Prediction of flood quantiles at ungauged catchments for the contiguous USA using Artificial Neural Networks

In this study, we utilise Artificial Neural Network (ANN) models to estimate the 100- and 1500-year return levels for around 900,000 ungauged catchments in the contiguous USA. The models were trained and validated using 4,079 gauges and several selected catchment descriptors out of a total of 25 available. The study area was split into 15 regions, which represent major watersheds. ANN models were developed for each region and evaluated by calculating several performance metrics such as root-mean-squared error (RMSE), coefficient of determination (R2) and absolute percent error. The availability of a large dataset of gauges made it possible to test different model architectures and assess the regional performance of the models. The results indicate that ANN models with only one hidden layer are sufficient to describe the relationship between flood quantiles and catchment descriptors. The regional performance depends on climate type as models perform worse in arid and humid continental climates. Overall, the study suggests that ANN models are particularly applicable for predicting ungauged flood quantiles across a large geographic area. The paper presents recommendations about future application of ANN in regional flood frequency analysis.

Uncertainties and their interaction in flood hazard assessment with climate change

The assessment of future impacts of climate change is associated with a cascade of uncertainty linked to the modelling chain employed in assessing local-scale changes. Understanding and quantifying this cascade is essential for developing effective adaptation actions. We evaluate and quantify uncertainties in future flood quantiles associated with climate change for four catchments, incorporating within our modelling chain uncertainties associated with 12 global climate models contained in the Coupled Model Intercomparison Project Phase 6, five different bias correction approaches, hydrological model parameter uncertainty and the use of three different extreme value distributions for flood frequency analysis. Results indicate increased flood hazard in all catchments for different Shared Socioeconomic Pathways (SSPs), with changes in flooding consistent with changes in annual maximum precipitation. We use additive chains and analysis of variance (ANOVA) to quantify and decompose uncertainties and their interactions in estimating selected flood quantiles for each catchment. We find that not only do the contributions of different sources of uncertainty vary by catchment, but that the dominant sources of uncertainty can be very different on a catchment-by-catchment basis. While uncertainties in future projections are widely assumed to be dominated by the ensemble of climate models used, we find that in one of our catchments uncertainties associated with bias correction methods dominate, while in another the uncertainty associated with the use of different extreme value distributions outweighs the uncertainty associated with the ensemble of climate models. These findings highlight the inability to generalise a priori about the importance of different components of the cascade of uncertainty in future flood hazard at the catchment scale. Moreover, we find that the interaction of components of the modelling chain employed are substantial (> 20 % of overall uncertainty in two catchments). While our sample is small, there is evidence that the dominant components of the cascade of uncertainty may be linked to catchment characteristics and rainfall–runoff processes. Future work that seeks to further explore the characteristics of the uncertainty cascade as they relate to catchment characteristics may provide insight into a priori identifying the key components of modelling chains to be targeted in climate change impact assessments.

Do small and large floods have the same drivers of change? A regional attribution analysis in Europe

Recent studies have shown evidence of increasing and decreasing trends for average floods and flood quantiles across Europe. Studies attributing observed changes in flood peaks to their drivers have mostly focused on the average flood behaviour, without distinguishing small and large floods. This paper proposes a new framework for attributing flood changes to potential drivers, as a function of return period (T), in a regional context. We assume flood peaks to follow a non-stationary regional Gumbel distribution, where the median flood and the 100-year growth factor are used as parameters. They are allowed to vary in time and between catchments as a function of the drivers quantified by covariates. The elasticities of floods with respect to the drivers and the contributions of the drivers to flood changes are estimated by Bayesian inference. The prior distributions of the elasticities of flood quantiles to the drivers are estimated by hydrological reasoning and from the literature. The attribution model is applied to European flood and covariate data and aims at attributing the observed flood trend patterns to specific drivers for different return periods at the regional scale. We analyse flood discharge records from 2370 hydrometric stations in Europe over the period 1960–2010. Extreme precipitation, antecedent soil moisture and snowmelt are the potential drivers of flood change considered in this study. Results show that, in northwestern Europe, extreme precipitation mainly contributes to changes in both the median (q2) and 100-year flood (q100), while the contributions of antecedent soil moisture are of secondary importance. In southern Europe, both antecedent soil moisture and extreme precipitation contribute to flood changes, and their relative importance depends on the return period. Antecedent soil moisture is the main contributor to changes in q2, while the contributions of the two drivers to changes in larger floods (T>10 years) are comparable. In eastern Europe, snowmelt drives changes in both q2 and q100.

