Fig 1 - uploaded by Korbinian Breinl
Content may be subject to copyright.
(a) Topography of Austria with river network and glaciated areas. (b) Five rainfall regions representing different dominating rainfall mechanisms, rain gauge network and catchment centroids of gauged catchments. Catchments impacted by karst are marked with crosses, those by glaciers with asterisks. (c) Geology of Austria.
Source publication
The aim of this paper is to explore how rainfall mechanisms and catchment characteristics shape the relationship between rainfall and flood probabilities. We propose a new approach of comparing intensity-duration-frequency statistics of maximum annual rainfall with those of maximum annual streamflow in order to infer the catchment behavior for runo...
Contexts in source publication
Context 1
... & Yang, 2011). For extreme rainfall and floods, there seems to similarly exist a stronger sensitivity in dry catchments. For example, the data of Paquet et al. (2013) suggest that a 1% increase of an extreme 72hr rainfall (300 mm) produced a 2.6%, 2.0% and 1.5% increase in extreme runoff for dry, intermediate and wet conditions (estimated from Fig. 16 of Paquet et al. (2013)). However, to the best of our knowledge we are not aware of data-based regional analyses that have explored flood sensitivities to extreme rainfall. Most of the above studies have analyzed the rainfall-to-flood probability transformation by a model-based derived flood frequency concept (e.g. Sivapalan et al., ...
Context 2
... study is set in Austria (84,000 km 2 ). In the North, East and South East elevations are below 200 m above sea level, while the highest Alpine summits reach over 3,500 m (Fig. 1a). Mean annual rainfall ranges from less than 400 mm yr 1 in the East to more than 3000 mm yr 1 in the ...
Context 3
... data from stream gauges in Austria for the period 1976-2015 (June to August), with a mean length of 36 years (maximum of 40 years, minimum of 21 years). Time series of catchments strongly influenced by hydraulic infrastructure were excluded, which resulted in 428 catchments with areas ranging from 3.9 km 2 to 4,792 km 2 and a median of 125 km 2 (Fig. 1b). We also used hourly rainfall observations from 314 Austrian rain gauges for the period 1950-2016 (June to August) with a mean length of 20 years (maximum of 43 years, minimum of 10 years) (Fig. 1b). Catchments impacted by Karst springs were identified based on a visual comparison with hydrogeological ...
Context 4
... by hydraulic infrastructure were excluded, which resulted in 428 catchments with areas ranging from 3.9 km 2 to 4,792 km 2 and a median of 125 km 2 (Fig. 1b). We also used hourly rainfall observations from 314 Austrian rain gauges for the period 1950-2016 (June to August) with a mean length of 20 years (maximum of 43 years, minimum of 10 years) (Fig. 1b). Catchments impacted by Karst springs were identified based on a visual comparison with hydrogeological ...
Context 5
... and flood mechanism vary considerably within the study region. In general, convective rainfall and flash floods are most frequent in the lowlands, identifiable by a higher frequency of hailstorms, severe wind gusts and tornados in these regions (Fig. A1, Dotzek et al. (2009) and Merz and Blöschl (2003)), while short-rain and long-rain floods from orographic rainfall are most frequent along the Alpine ridge (Merz & Blöschl, 2003), aligned with longer wet spells and shorter dry spells along the Alpine ridge compared to the lowlands (Fig. A2). Throughout Austria, floods are more frequent ...
Context 6
... investigate the regional differences, rain gauges and catchments were grouped into five regions based on the previous rainfall-based classifications by Matulla et al. (2003), Seibert et al. (2007) and Breinl et al. (2020) (Fig. 1b). We chose a rainfall-based classification in order to better understand regionally dominant factors in runoff formation. (i) The "Western orographic" region is dominated by orographic rainfall due to airflows from Western, NNW and NW directions. Long-rain floods are most frequent and the highest catchments are affected by glacier melt ...
Context 7
... response is also influenced by geology (Fig. 1c). The Flysch zone along the Northern fringe of the Alps is characterized by low subsurface permeability and thus surface or near surface flow paths, leading to flashy response. Similarly, shallow soils and efficient drainage networks in the Southeast lead to short response times ( Gaál et al., 2012). On the other hand, in the Phyllite ...
Context 8
... model parameters of the rainfall model λ P (scale parameter representing rainfall variability), ψ P (location parameter representing rainfall magnitude, ceteris paribus) and η P (scaling parameter representing the dependence on duration). Point sizes relate to catchment centroid elevations and colors indicate rainfall regions in Austria (Fig. 1b). Fig. 5. Cumulative distribution functions of rainfall i P (upper row) and specific streamflow i q (bottom row) extremes with a duration of 1hr in the five rainfall regions. Colored lines represent the median CDFs in each region, numbers in the plot area refer to the 5% and 99% quantiles (italic) of the median. Dashed and dotted grey ...
