Predicting nonstationary flood frequencies: Evidence supports an updated stationarity thesis in the United States

Nonstationary extreme value analysis (NEVA) can improve the statistical representation of observed flood peak distributions compared to stationary (ST) analysis, but management of flood risk relies on predictions of out-of-sample distributions for which NEVA has not been comprehensively evaluated. In this study, we apply split-sample testing to 1250 annual maximum discharge records in the United States and compare the predictive capabilities of NEVA relative to ST extreme value analysis using a log-Pearson Type III (LPIII) distribution. The parameters of the LPIII distribution in the ST and nonstationary (NS) models are estimated from the first half of each record using Bayesian inference. The second half of each record is reserved to evaluate the predictions under the ST and NS models. The NS model is applied for prediction by (1) extrapolating the trend of the NS model parameters throughout the evaluation period and (2) using the NS model parameter values at the end of the fitting period to predict with an updated ST model (uST). Our analysis shows that the ST predictions are preferred, overall. NS model parameter extrapolation is rarely preferred. However, if fitting period discharges are influenced by physical changes in the watershed, for example from anthropogenic activity, the uST model is strongly preferred relative to ST and NS predictions. The uST model is therefore recommended for evaluation of current flood risk in watersheds that have undergone physical changes. Supporting information includes a MATLAB® program that estimates the (ST/NS/uST) LPIII parameters from annual peak discharge data through Bayesian inference.

[1]  Francesco Serinaldi,et al.  Stationarity is undead: Uncertainty dominates the distribution of extremes , 2015 .

[2]  A. Aghakouchak.  Extremes in a Changing Climate: Detection, Analysis and Uncertainty , 2013 .

[3]  Demetris Koutsoyiannis,et al.  Negligent killing of scientific concepts: the stationarity case , 2015 .

[4]  R. Hirsch,et al.  Fragmented patterns of flood change across the United States , 2016, Geophysical research letters.

[5]  L. Brekke,et al.  A framework for assessing flood frequency based on climate projection information , 2009 .

[6]  K. Trenberth.  Changes in precipitation with climate change , 2011 .

[7]  Reinhard Viertl,et al.  A fuzzy Bayesian approach to flood frequency estimation with imprecise historical information , 2016, Water resources research.

[8]  Michel Lang,et al.  Bayesian Methods for Non-stationary Extreme Value Analysis , 2013 .

[9]  S. Vicente‐Serrano,et al.  Assessing trends in extreme precipitation events intensity and magnitude using non‐stationary peaks‐over‐threshold analysis: a case study in northeast Spain from 1930 to 2006 , 2011 .

[10]  J. R. Wallis,et al.  A daily hydroclimatological data set for the continental United States , 1991 .

[11]  Casey Brown,et al.  Special Section on Climate Change and Water Resources: Climate Nonstationarity and Water Resources Management , 2012 .

[12]  Daniel Cooley,et al.  Extreme value analysis and the study of climate change , 2009 .

[13]  Ilaria Prosdocimi,et al.  Non-stationarity in annual and seasonal series of peak flow and precipitation in the UK , 2013 .

[14]  M. Lang,et al.  Statistical analysis of extreme events in a non-stationary context via a Bayesian framework: case study with peak-over-threshold data , 2006 .

[15]  Jery R. Stedinger,et al.  Flood Frequency Analysis in the Context of Climate Change , 2015 .

[16]  Timothy A. Cohn,et al.  Nature's style: Naturally trendy , 2005 .

[17]  A. Aghakouchak,et al.  Nonstationary Precipitation Intensity-Duration-Frequency Curves for Infrastructure Design in a Changing Climate , 2014, Scientific Reports.

[18]  Günter Blöschl,et al.  Hydrological modelling in a changing world , 2011 .

[19]  H. B. Mann.  Nonparametric Tests Against Trend , 1945 .

[20]  Laura Read,et al.  Hazard function analysis for flood planning under nonstationarity , 2016 .

[21]  T. McMahon,et al.  Log Pearson III distribution — Effect of dependence, distribution parameters and sample size on peak annual flood estimates , 1981 .

