Fig 3 - uploaded by Korbinian Breinl
Content may be subject to copyright.
Relationship between flood i q and rainfall i P intensities (a,b) and related elasticities (c,d) estimated from observed rainfall time series at the rain gauge "Hohe Warte" and a linear reservoir over a range of storage coefficients k. (a) Relationship for d = 1hr and (b) T = 2yrs. (c) Elasticities ε 1 assuming d = 1hr is constant at a return period of T = 2yrs. Blue to red points correspond to varying k and a runoff coefficient rc = 1; Gray points correspond to the median of 100 simulations for each k with rc varying randomly according to a beta distribution with parameters α = 2 and β = 5, i.e. a mean of about 0.29. The shaded area corresponds to the 5th and 95th percentiles of the simulations. (d) Elasticities ε 2 assuming T = 2yrs is constant at a duration of d = 1hr. Blue to red points correspond to varying k and rc = 1; Gray line and shaded area correspond to randomly varying rc as in (c).
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
... give a more intuitive illustration of the IDF-QDF relationship, we performed simulations convoluting an observed rainfall time series (43 years) with a linear reservoir, and analyzing the resulting hydrographs according to Equation (3). Fig. 3a shows the mapping of IDF and QDF curves for d = 1hr and different return periods T implicit in the magnitudes of i q and i P . For the limiting case of a response time of 0 (i.e. storage coefficient of the linear reservoir k = 0) and a runoff coefficient r c = 1, streamflow is equal to rainfall and the relationship plots on the 1:1 ...
Context 2
... return periods T implicit in the magnitudes of i q and i P . For the limiting case of a response time of 0 (i.e. storage coefficient of the linear reservoir k = 0) and a runoff coefficient r c = 1, streamflow is equal to rainfall and the relationship plots on the 1:1 line. As k increases, the streamflow event peaks decrease and so does i q . Fig. 3b shows the relationship for T q = T P = 2yrs and different durations d implicit in the magnitudes of i q and i P . Again streamflow decreases with increasing k, particularly for short durations. For long durations, as demonstrated through experiments assuming block rainfall and varying storm durations d s (Blöschl & Sivapalan, 1997), ...
Context 3
... simulations also allow an illustration of the behavior of the elasticities ε 1 and ε 2 . For a storage coefficient k = 0 and a runoff coefficient rc = 1, both ε 1 and ε 2 are unity, due to the 1:1 mapping of precipitation to streamflow, meaning a 1% change in rainfall relates to a 1% change in streamflow (Fig. 3c). With increasing k, the elasticity ε 1 decreases to about 0.85. This is because, with higher k, rainfall in the time steps adjoining the annual maxima becomes increasingly more relevant in determining the maximum annual streamflow due to the convolution of the linear reservoir. While for small k the annual maxima of the rainfall and ...
Context 4
... of the peak reduction on d/k. In order to test the sensitivity of the elasticities to the presence of randomness in the runoff coefficient, we assumed the runoff coefficient for each event (defined as rainfall wet spell) to vary randomly between 0 and 1 according to a beta distribution with a mean of 0.29 (α = 2, β = 5) instead of using rc = 1 (Fig. 3c and Fig. 3d). These values of α and β are representative of medium rainfall regions in Austria ( Merz et al., 2006). Random rc increase the elasticity ε 1 , which is related to the possibility of combinations of large rainfall with large runoff coefficients, which steepens the flood frequency curve and thus increases ε 1 ( ). More skewed beta ...
Context 5
... rainfall and streamflow model parameters reflecting the rainfall and flood intensities (shown in Fig. 5 and Fig. 6) are given in Fig. 7 (corresponding maps in Fig. A3). The "Eastern mixed" region exhibits the highest values of the scale parameter λ P , reflecting the high variability of rainfall extremes, while the variability of streamflow extremes (λ q ) is highest in the Northern orographic region (Fig. 7a, also see Fig. A3a, b). Hence, the variability of streamflow extremes is spatially not ...