Context 9
... specific streamflow i q (bottom row) extremes with a duration of 1hr in the five rainfall regions. Colored lines represent the median CDFs in each region, numbers in the plot area refer to the 5% and 99% quantiles (italic) of the median. Dashed and dotted grey lines represent catchments impacted by Karst springs and glaciers, respectively (see Fig. ...
Context 10
... specific streamflow i q (bottom row) extremes with a duration of 24hrs in the five rainfall regions. Colored lines represent the median CDFs in each region, numbers in the plot area refer to the 5% and 99% quantiles (italic) of the median. Dashed and dotted grey lines represent catchments impacted by Karst springs and glaciers, respectively (see Fig. 1b). Fig. 7. Distribution of rainfall and streamflow model parameters in five rainfall regions in Austria (Fig. 1b). Left boxplot in each region refers to the rainfall model parameters, right boxplot to the streamflow model parameters. (a) Scale parameters λ P and λ q of the rainfall and streamflow model, respectively. (b) location ...
Context 11
... represent the median CDFs in each region, numbers in the plot area refer to the 5% and 99% quantiles (italic) of the median. Dashed and dotted grey lines represent catchments impacted by Karst springs and glaciers, respectively (see Fig. 1b). Fig. 7. Distribution of rainfall and streamflow model parameters in five rainfall regions in Austria (Fig. 1b). Left boxplot in each region refers to the rainfall model parameters, right boxplot to the streamflow model parameters. (a) Scale parameters λ P and λ q of the rainfall and streamflow model, respectively. (b) location parameters ψ P and ψ q . (c) scaling parameters η P and η q ...
Context 12
... regions. The CV of streamflow is likewise higher in the dry lowlands and decreases with increasing elevation (and thus increasing annual rainfall), possibly due to the smaller variability of runoff coefficients (Merz & Blöschl, 2009b). The lowest CV of streamflow occurs in the glaciated areas as the values of ψ q are the highest (also see Fig. 1a). Thus, as opposed to the scale parameters (see above), the CVs of rainfall and floods are spatially ...
Context 13
... q to changes in extreme rainfall i P assuming duration d = const. = 1hr in Austria. Elasticity is the change in percentage of i q when i P is changed by 1% and is shown as points for each gauged catchment where the sizes represent the values of the elasticities and the colors the rainfall regions (Fig. 1b). Elasticities interpolated from these catchment values (ordinary kriging) are shown as background pattern. (a) ε 1 d = 1hr at T = 2yrs, (b) ε 1 d = 1hr at T = ...
Context 14
... elasticity ε 1 of streamflow relative to changes of the rainfall assuming d = const. (Equation (8)) is characterized by a distinct spatial pattern ( Fig. 9 and Fig. 10). The elasticities ε 1 are shown in Fig. 9 as boxplots and in Fig. 10 as maps. The sizes of the points in Fig. 10 represent the values of ε 1 for each gauged catchment (at the catchment centroids), the background pattern is generated by interpolation from these catchment values (ordinary kriging) and is intended to better highlight the ...
Context 15
... elasticity ε 1 of streamflow relative to changes of the rainfall assuming d = const. (Equation (8)) is characterized by a distinct spatial pattern ( Fig. 9 and Fig. 10). The elasticities ε 1 are shown in Fig. 9 as boxplots and in Fig. 10 as maps. The sizes of the points in Fig. 10 represent the values of ε 1 for each gauged catchment (at the catchment centroids), the background pattern is generated by interpolation from these catchment values (ordinary kriging) and is intended to better highlight the regional patterns. For T = 2yrs, the rainfall-streamflow ...
Context 16
... elasticity ε 1 of streamflow relative to changes of the rainfall assuming d = const. (Equation (8)) is characterized by a distinct spatial pattern ( Fig. 9 and Fig. 10). The elasticities ε 1 are shown in Fig. 9 as boxplots and in Fig. 10 as maps. The sizes of the points in Fig. 10 represent the values of ε 1 for each gauged catchment (at the catchment centroids), the background pattern is generated by interpolation from these catchment values (ordinary kriging) and is intended to better highlight the regional patterns. For T = 2yrs, the rainfall-streamflow relationships are highly elastic (i.e. ε 1 ≫1) in the ...