Impact of climate change on floods in Pamplona (Spain) by using climate change projections and a distributed hydrological model

<p>Understanding how floods are expected to change is essential for decision making and flood risk management, as flood risks are expected to increase in the future. Several studies have analysed the impact of climate change on flood risks with rainfall-runoff models and climate projections as input data. Nevertheless, most of these studies involve large-scale river basins instead of focusing on smaller river basins or points of interest like urban areas. This study quantifies the expected changes in flood quantiles at the River Arga in the city of Pamplona (Spain) within the SAFERDAMS project (PID2019-107027RB-I00) funded by the Spanish Ministry of Science and Innovation. It uses climate change projections from 12 climate models of the EURO-CORDEX programme for two Representative Concentration Pathways - RCPs as input data of the RIBS distributed hydrological model (Garrote and Bras 1995 ab, JoH). The analysis considers seven return periods (2, 5, 10, 50, 100, 500 and 1000 years), two greenhouse gas emission scenarios (RCP4.5 and RCP8.5) and three time windows (2011-2040, 2041-2070 and 2070-2100).</p><p>First, the RIBS model has been calibrated with a set of objective functions to minimise the bias between simulations and observations recorded at a streamflow-gauging station located in the Arga River in Pamplona. The seven greatest flood events occurred in Pamplona in the last decade are considered. A long set of random combinations of model parameter values are used. The combination of parameter values that led to the smallest errors were selected.</p><p>Second, 24-h design rainfall storms with a time step of 1 h in the current scenario at a set of rainfall gauge stations in the Arga River catchment are obtained by using an extreme frequency analysis. Expected changes in daily rainfall quantiles in the Arga River catchment obtained by processing climate change projections are used (Garijo and Mediero 2019, Water). Current and future design rainfall storms were obtained for the seven return periods, two RCPs and three time windows. The input data in the RIBS model are provided in a raster format. Hence, design rainfall storms were transformed into spatial distributions of precipitation with the Thiessen polygons technique.</p><p>The findings show a decrease in design peak discharges for return periods smaller than 10 years and an increase for the 500- and 1000-year floods for both RCPs in the three time windows. However, 50- and 100-year return period flood quantiles are expected to increase especially in the 2041-2070 and 2071-2100 time windows only in the emission scenario RCP8.5. The emission scenario RCP8.5 always provides greater increases in flood quantiles than RCP4.5, except for the more frequent floods (2, 5 and 10 years) in the time window 2011-2040. The increases of design discharges are 10-30% higher in RCP8.5 than in RCP4.5 for the greatest return periods. Therefore, flood magnitude changes for the most extreme events seem to be related to the evolution of greenhouse gasses emissions, following the same behaviour of the RCPs: the greatest expected changes are in the 2040 for the RCP4.5 and in the 2100 for the RCP8.5.</p>

Data-based attribution of changes in flood quantiles across Europe between 1960 and 2010