Context 6
... intensities (shown in Fig. 5 and Fig. 6) are given in Fig. 7 (corresponding maps in Fig. A3). The "Eastern mixed" region exhibits the highest values of the scale parameter λ P , reflecting the high variability of rainfall extremes, while the variability of streamflow extremes (λ q ) is highest in the Northern orographic region (Fig. 7a, also see Fig. A3a, b). Hence, the variability of streamflow extremes is spatially not aligned with that of the rainfall ...
Context 7
... location parameters ψ P and ψ q (Fig. 7b) have a similar spatial distribution (r s = 0.69). Both are smallest in the Northeastern convective region and largest in the Western orographic and the Southern mixed regions. Both parameters are highest in the high elevation zones of Austria (Fig. A3c, d). This result can be interpreted as the influence of orographic rainfall and perhaps also higher runoff coefficients (Merz & Blöschl, 2009b). Also, as ψ P and ψ q define the CV of the extremes (see Equation (6)), one can conclude that the CV of rainfall is higher in the lowlands influenced by convective mechanisms compared to the ...
Context 8
... η q are generally lower than the ones of the rainfall (η P ) reflecting the dampening effects of catchment processes (Fig. 7c). Moreover, η P and η q are spatially aligned with the largest values in the Northeastern convective and Eastern mixed regions. This behavior reflects the higher convective activity and flashier flood response (also see Fig. A3e, f). There are striking similarities between the spatial distribution of η q and the concentration times of catchments as analyzed by Gaal et al. (2012) from flood events ( Gaal et al. (2012) , Fig. 4), which is not surprising as η q reflects the catchment response ...
Context 9
... of annual rainfall and streamflow maxima that occurred within 24hrs of each other for the duration of 1hr (left boxplots) and 24hrs (right boxplots), stratified by rainfall regions. The synchronicity of both durations is highest in the lowlands of the North and East, which are regions of flashy catchment response represented by large η q (Fig. A3f). In these regions, also elasticities ε 1 (Fig. 9) are highest. This is in line with our simulation experiments (Fig. 3c), which indicated small storage coefficients k to cause high elasticities ε 1 ...
Context 10
... and 24hrs (right boxplots), stratified by rainfall regions. The synchronicity of both durations is highest in the lowlands of the North and East, which are regions of flashy catchment response represented by large η q (Fig. A3f). In these regions, also elasticities ε 1 (Fig. 9) are highest. This is in line with our simulation experiments (Fig. 3c), which indicated small storage coefficients k to cause high elasticities ε 1 ...
Context 11
... 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 ...
Context 12
... 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 of rainfall and streamflow (Fig. 5 and Fig. 6), with more ...
Context 13
... moments and parameters of the rainfall distribution are aligned with the regional rainfall mechanisms. In the high elevation catchments where orographic rainfall dominates, the location parameter ψ P tends to be high and the scale parameter λ P tends to be small, i.e. the rainfall extremes tend to be large with little temporal variability (see Fig. A3a and A3c). The opposite applies to catchments in the lowlands where convective rainfall extremes are more frequent. The relationship between rainfall magnitude and variability is reflected by a high negative correlation between ψ P and λ P (Fig. 4a and Fig. 8, r S = 0.62). The regional rainfall mechanisms also manifest themselves in the ...
Context 14
... and variability is reflected by a high negative correlation between ψ P and λ P (Fig. 4a and Fig. 8, r S = 0.62). The regional rainfall mechanisms also manifest themselves in the spatial distribution of the scaling parameter η P , which tends to be higher in the convective lowlands than in the mountainous regions with dominant orographic rainfall (Fig. A3e). A higher parameter η P implies that the intensity decreases more strongly with duration which can be expected for the short rainstorms that occur frequently in the ...
Context 15
... in the case of rainfall, the location parameter of streamflow ψ q is higher in the mountain catchments than in the lowlands (see Fig. A3d), reflected by a strong positive correlation of ψ q with catchment elevation (Fig. 8, r S = 0.77) and mean summer rainfall (Fig. 8, r S = 0.55). This means, the magnitude of floods tends to be higher in the orographic rainfall regions, not only due the higher extreme rainfall magnitudes but also due to high annual rainfall amounts, ...