Context 17
... these catchment values (ordinary kriging) and is intended to better highlight the regional patterns. For T = 2yrs, the rainfall-streamflow relationships are highly elastic (i.e. ε 1 ≫1) in the lowlands in the North, Northeast and Southeast (e.g. Northeastern convective rainfall region) compared to the high Alpine regions (e.g. Western orographic Fig. 11. Elasticity ε 2 assuming T = const. of streamflow to rainfall extremes of all catchments in the five rainfall regions in Austria (Fig. 1b). Elasticity is the change in percentage of i q when i P is increased by 1%. Boxes represent the 25th and 75th percentiles, lines within the median. Fig. 12. Spatial patterns of the elasticity ε 2 of ...
Context 18
... relationships are highly elastic (i.e. ε 1 ≫1) in the lowlands in the North, Northeast and Southeast (e.g. Northeastern convective rainfall region) compared to the high Alpine regions (e.g. Western orographic Fig. 11. Elasticity ε 2 assuming T = const. of streamflow to rainfall extremes of all catchments in the five rainfall regions in Austria (Fig. 1b). Elasticity is the change in percentage of i q when i P is increased by 1%. Boxes represent the 25th and 75th percentiles, lines within the median. Fig. 12. Spatial patterns of the elasticity ε 2 of all catchments assuming T = const. for the five rainfall regions in Austria (Fig. 1b). Elasticity is the change in percentage of i q when ...
Context 19
... to the high Alpine regions (e.g. Western orographic Fig. 11. Elasticity ε 2 assuming T = const. of streamflow to rainfall extremes of all catchments in the five rainfall regions in Austria (Fig. 1b). Elasticity is the change in percentage of i q when i P is increased by 1%. Boxes represent the 25th and 75th percentiles, lines within the median. Fig. 12. Spatial patterns of the elasticity ε 2 of all catchments assuming T = const. for the five rainfall regions in Austria (Fig. 1b). Elasticity is the change in percentage of i q when i P is increased by 1% and is shown as points for each gauged catchment where the sizes represent the values of the elasticities and the colors the rainfall ...
Context 20
... of all catchments in the five rainfall regions in Austria (Fig. 1b). Elasticity is the change in percentage of i q when i P is increased by 1%. Boxes represent the 25th and 75th percentiles, lines within the median. Fig. 12. Spatial patterns of the elasticity ε 2 of all catchments assuming T = const. for the five rainfall regions in Austria (Fig. 1b). Elasticity is the change in percentage of i q when i P is increased by 1% and is shown as points for each gauged catchment where the sizes represent the values of the elasticities and the colors the rainfall regions (Fig. 1b). Elasticities interpolated from these catchment values (ordinary kriging) are shown as background pattern. ...
Context 21
... 12. Spatial patterns of the elasticity ε 2 of all catchments assuming T = const. for the five rainfall regions in Austria (Fig. 1b). Elasticity is the change in percentage of i q when i P is increased by 1% and is shown as points for each gauged catchment where the sizes represent the values of the elasticities and the colors the rainfall regions (Fig. 1b). Elasticities interpolated from these catchment values (ordinary kriging) are shown as background pattern. region) where the elasticity is closer to unity i.e. ε 1 ∼ 1 (left boxplots of Fig. 9 and Fig. 10a). In the glaciated regions of the highest summits of Austria (regions in red colors in Fig. 10, also see glaciated areas in Fig. ...
Context 22
... by 1% and is shown as points for each gauged catchment where the sizes represent the values of the elasticities and the colors the rainfall regions (Fig. 1b). Elasticities interpolated from these catchment values (ordinary kriging) are shown as background pattern. region) where the elasticity is closer to unity i.e. ε 1 ∼ 1 (left boxplots of Fig. 9 and Fig. 10a). In the glaciated regions of the highest summits of Austria (regions in red colors in Fig. 10, also see glaciated areas in Fig. 1a) the relationships are inelastic (ε 1 < 1). For a return period of T = 100yrs (right boxplots in Fig. 9 and Fig. 10b), the spatial pattern remains similar, but the elasticities tend towards ...
Context 23
... and the colors the rainfall regions (Fig. 1b). Elasticities interpolated from these catchment values (ordinary kriging) are shown as background pattern. region) where the elasticity is closer to unity i.e. ε 1 ∼ 1 (left boxplots of Fig. 9 and Fig. 10a). In the glaciated regions of the highest summits of Austria (regions in red colors in Fig. 10, also see glaciated areas in Fig. 1a) the relationships are inelastic (ε 1 < 1). For a return period of T = 100yrs (right boxplots in Fig. 9 and Fig. 10b), the spatial pattern remains similar, but the elasticities tend towards ...
Context 24
... regions (Fig. 1b). Elasticities interpolated from these catchment values (ordinary kriging) are shown as background pattern. region) where the elasticity is closer to unity i.e. ε 1 ∼ 1 (left boxplots of Fig. 9 and Fig. 10a). In the glaciated regions of the highest summits of Austria (regions in red colors in Fig. 10, also see glaciated areas in Fig. 1a) the relationships are inelastic (ε 1 < 1). For a return period of T = 100yrs (right boxplots in Fig. 9 and Fig. 10b), the spatial pattern remains similar, but the elasticities tend towards ...