<p>Changes in European floods during past decades have been analysed and detected by several studies. These studies typically focused on the mean flood behaviour, without distinguishing small and large floods. In this work, we investigate the causes of the detected flood trends across Europe over five decades (1960-2010), as a function of the return period. We adopt a regional non-stationary flood frequency approach to attribute observed flood changes to potential drivers, used as covariates of the parameters of the regional probability distribution of floods. The elasticities of floods with respect to the drivers and the regional contributions of the drivers to changes in flood quantiles associated with small and large return periods (i.e. 2-year and 100-year floods, respectively) are estimated by Bayesian inference, with prior information on the elasticity parameters obtained from expert knowledge and the literature. The data-based attribution approach is applied to annual maximum flood discharge seires from 2370 hydrometric stations in Europe. Extreme precipitation, antecedent soil moisture and snowmelt are the potential drivers considered. Results show that extreme precipitation mainly contributes to positive flood changes in North-western Europe. Both antecedent soil moisture and extreme precipitation contribute to negative flood changes in Southern Europe, with relative contributions varying with the return period. Antecedent soil moisture contributes the most to changes in small floods (i.e. T=2-10 years), while the two drivers contribute with comparable magnitude to changes in more extreme events. In eastern Europe, snowmelt clearly drives negative changes in both small and large floods.</p>

Uncertainties and their interaction in flood risk assessment with climate change

The assessment of future impacts of climate change is associated with a cascade of uncertainty linked to the modelling chain employed in assessing local scale changes. Understanding and quantifying this cascade is essential to developing effective adaptation actions. We evaluate and quantify uncertainties in future flood quantiles associated with climate change for four Irish catchments, incorporating within our modelling chain uncertainties associated with 12 Global Climate Models contained in the Coupled Model Inter-comparison Project Phase 6, five different bias correction approaches, hydrological model parameter uncertainty and use of three different extreme value distributions for flood frequency analysis. Results indicate increased flood risk in all catchments for different Shared Socioeconomic Pathways (SSPs), with changes in flooding related to changes in annual maximum precipitation. We use a sensitivity test based on the analysis of variance (ANOVA) to decompose uncertainties and their interactions in estimating selected flood quantiles in the 2080s for each catchment. We find that the dominant sources of uncertainty vary between catchments, calling into question the ability to generalise about the importance of different components of the cascade of uncertainty in future flood risk. For two of our catchments, uncertainties associated with bias correction methods and extreme value distributions outweigh the uncertainty associated with the ensemble of climate models. For all catchments and flood quantiles examined, hydrological model parameter uncertainty is the least important component of our modelling chain, while the uncertainties derived from the interaction of components are substantial (>20 percent of overall uncertainty in two catchments). While our sample is small, there is evidence that the dominant components of the cascade of uncertainty may be linked to catchment characteristics and rainfall runoff processes. Future work that seeks to further explore the dominant components of uncertainty as they relate to catchment characteristics may provide insight into a priori identifying the key components of modelling chains to be included in climate change impact assessments.

Do small and large floods have the same drivers of change? A regional attribution analysis in Europe

Recent studies have shown evidence of increasing and decreasing trends in mean annual floods and flood quantiles across Europe. Studies attributing observed changes in flood peaks to their drivers have mostly focused on mean annual floods. This paper proposes a new framework for attributing flood changes to potential drivers, as a function of return period (T), in a regional context. We assume flood peaks to follow a non-stationary regional Gumbel distribution, where the median flood and the 100-year growth factor are used as parameters. They are allowed to vary in time and between catchments as a function of the drivers quantified by covariates. The elasticities of floods with respect to the drivers and the contributions of the drivers to flood changes are estimated by Bayesian inference. The prior distributions of the elasticities of flood quantiles to the drivers are estimated by hydrological reasoning and from the literature. The attribution model is applied to European flood and covariate data and aims at attributing the observed flood trend patterns to specific drivers for different return periods. We analyse flood discharge records from 2370 hydrometric stations in Europe over the period 1960–2010. Extreme precipitation, antecedent soil moisture and snowmelt are the potential drivers of flood change considered in this study. Results show that, in northwestern Europe, extreme precipitation mainly contributes to changes in both the median (q2) and 100-year flood (q100), while the contributions of antecedent soil moisture are of secondary importance. In southern Europe, both antecedent soil moisture and extreme precipitation contribute to flood changes, and their relative importance depends on the return period. Antecedent soil moisture is the main contributor to changes in q2, while the contributions of the two drivers to changes in larger floods (T > 10 years) are comparable. In eastern Europe, snowmelt drives changes in both q2 and q100.