Context 16
... propensity towards saturation excess overflow. The variability of streamflow represented by the scale parameter λ q , however, is controlled by catchment topography, soil type and the geology (e.g. correlations with slope r S = 0.33, Rendzina soils r S = 0.21 or Carbonate rock geology r S = 0.28, Fig. 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 ...
Context 17
... correlations with slope r S = 0.33, Rendzina soils r S = 0.21 or Carbonate rock geology r S = 0.28, Fig. 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, ...
Context 18
... 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 ( Li et al., 2017) and thus more ...
Context 19
... the rainfall may be stored in the fractured carbonic rocks, while more extreme rainfall events can saturate the epikarst zone inducing large streamflow extremes ( Li et al., 2017) and thus more variable floods. Such step changes in streamflow extremes may also occur in other geological formations (Rogger et al., 2012). The scaling parameter η q (Fig. A3f), on the other hand, is highest in the lowlands, showing a similarity with the spatial distribution of the rainfall scaling parameter η P , which is also lowest in the ...
Context 20
... a steeper rainfall frequency curve is associated with a steeper flood frequency curve, which is in line with previous studies (e.g. Merz & Blöschl, 2003;Smith et al., 2011;Villarini & Smith, 2010). The scale parameter of rainfall is not aligned with that of streamflow, reflected by r S = 0.14 between λ P and λ q and different spatial patterns (Fig. A3a, A3b and Fig. ...
Context 21
... 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 synchronicity (Fig. A4). The decoupling is even more pronounced in the glaciated ...
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
... 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. ...
... Typically, daily predictions are more complex than monthly or annual predictions. In general, precipitation and stream discharge are highly correlated (Breinl et al., 2021;Rohith et al., 2021). The minimal reduction in model performance (Table 1) for the case without discharge (Q) as input (Case 4) indicates precipitation as a good surrogate for discharge in modeling continuous daily stream nitrate concentration. ...
High-frequency stream nitrate concentration provides critical insights into nutrient dynamics and can help to improve the effectiveness of management decisions to maintain a sustainable ecosystem. However, nitrate monitoring is conventionally conducted through lab analysis using in situ water samples and is typically at coarse temporal resolution. In the last decade, many agencies started collecting high-frequency (5-60 min intervals) nitrate data using optical sensors. The hypothesis of the study is that the data-driven models can learn the trend and temporal variability in nitrate concentration from high-frequency sensor-based nitrate data in the region and generate continuous nitrate data for unavailable data periods and data-limited locations. A Long Short-Term Memory (LSTM) model-based framework was developed to estimate continuous daily stream nitrate for dozens of gauge locations in Iowa, USA. The promising results supported the hypothesis; the LSTM model demonstrated median test-period Nash-Sutcliffe efficiency (NSE) = 0.75 and RMSE = 1.53 mg/L for estimating continuous daily nitrate concentration in 42 sites, which are unprecedented performance levels. Twenty-one sites (50 % of all sites) and thirty-four sites (76 % of all sites) demonstrated NSE >0.75 and 0.50, respectively. The average nitrate concentration of neighboring sites was identified as a crucial determinant of continuous daily nitrate concentration. Seasonal model performance evaluation showed that the model performed effectively in the summer and fall seasons. About 26 sites showed correlations >0.60 between estimated nitrate concentration and discharge. The concentration-discharge (c-Q) relationship analysis showed that the study watersheds had four dominant nitrate transport patterns from landscapes to streams with increasing discharge, including the flushing pattern being the most dominant one. Stream nitrate estimation impedes due to data inadequacy. The modeling framework can be used to generate temporally continuous nitrate at nitrate data-limited regions with a nearby sensor-based nitrate gauge. Watershed planners and policymakers could utilize the continuous nitrate data to gain more information on the regional nitrate status and design conservation practices accordingly.
... 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. ...