Context 25
... pattern. region) where the elasticity is closer to unity i.e. ε 1 ∼ 1 (left boxplots of Fig. 9 and Fig. 10a). In the glaciated regions of the highest summits of Austria (regions in red colors in Fig. 10, also see glaciated areas in Fig. 1a) the relationships are inelastic (ε 1 < 1). For a return period of T = 100yrs (right boxplots in Fig. 9 and Fig. 10b), the spatial pattern remains similar, but the elasticities tend towards ...
Context 26
... elasticity ε 2 of streamflow relative to rainfall changes assuming T = const. (Equation (11)) is highest in the dry lowlands in the Northeast and East (Northeastern convective and Eastern mixed rainfall region) ( Fig. 11 and Fig. 12). As in the case of Fig. 10, the sizes of the points in Fig. 12 represent the values of ε 2 for each gauged catchment (at the catchment centroids), while the background pattern is generated by interpolation from these catchment values (ordinary kriging). In these regions catchment concentration times tend to be short as indicated by ...
Context 27
... elasticity ε 2 of streamflow relative to rainfall changes assuming T = const. (Equation (11)) is highest in the dry lowlands in the Northeast and East (Northeastern convective and Eastern mixed rainfall region) ( Fig. 11 and Fig. 12). As in the case of Fig. 10, the sizes of the points in Fig. 12 represent the values of ε 2 for each gauged catchment (at the catchment centroids), while the background pattern is generated by interpolation from these catchment values (ordinary kriging). In these regions catchment concentration times tend to be short as indicated by high η q (Fig. 7c), and storm ...
Context 28
... elasticity ε 2 of streamflow relative to rainfall changes assuming T = const. (Equation (11)) is highest in the dry lowlands in the Northeast and East (Northeastern convective and Eastern mixed rainfall region) ( Fig. 11 and Fig. 12). As in the case of Fig. 10, the sizes of the points in Fig. 12 represent the values of ε 2 for each gauged catchment (at the catchment centroids), while the background pattern is generated by interpolation from these catchment values (ordinary kriging). In these regions catchment concentration times tend to be short as indicated by high η q (Fig. 7c), and storm durations tend to be short indicated ...
Context 29
... pronounced similarities of the spatial patterns of ε 1 (Fig. 10) and the location parameter of the streamflow ψ q (Fig. A3d), as well as the spatial patterns of ε 2 (Fig. 12) and the scaling parameter of the streamflow η q (Fig. A3f) suggest that catchment processes dominate the runoff transformation, since ε 1 is defined by the location parameters (Equation (9)) and ε 2 by the scaling parameters ...
Context 30
... pronounced similarities of the spatial patterns of ε 1 (Fig. 10) and the location parameter of the streamflow ψ q (Fig. A3d), as well as the spatial patterns of ε 2 (Fig. 12) and the scaling parameter of the streamflow η q (Fig. A3f) suggest that catchment processes dominate the runoff transformation, since ε 1 is defined by the location parameters (Equation (9)) and ε 2 by the scaling parameters (Equation (11)). The dominance of catchment processes is also visible in the cumulative distribution functions ...
Context 31
... 8) and is thus highest along the Alpine ridge (Fig. A3b), while the variability of the rainfall represented by λ P is mainly controlled by elevation (see negative correlation between λ P and elevation in Fig. 8) and is highest in the lowlands (Fig. A3a). Some of the highest values of λ q relate to karstic catchments along the Alpine divide (see Fig. 1a catchments with crosses, Fig. A3b and positive correlation between λ q and Carbonate rock in Fig. 8). In karstic catchments, during periods of average rainfall events, most of the rainfall may be stored in the fractured carbonic rocks, while more extreme rainfall events can saturate the epikarst zone inducing large streamflow extremes ...
Context 32
... to an interplay of climatic and catchment processes over time, modulated by the geology ( Gaal et al., 2012). For example, the efficient drainage network in the Southeast of Austria leads to short response times and may have evolved from the dominating convective rainfall mechanisms in the region (see for example high frequency of hailstorms in Fig. A1) as intense convective rainstorms increase the overland flow, which in return influences the drainage network (Abrahams & Ponc- zynski, 1984;Tucker & Bras, ...
