOPEX repository, we investigate the statistical behavior of the corresponding annual rainfall and streamflow maxima, also accounting for the influence of antecedent soil moisture conditions. The latter are quantified by means of accumulated daily rainfall at various aggregation scales (i.e., from 5 up to 30 days) before each extreme rainfall and streamflow event. Analysis of maxima is employed by fitting the Generalized Extreme Value (GEV) distribution, using the L-moments method for extracting the associated parameters (shape, scale, location). Significant attention is paid for ensuring statistically consistent estimations of the shape parameter, which is empirically adjusted in order to minimize the influence of sample uncertainty. Finally, we seek for the possible correlations among the derived parameter values and hydroclimatic characteristics of the studied basins, and also depict their spatial distribution across USA.

How to cite: Nezi, M., Tsoukalas, I., Ntigkakis, C., and Efstratiadis, A.: Multidimensional context for extreme analysis of daily streamflow, rainfall and accumulated rainfall across USA, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-19674, https://doi.org/10.5194/egusphere-egu2020-19674, 2020.

In the field of extreme hydrological events, design storm identification represents key element due to the links with flood risk as well as water resources availability and management.

In order to obtain a regional frequency analysis for studying and understanding the annual maximum of daily rainfall, two different statistic methods are proposed here on Tuscany Region (Central Italy). The first method concerns with the hierarchical approach on three levels: the studied area is divided into homogeneous regions and sub-regions, then the statistical homogeneity within the regions is verified through several homogeneity tests. Furthermore, the Two-Component Extreme Value (TCEV) probability distribution of the extreme rainfall is considered identical within each homogeneous region unless a scale factor, i.e. the index rainfall, estimated through a multivariate model based on climatic and geomorphological characteristics.

A Generalized Additive Model (GAM) for extremes is also implemented on the studied area assuming that the observations follow a Generalized Extreme Value - GEV distribution whose locations are spatially dependent. The research has been carried out starting from a general set of 922 rain gauges (Regional Hydrological Service of Tuscany – SIR), on time series of annual maximum of daily rainfall recorded from 1916 to 2017. The application of the two methods is discussed based on the comparison between the maps of the design storm for daily duration and 2, 20, 50, 100 e 200 years return periods.

How to cite: Pampaloni, M., Vannacci, V., Caporali, E., Bocci, C., Chiarello, V., and Petrucci, A.: Design storm estimation in Tuscany (Italy) through regional frequency analysis and generalized additive modelling , EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-18598, https://doi.org/10.5194/egusphere-egu2020-18598, 2020.

Urbanization causes extension of impervious surface interrupting natural hydrological cycle, which may increase in the number of disaster factors causing difficulties in terms of flood management. Flood control measures should prioritize identification of areas where flooding is expected to occur, considering various spatial characteristics distributed over the areas at risk. In this study, a probabilistic flood risk assessment was performed. The flood hazard map for a 100-year return level was used to illustrate the concept of a probabilistic model. Here, we trained the model to obtain the relationship between the estimated inundation area and potential predictors such as elevation, slope, curve number, and distance to the river. In this study, a Bayesian logistic regression analysis was performed to impose probabilities on the inundation for each grid. Finally, the flood risk was provided with the population for the entire target area through the model.

Keywords: Bayesian Inference, Flood Hazard Map, Geographical Information, Logistic Regression

Acknowledgement

This work is supported by the Korea Agency for Infrastructure Technology Advancement(KAIA) grant funded by the Ministry of Land, Infrastructure and Transport (Grant 19AWMP-B121100-04)

How to cite: Yu, J.-U., Jung, M., Kim, J.-Y., and Kwon, H.-H.: Proababilistic Approach to Deterministic Inundation Map Informed by Geographical Factors, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-21498, https://doi.org/10.5194/egusphere-egu2020-21498, 2020.

Title:

Evaluation of the spatio-temporal development of hydrological droughts and its sensitivity to the choice of different parameters of the hydrological model. Case study: Magdalena-Cauca River basin – Colombia

Abstract:

Droughts in Colombia have been studied using local and regional indicators; however, the nature of events and the heterogeneity of mountains, and the high variability of climate and hydrological process, indicate that this should follow a more dynamic spatiotemporal analysis. In previous studies about drought, it has been possible to identify how natural drought phenomena tend to spread irregularly through large regions. This research aims to develop a spatiotemporal evaluation of hydrological droughts in Colombia. The process of the analysis followed three main components, one estimating the drought indicator to the interpolated data set from the local agency IDEAM. This step aims to find the optimal combination of parameters sets to characterize the hydrological behavior; to determine standardized runoff, soil moisture and evaporation deficit indices (SRI, SSMI, SEDI respectively). Second, the determination of the patterns using the Contiguous Drought Area (CDA) and Non-Contiguous Drought Area (NCDA) methodologies to characterise the spatio-temporal behaviour of the hydrological droughts. And before concluding an assessment of the robustness of the drought events, a threshold sensitivity analysis was performed. The body of the study includes a complete conceptual framework with the definition of hydrological droughts and drought indices (DI). This methodology is based on the characterization spatiotemporal droughts that examines the patterns of events using results from previous studies. The results of this analysis are key for the preparedness of the region to extreme events.

How to cite: Corzo P, G. A., Aguilar Carrillo, F., Cotrino Palma, J. M., and Santos Granados, G. R.: Evaluation of the spatio-temporal development of hydrological droughts and its sensitivity to the choice of different parameters of the hydrological model. Case study: Magdalena-Cauca River basin – Colombia, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-22037, https://doi.org/10.5194/egusphere-egu2020-22037, 2020.