... These methods mainly focused on at-site estimates centered on the multifractality of flood quantiles (Pandey et al., 1998;Sauquet et al., 2008) and streamflow and rainfall dynamics (Tessier et al., 1996). In addition, a large body of flood scaling literature explores evaluating the flood frequency curves, intensity-duration-frequency curves (Breinl et al., 2021), flood elasticity (Sankarasubramanian et al., 2001), Fourier power spectrum (Telesca et al., 2012), regional flood frequency analysis using index flood method (Ishak et al., 2011;Stephens et al., 2015), and varying quantile regression techniques (Gupta & Waymire, 1990;Telesca et al., 2012;Tessier et al., 1996;Zaman et al., 2013). For example, the flood frequency analysis utilized by the United State Geological Survey (USGS) fits a log-Pearson Type III (LP3) probability distribution to annual flood peaks observed at individual streamgage in a region (Benson, 1962;US-Water-Resources-Council, 1982). ...
Accurate flood risk assessment requires a comprehensive understanding of flood sensitivity to regional drivers and climate factors. This paper presents the scaling of floods (duration, peak, volume) with geomorphologic characteristics of the basin (i.e., drainage area, slope, elevation) and precipitation patterns (rainfall accumulation, variability). Long‐term daily streamflow observations over the 20th and early 21st centuries from Hydro‐Climatic Data Network streamgages across the conterminous United States are used to create a flood event database based on their flood stage information. Antecedent daily rainfall accumulation and variability corresponding to these floods are computed using Global Historical Climatology Network daily data set. Two Bayesian scaling models are developed, and the spatial organization of scaling exponents is investigated. The baseline model quantifies the scaling of floods to geomorphologic characteristics. The dynamic model quantifies the scaling of floods to antecedent precipitation distribution which is further conditioned on geomorphologic characteristics. Results show that small and low‐elevation basins have a stronger response to antecedent rainfall distribution in amplifying flood peaks, while high‐elevation steeper basins have a lower response for flood duration and volume. The dynamic models demonstrate that there are significant variations in the flood scaling rates, with the largest rates up to 40% and 4.5% for flood duration, 64% and 44% for peak, and 98% and 40% for volume found across the Northeast, Coastal Southeast, and Northwest with intensifying rainfall accumulation and variability, respectively. This study advances flood predictions by better informing the flood attributes in the context of dynamical land‐atmosphere perturbations.
... lokasinya yang berada didekat pantai tersebut. Namun, saluran drainase sudah tidak mampu menampung air saat hujan dengan intensitas tinggi dan pasang surut air laut yang terjadi secara beriringan [4] [5]. ...
Drainage is a basic facility that must be provided to meet community needs, so it’s very important in urban spatial planning. The benefit of building drainage is to prevent stagnant water so that it does not interfere with community activities. The condition of the drainage channels on Jalan Marina isn’t maintained because the dimensions of the channels are small and sedimentation occurs, causing silting of the channels. This causes floods that always occur every year. Re-dimension of the drainage channel was carried out because the initial dimensions could not accommodate the planned flood discharge. Some of data the basic for determining the re-dimensional drainage channels are rainfall data, sea tide data, and initial drainage channel dimension data. From the calculation results, the value of planned flood discharge is 172.634 mᶟ/s. So, the dimensions of the drainage channel are re-planned because the channel discharge value is smaller than the planned flood discharge value. The results of this study are that the planned flood discharge is greater than the drainage channel water discharge so that the initial channel dimensions cannot accommodate water. The results of re-planning the dimensions of the drainage channel based on the planned flood discharge value are h of 8.5 meters, b of 9.8 meters, and W of 2 meters. Re-dimensional planning of the drainage channel is fulfilled because Qs<Qp (172.634 mᶟ/s <227.103 mᶟ/s). This research is expected to provide dimensions of drainage channels that are in accordance with field conditions so that no more flooding occurs.
... To depict the characteristics of storms and floods, various statistical methods have been adopted. Intensity-duration-frequency (IDF) curve is a widely-used method to depict the precipitation process of storms (Baeck et al., 2011;Breinl et al., 2021;Breinl et al., 2020;Cheng and AghaKouchak, 2014;Hosseinzadehtalaei et al., 2020;Ombadi et al., 2018;Sadegh et al., 2017), and the flood peaks are depicted by flood frequency curves in previous studies (Baeck et al., 2011;Blöschl and Sivapalan, 1997;Breinl et al., 2021;Serinaldi, 2011;Villarini and Smith, 2010). Based on extreme value theory, previous studies suggest that Generalized Extreme Value (GEV) and Gumbel distribution are used to capture the distributions of peak values in annual maximum (AM) methods, and the exponential and Generalized Pareto distribution are used to capture the distributions of peak values in peak over threshold methods (Gao et al., 2016;Serinaldi and Kilsby, 2014;Wang, 1991). ...