Context 33
... ε 1 are highest in the dry lowlands in the Northeast and Southeast ( Fig. 9 and Fig. 10). For T = 2yrs, they are more than four times higher than the mean of the Alpine catchments. That is, in the dry lowlands, flood frequency curves are considerably steeper than the rainfall frequency curves, which is consistent with the regional analyses (Merz & Blöschl, 2009a, 2009b). suggested that in dry regions, steep flood ...
Context 34
... the mountainous catchments of the study region, where annual rainfall is above 1000 mm yr 1 , elasticities ε 1 are closer to unity ( Fig. 9 and Fig. 10). In these catchments the high orographic rainfall may frequently lead to soil saturation and thus event runoff coefficients tend to be high (Merz & Blöschl, 2009b), and therefore the steepness of the flood frequency curves is similar to that of the IDF curves. Another possible explanation is the slower catchment response in parts of ...
Context 35
... the steepness of the flood frequency curves is similar to that of the IDF curves. Another possible explanation is the slower catchment response in parts of the high rainfall regions ( Gaal et al., 2012), as a result of a more pervious geology, for example in the Southern mixed rainfall zone where the Phylitte geology reduces the response time (Fig. 1c) and thus may reduce elasticity as illustrated in the simulation experiment (Fig. 3). This is because, for slow response times, the highest rainfall extreme of year and a given duration does not necessarily cause the highest flood peak in that year. The lower correspondence of rainfall and flood events is also in line with their lower ...
Context 36
... 3). This is because, for slow response times, the highest rainfall extreme of year and a given duration does not necessarily cause the highest flood peak in that year. The lower correspondence of rainfall and flood events is also in line with their lower synchronicity (Fig. A4). The decoupling is even more pronounced in the glaciated catchments ( Fig. 10 red areas) where elasticities drop below unity, as a consequence of glacier melt that is more relevant to floods than rainfall ...
Context 37
... assuming the return period T is constant. In the model ε 2 is represented by η q η P (Equation (11)), i.e. the ratio of the temporal scaling of streamflow and extreme rainfall. Regions of high elasticities ε 2 are regions of fast catchment response as indicated by ( Gaal et al., 2012), which is aligned with values of η q and to some degree η P (Fig. 11, Fig. 12, and Fig. A3f). Fast catchment response is mainly related to shallow soils, and clay and marl and Flysch geologies with low infiltration (Fig. 1c), as well as an efficient drainage network ( Gaal et al., 2012;Holko et al., 2011). In these regions, also relevant storm durations tend to be short due to the dominance of convective rainfall, which in ...
Context 38
... and extreme rainfall. Regions of high elasticities ε 2 are regions of fast catchment response as indicated by ( Gaal et al., 2012), which is aligned with values of η q and to some degree η P (Fig. 11, Fig. 12, and Fig. A3f). Fast catchment response is mainly related to shallow soils, and clay and marl and Flysch geologies with low infiltration (Fig. 1c), as well as an efficient drainage network ( Gaal et al., 2012;Holko et al., 2011). In these regions, also relevant storm durations tend to be short due to the dominance of convective rainfall, which in combination with the fast catchment response explains the high elasticities ε 2 . Regions of low elasticities ε 2 on the other hand are ...
Context 39
... ε 1 tend towards unit elasticity with increasing return period, as can be seen from Equation (9), Fig. 9 and Fig. 10. This behavior is a consequence of the model formulation and fully consistent with hydrological reasoning. For both, infiltration excess and saturation excess mechanisms, runoff tends to become more similar to rainfall as the rainfall intensity increases, and if soil saturation is reached, any additional rainfall transforms into ...
Context 40
... authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. (Fig. 1b). (Fig. 1b). The rainfall parameters include λ P (scale parameter representing rainfall variability), ψ P (location parameter representing the magnitude, ceteris paribus) and η P (scaling parameter indicating convective activity). The streamflow model parameters include λ q (scale parameter representing streamflow variability), ψ q ...
Context 41
... authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. (Fig. 1b). (Fig. 1b). The rainfall parameters include λ P (scale parameter representing rainfall variability), ψ P (location parameter representing the magnitude, ceteris paribus) and η P (scaling parameter indicating convective activity). The streamflow model parameters include λ q (scale parameter representing streamflow variability), ψ q (location ...
Context 42
... parameter representing the magnitude, ceteris paribus) and η P (scaling parameter indicating convective activity). The streamflow model parameters include λ q (scale parameter representing streamflow variability), ψ q (location parameter representing the magnitude, ceteris paribus) and η q (scaling parameter representing catchment response). (Fig. 1b). Boxes represent the 25th and 75th percentiles, lines within the ...
Similar publications
In this study, a new approach for rainfall spatial interpolation in the Luxembourgian case study is introduced. The method used here is based on a Fuzzy C-Means (FCM) clustering method. In a typical FCM procedure, there are a lot of available data and each data point belongs to a cluster, with a membership degree [0 1]. On the other hand, in our me...
Citations
... However, apart from rainfall in the area, there are several factors that influence the occurrence of inundation and catastrophic flooding. These factors can be climate change, topography, soil type, and the geology of the area [16]. ...
Floods are hydrometeorological disasters that often occur in various regions, including in Nganjuk Regency. From 2019 to 2022, there were urban floods that inundated highways, markets and hospitals and disrupted community and government activities. Heavy rain is one of the causes of flooding. Rain intensity, duration and frequency (IDF) curves were constructed using rainfall data from six rain gauge stations. This curve is used to determine the 10-year and 25-year probabilities of PUH intensity and the probability of causing flooding in Nganjuk District. Rain intensity in Nganjuk Regency is PUH 10 and 25 with an average rain duration of 15-60 minutes which is included in the moderate rain category, namely 20-50 mm/day. The results of this analysis will become the basis for consideration in formulating policies and strategies for urban flood disaster prevention and management in Nganjuk Regency. Therefore, coordination between the city government and the government is needed to reduce the risk of urban flooding in Nganjuk Regency.
... Precipitation events are occurring at higher frequencies and intensities as a result of climate change. Because precipitation is the main driver of floods, an increase in precipitation results in extensive flooding (Breinl et al., 2021;G. Sofia & Nikolopoulos, 2020). ...
In many regions of the world, including the United States, human and animal fecal genetic markers have been found in flood waters. In this study, we use high‐resolution whole genomic sequencing to examine the origin and distribution of Salmonella enterica after the 2018 Hurricane Florence flooding. We specifically asked whether S. enterica isolated from water samples collected near swine farms in North Carolina shortly after Hurricane Florence had evidence of swine origin. To investigate this, we isolated and fully sequenced 18 independent S. enterica strains from 10 locations (five flooded and five unflooded). We found that all strains have extremely similar chromosomes with only five single nucleotide polymorphisms (SNPs) and possessed two plasmids assigned bioinformatically to the incompatibility groups IncFIB and IncFII. The chromosomal core genome and the IncFIB plasmid are most closely related to environmental Salmonella strains isolated previously from the southeastern US. In contrast, the IncFII plasmid was found in environmental S. enterica strains whose genomes were more divergent, suggesting the IncFII plasmid is more promiscuous than the IncFIB type. We identified 65 antibiotic resistance genes (ARGs) in each of our 18 S. enterica isolates. All ARGs were located on the Salmonella chromosome, similar to other previously characterized environmental isolates. All isolates with different SNPs were resistant to a panel of commonly used antibiotics. These results highlight the importance of environmental sources of antibiotic‐resistant S. enterica after extreme flood events.
... Flood risk analysis has been previously conducted using precipitation-based indicators, especially when observed data is challenging to obtain (Guerrerio & Lajinha, 2007;Breinl et al., 2021). In this context, the patterns and trends derived from such analysis are of great value to decisionmakers for efficient flood risk management and the development of preventive strategies to mitigate the impacts of floods (Olanrewaju & Reddy, 2022). ...
... precisely 1 or 24 hours. This approach is used in, for example, Breinl et al. (2021) and Barna et al. (2023). For each station, even spacing in the streamflow time series was enforced via regular sampling of a linear interpolation of the observed data. ...
Estimation of flood quantiles at ungauged basins is often achieved through regression based methods. In situations where flood retention is important, e.g. floodplain management and reservoir design, flood quantile estimates are often needed at multiple durations. This poses a problem for regression-based models as the form of the functional relationship between catchment descriptors and the response may not be constant across different durations. A particular type of regression model that is well-suited to this situation is a generalized additive model (GAM), which allows for flexible, semi-parametric modeling and visualization of the relationship between predictors and the response. However, in practice, selecting predictors for such a flexible model can be challenging, particularly given the characteristics of available catchment descriptor datasets. We employ a machine learning-based variable pre-selection tool which, when combined with domain knowledge, enhances the practicality of constructing GAMs. In this study, we develop a GAM for index (median) flood estimation with the primary objective of investigating duration-specific differences in how catchment descriptors influence the median flood. As the accuracy of this explainable approach is dependent on the fitted GAM being adequate, the secondary objective of our study is prediction of the median flood at ungauged locations and multiple durations, where predictive performance and reliability at ungauged locations are used as proxies for adequacy of the GAM. Predictive performance of the GAM is compared to two benchmark models: the existing log-linear model for median flood estimation in Norway and a fully data-driven machine learning model (an extreme gradient boosting tree ensemble, XGBoost). We find that the predictive accuracy and reliability of the GAM matched or exceeded that of the benchmark models at both durations studied. Within the predictor set selected for this study, we observe duration-specific differences in the relationship between the median flood and the two catchment descriptors effective lake percentage and catchment shape. Ignoring these differences results in a statistically significant decline in predictive performance. This suggests that models developed and estimated for prediction of the index flood at one duration may have reduced performance when applied directly to situations outside of that specific duration.