... To depict the characteristics of storms and floods, various statistical methods have been adopted. Intensity-duration-frequency (IDF) curve is a widely-used method to depict the precipitation process of storms (Baeck et al., 2011;Breinl et al., 2021;Breinl et al., 2020;Cheng and AghaKouchak, 2014;Hosseinzadehtalaei et al., 2020;Ombadi et al., 2018;Sadegh et al., 2017), and the flood peaks are depicted by flood frequency curves in previous studies (Baeck et al., 2011;Blöschl and Sivapalan, 1997;Breinl et al., 2021;Serinaldi, 2011;Villarini and Smith, 2010). Based on extreme value theory, previous studies suggest that Generalized Extreme Value (GEV) and Gumbel distribution are used to capture the distributions of peak values in annual maximum (AM) methods, and the exponential and Generalized Pareto distribution are used to capture the distributions of peak values in peak over threshold methods (Gao et al., 2016;Serinaldi and Kilsby, 2014;Wang, 1991). ...
... The relationship between storms and floods is often characterized by the flood runoff coefficient (the ratio of flood peak to storm peak), which is found to be affected by land surface conditions, including catchment size, land cover type, and antecedent soil moisture (Sriwongsitanon and Taesombat, 2011), as well as rainfall characteristics, including the causative mechanisms of precipitation extremes, rainfall intensity and rainfall process (Breinl et al., 2021;Sharma et al., 2018). The flood runoff coefficient captures the relationship between the peak storm and peak flood directly in a certain flood event. ...
Storms and the resultant floods have always been catastrophic disasters and raised increasing global concerns in the context of climate change. However, the relationships between storms and floods remain largely elusive. Here we examine the storm-flood relationship and its variations in the Upper Chao Phraya River Basin (UCPRB), a typical tropical monsoon basin in southeast Asia. The distributions of storms and floods are characterized by statistical models with the aid of climatic and land surface covariates. The storm-flood relationship is depicted by the concept of storm-flood elasticity, which represents the corresponding changes in flood peaks in response to changes in the storm peaks with the same return period. The storm-flood elasticity coefficients for 100-year return period events range from 0.61 to 1.20, and the values of storm-flood elasticity coefficients tend to be smaller for long-return period events than for short-return-period events, under high-typhoon-precipitation (high-TP), high-non-typhoon-precipitation (high-nTP), and low-forest (low-F) conditions, and in humid regions than in arid regions. The climatic covariates are shown to have stronger effects on the storm-flood elasticity coefficient than the land surface covariate in most basins. In the basins where deforestation shows strong impacts on the storm-flood relationship, afforestation can be an effective approach for flood control. In most basins in the UCPRB, the variation of typhoon precipitation has larger impacts than those of non-typhoon precipitation, indicating that typhoon precipitation should be paid more attention to when considering the future changes in floods. The findings help develop a better understanding of storm-flood relationships in tropical monsoon regions, and the methods of this study can also be applied in other climate regions.
... However, in real world basins this is not the case, and the return period of rainfall is not the same as the return period of peak discharge. As a result, researchers have raised concerns regarding the design storm method (e.g., [4,5]). ...
In order to examine the relationship between rainfall return periods and flood return periods, the design storm approach is compared to the rainfall–runoff continuous simulation and flood frequency analysis approach. The former was based on rainfall frequency analysis and event-based hydrological simulations, while the latter was based on continuous hydrological simulations and flood frequency analysis. All hydrological simulations were undertaken employing the HEC-HMS software. For the rainfall frequency analysis, the Generalized Extreme Value (GEV) probability distribution was used. For the flood frequency analysis, both the Extreme Value Type I (Gumbel) and GEV theoretical distributions were used and compared to each other. Flood hazard (inundation depth, flow velocities and flood extent) was estimated based on hydrodynamic simulations employing the HEC-RAS software. The study area was the Pineios catchment, upstream of Larissa city, Greece. The results revealed that the assumption of equivalent return periods of rainfall and discharge is not valid for the study area. For instance, a 50-year return period flood corresponds to a rainfall return period of about 110 years. Even if flow measurements are not available, continuous simulation based on re-analysis datasets and flood frequency analysis may be alternatively used.