... Elevation, slope, precipitation, and flow accumulation were numerical factors while LULC and soil maps are descriptive factors. According to Breinl et al. (2021), the relationship between rainfall and flood probabilities is shaped by precipitation mechanisms and watershed features. The precipitation data used in the study is a historical monthly 2000 climate data, which has a 30 s (~ 1 km 2 ) spatial resolution, and it was resampled to 30 m to match with spatial resolution of the other factors. ...
Climate change, urbanization, and land subsidence are increasing risks of dangerous and costly flooding in Africa’s coastal regions. As the intensity and the number of flooding occurrences increase, real-time solutions for flood risks and hazards mapping are required to assess flooding and its potential impacts on the environment and humans. Flood risk mapping has become an important tool for assessing flood risks in coastal areas to guide planning and mitigation efforts by local officials. However, this tool has not been leveraged to support flood risk analysis. Using GIS weighted sum method, this study mapped and assessed flood risk zones in Limbe, Cameroon, by utilizing a Shuttle Radar Topography Mission-Digital Elevation Model (SRTM-DEM). Flood risk zones were delineated based on several environmental factors, including precipitation, elevation, slope, soil texture, land use/land cover, and flow accumulation. The findings classified about 50% of land in Limbe as having either high or very high vulnerability to flooding. The highest risk areas were concentrated in the upstream and downstream areas of Ejengele and Limbe Rivers. Only about 9% of Limbe was classified as very low risk, suggesting that most of the land area is susceptible to flooding. The study reveals that more than 90% of the Limbe’s urban areas are in flood-prone areas, which requires special attention from stakeholders to take necessary steps toward the preparation and mitigation of extreme flood events. The resulting flood risk maps provide vital information for urban and risk management planning professionals.
... Rainfall depths tend to be largest in the north-west of the region, mainly as a result of orographic effects. At the northern fringe of the Alps in the north-west of Austria, rainfall associated with a return period of 100 years is of the order of 150 mm (Breinl et al., 2021). In the basins at the northern fringe of the high Alps, long duration rainfall driven floods are particularly common. ...
This study compares the flood regime of rivers in Ukraine and Austria over the last decades. We used data from mountain and lowland watersheds, where floods are caused by different processes. In order to identify possible shifts in the day of occurrence of annual flood maxima, we apply the kernel density method to the time series of two subperiods (1960-1987 and 1988-2015). We use the Mann Kendall test at a 5% significance level to identify significant positive or negative trends in the series of annual maximum discharges. In Austria, we observe an increasing trend in summer floods associated with increasing precipitation. In the lowland areas of Ukraine, a clear reduction in spring floods is observed, linked to shallower snow packs in a warming climate. In the Ukrainian Carpathians, on the other hand, where floods occur throughout the year, an increase in the portion of liquid precipitation during the cold period of the year leads to earlier floods and an increase in the probability of flooding in winter.
... For instance, the 1989 floods in SW Mallorca reached peaks over 700 m 3 /s, with a highest value of 1054 m 3 /s in the area of Campos [60]. Again, these values have to be considered in the knowledge that rainfall intensities and its spatial distribution over catchments play an important role in the flood generation processes and such information is not available for the vast majority of Mallorca's events [61,62]. ...
The research presented herein studies three episodes of flooding that affected the ephemeral basin of the Sant Jordi stream in northwestern Mallorca. These events are considered common since they do not reach the proportions in terms of the flow rates of other cases that have occurred in Mallorca, but they are nevertheless important due to the impact they have on human activity and also due to the morphological changes caused in the basin itself. On the one hand, the development of the field work to characterize and calculate the peak flows is presented, and on the other hand, the geomorphic changes caused by the water and the materials carried away are explained. The results allow us to identify a type of Mediterranean flood, which happens on a regular basis, but which does not stand out for its flows or for its major socioeconomic impacts but still has an effect on the natural and anthropic environment. This information can be valuable for local and regional authorities as well as for the public to avoid risk situations and prevent impacts on public and private property caused by future events.
... The detection of rainfall plays a vital role in preventing floods, as it enables the identification of suitable thresholds that are liable to cause flood damage [1], understanding of the relationship between rainfall and flood probabilities [2], identify spatiotemporal and fluvial-pluvial sources of flooding [3], and evaluate the impact of climate change on flood and extreme precipitation events [4]. Additionally, rapid onset flooding, commonly known as flash floods, can rise within a brief duration of time, varying from a few minutes to a few hours, triggered by intense rainfall, a sudden release of water, or a failure of a dam or levee [5]. ...