... Rainfall-runoff simulation is one of the core approaches in hydrology. However, there has been a struggle to thoroughly understand the link between rainfall and runoff because of the intricate interactions between land use, soil properties, and precipitation patterns [1][2][3][4]. Hydrological processes can be captured using physical-based or data-driven simulation models. Physically based models need much time and effort to comprehend the water cycle thoroughly. ...
Accurate streamflow simulation is crucial for many applications, such as optimal reservoir operation and irrigation. Conceptual techniques employ physical ideas and are suitable for representing the physics of the hydrologic model, but they might fail in competition with their more advanced counterparts. In contrast, deep learning (DL) approaches provide a great computational capability for streamflow simulation, but they rely on data characteristics and the physics of the issue cannot be fully understood. To overcome these limitations, the current study provided a novel framework based on a combination of conceptual and DL techniques for enhancing the accuracy of streamflow simulation in a snow-covered basin. In this regard, the current study simulated daily streamflow in the Kalixälven river basin in northern Sweden by integrating a snow-based conceptual hydrological model (MISD) with a DL model. Daily precipitation, air temperature (average, minimum, and maximum), dew point temperature, evapotranspiration, relative humidity, sunshine duration, global solar radiation, and atmospheric pressure data were used as inputs for the DL model to examine the effect of each meteorological variable on the streamflow simulation. Results proved that adding meteorological variables to the conceptual hydrological model underframe of parallel settings can improve the accuracy of streamflow simulating by the DL model. The MISD model simulated streamflow had an MAE = 8.33 (cms), r = 0.88, and NSE = 0.77 for the validation phase. The proposed deep-conceptual learning-based framework also performed better than the standalone MISD model; the DL method had an MAE = 7.89 (cms), r = 0.90, and NSE = 0.80 for the validation phase when meteorological variables and MISD results were combined as inputs for the DL model. The integrated rainfall-runoff model proposed in this research is a new concept in rainfall runoff modeling which can be used for accurate streamflow simulations.
... Nevertheless, there are reasons to believe that climate change may indeed increase the flood hazard in small catchments. For large return periods, the increase may be of a similar magnitude to that of precipitation since the rainfall-runoff relationship tends to become linear, while for smaller return periods, the relative increase may be larger Breinl et al., 2021). ...
There is serious concern that the hazard, or probability,
of river floods is increasing over time. Starting from narratives that are
sometimes discussed in public, the article addresses three hypotheses. The
first suggests that land-use changes, such as deforestation, urbanisation
and soil compaction by agriculture, increase flood hazards. This review finds that land-use effects on floods are particularly pronounced in small
catchments since soil permeability plays an important role in infiltration at this scale. For regional floods, and the most extreme events, land use is usually not the most important control, since areas of soil saturation play a greater role in runoff generation, which are less dependent on soil
permeability. The second hypothesis suggests that hydraulic interventions
and structures, such as river training, levees and dams, increase flood
hazards. This review finds that hydraulic structures have the greatest impact on events of medium magnitude, associated with return periods of tens to hundreds of years, and that their effects are usually local. Long-term interactions between humans and floods must be taken into account when predicting future flood hazards. The third hypothesis suggests that climate change increases flood hazard. This review finds that, in small catchments of a few hectares, flood hazards may increase due to convective storms. In large catchments, where regional floods occur, changes are not necessarily directly related to precipitation, nor are they directly related to rising air temperatures, but are determined by the seasonal interplay of soil moisture, snow and extreme precipitation via runoff generation. Increases and decreases in flood hazards have been observed worldwide. It is concluded that significant progress has been made in recent years in understanding the role of land use, hydraulic structures and climate in changing river flood hazards. It is crucial to consider all three factors of change in flood risk management and communicate them to the general public in a nuanced way.