Rainfall is crucial for flood prevention and comprehending the correlation between rainfall and flooding. Cavite province in the Philippines is vulnerable to flooding caused by heavy rainfall and climate change impacts. Early detection of flooding through early warning systems can prevent excessive damage loss and potentially save lives. It can also provide major savings in terms of monetary benefit and increased interagency coordination for rapid decision-making. Machine learning is an important tool for predicting rainfall which can be used to predict rainfall in the province. The objective of this study is to conduct a comparative analysis of various models for predicting daily rainfall, using relevant atmospheric features such as maximum, minimum, and mean temperature, relative humidity, wind speed, wind direction, cloud cover, pressure, and evaporation. The study seeks to identify the most effective model for accurately predicting rainfall in the Cavite Province to benefit the local community. Among the five machine learning models evaluated, the Gaussian Process Regression model demonstrated the highest accuracy in predicting daily rainfall. The findings of this study can be leveraged to mitigate the damage caused by flooding in the Cavite Province and serve as a useful reference for similar studies in other regions prone to flooding.
... Quantifying severe weather events is of particular interest to actuaries, since events such as flooding account for a large part of global economic losses (Boudreault et al., 2020). An increase in extreme rainfall can lead to a possibly greater increase in river discharge (Breinl et al., 2021). Therefore, one would gain from obtaining reliable rainfall projections to assess flood risks. ...
... Through a combination of hydrological and hydraulic models such as Hydrotel (Fortin et al., 2001), HEC-RAS (Brunner, 2016) or the Hillslope Link Model (Demir & Krajewski, 2013), one can produce discharge flood projections under different rainfall scenarios. Breinl et al. (2021) used elasticity to illustrate the relationship between extreme precipitation and flooding, where depending on ground dampness, an increase in precipitation will have an at least equivalent increase in river discharge, leading to increased flood severity. Supposing that the reduction in ARF will mitigate the impact of an increase in quantiles due to more localised rainfall, such that for example we have an approximately 7% and 19% increase under, respectively, the Cooke and BMA-EM scenarios, the relationship between discharge and rainfall would clearly imply a greater risk of increased flood losses in the latter case. ...
Climate change is expected to increase the frequency and intensity of extreme weather events. To properly assess the increased economical risk of these events, actuaries can gain in relying on expert models/opinions from multiple different sources, which requires the use of model combination techniques. From non-parametric to Bayesian approaches, different methods rely on varying assumptions potentially leading to very different results. In this paper, we apply multiple model combination methods to an ensemble of 24 experts in a pooling approach and use the differences in outputs from the different combinations to illustrate how one can gain additional insight from using multiple methods. The densities obtained from pooling in Montreal and Quebec City highlight the significant changes in higher quantiles obtained through different combination approaches. Areal reduction factor and quantile projected changes are used to show that consistency, or lack thereof, across approaches reflects the uncertainty of combination methods. This shows how an actuary using multiple expert models should consider more than one combination method to properly assess the impact of climate change on loss distributions, seeing as a single method can lead to overconfidence in projections.
... In more recent years, the QDF model has been used to characterize flood events of different duration in Algeria (Renima et al., 2018), to inform development of a depth-duration-frequency relationship used to assess risk of rainfall-driven floods in Poland Markiewicz (2021) and as a comparison point to IDF models when assessing catchment behavior for runoff extremes in Austria (Breinl et al., 2021). As noted in Breinl et al. (2021), the relationship quantified by the QDF model is an analogue to the relationship quantified in IDF modeling for precipitation extremes: in the hypothetical situation where all rainfall becomes runoff and the time of concentration is instantaneous, the QDF and IDF models have identical relationships. ...
... In more recent years, the QDF model has been used to characterize flood events of different duration in Algeria (Renima et al., 2018), to inform development of a depth-duration-frequency relationship used to assess risk of rainfall-driven floods in Poland Markiewicz (2021) and as a comparison point to IDF models when assessing catchment behavior for runoff extremes in Austria (Breinl et al., 2021). As noted in Breinl et al. (2021), the relationship quantified by the QDF model is an analogue to the relationship quantified in IDF modeling for precipitation extremes: in the hypothetical situation where all rainfall becomes runoff and the time of concentration is instantaneous, the QDF and IDF models have identical relationships. ...
... Available QDF models usually assume that only the index flood changes with duration, with the growth curve assumed constant across durations (e.g. Cunderlik and Ouarda, 2006;Breinl et al., 2021). Here the index flood is the median annual maximum flood. ...