[22]  Jürgen Jensen,et al.  Nonstationary extreme value analysis of annual maximum water levels for designing coastal structures on the German North Sea coastline , 2010 .

[23]  David R. Anderson,et al.  Model selection and multimodel inference : a practical information-theoretic approach , 2003 .

[24]  Félix Francés,et al.  Non-stationary flood frequency analysis in continental Spanish rivers, using climate and reservoir indices as external covariates , 2013 .

[25]  W. J. Cody,et al.  Chebyshev Approximations for the Natural Logarithm of the Gamma Function , 1967 .

[26]  Richard W. Katz,et al.  Statistics of extremes in climate change , 2010 .

[27]  Sergio M. Vicente-Serrano,et al.  Mapping the hazard of extreme rainfall by peaks-over-threshold extreme value analysis and spatial regression techniques , 2006 .

[28]  D. Higdon,et al.  Accelerating Markov Chain Monte Carlo Simulation by Differential Evolution with Self-Adaptive Randomized Subspace Sampling , 2009 .

[29]  Donald H. Burn,et al.  Non-stationary pooled flood frequency analysis , 2003 .

[30]  T. Ouarda,et al.  Bayesian Nonstationary Frequency Analysis of Hydrological Variables 1 , 2011 .

[31]  Upmanu Lall,et al.  A climate informed model for nonstationary flood risk prediction: Application to Negro River at Manaus, Amazonia , 2015 .

[32]  Andrés Díez-Herrero,et al.  Flood frequency analysis of historical flood data under stationary and non-stationary modelling , 2015 .

[33]  Eric Gilleland,et al.  New Software to Analyze How Extremes Change Over Time , 2011 .

[34]  H. Lins,et al.  Stationarity: Wanted Dead or Alive? 1 , 2011 .

[35]  T. Ouarda,et al.  Generalized maximum likelihood estimators for the nonstationary generalized extreme value model , 2007 .

[36]  Luc Neppel,et al.  Non-stationary frequency analysis of heavy rainfall events in southern France , 2013 .

[37]  Bruno Merz,et al.  Floods and climate: emerging perspectives for flood risk assessment and management , 2014 .

[38]  Michel Lang,et al.  Understanding Flood Regime Changes in Europe: A state of the art assessment , 2013 .

[39]  Richard M. Vogel,et al.  Nonstationarity: Flood Magnification and Recurrence Reduction Factors in the United States 1 , 2011 .

[40]  B. Merz,et al.  Attribution of regional flood changes based on scaling fingerprints , 2016, Water resources research.

[41]  J. R. Wallis,et al.  Regional Frequency Analysis: An Approach Based on L-Moments , 1997 .

[42]  P. Bates,et al.  Flood frequency analysis for nonstationary annual peak records in an urban drainage basin , 2009 .

[43]  L. Perreaulta,et al.  Bayesian change-point analysis in hydrometeorological time series . Part 2 . Comparison of change-point models and forecasting , 2000 .

[44]  N. Matalas,et al.  Comment on the Announced Death of Stationarity , 2012 .

[45]  J. Salas,et al.  Revisiting the Concepts of Return Period and Risk for Nonstationary Hydrologic Extreme Events , 2014 .

[46]  F. Massey.  The Kolmogorov-Smirnov Test for Goodness of Fit , 1951 .

[47]  Jery R. Stedinger,et al.  Bayesian MCMC flood frequency analysis with historical information , 2005 .

[48]  R. Nicholls,et al.  Future flood losses in major coastal cities , 2013 .

[49]  G. Blöschl,et al.  The influence of non-stationarity in extreme hydrological events on flood frequency estimation , 2016 .

[50]  G. Villarini,et al.  Flood peak distributions for the eastern United States , 2009 .

[51]  V. Singh,et al.  Non-stationary approach to at-site flood frequency modelling I. Maximum likelihood estimation , 2001 .

[52]  J. Rolf Olsen,et al.  Climate Change and Floodplain Management in the United States , 2006 .

[53]  Jasper A. Vrugt,et al.  Markov chain Monte Carlo simulation using the DREAM software package: Theory, concepts, and MATLAB implementation , 2016, Environ. Model. Softw..

[54]  Mauro Naghettini,et al.  On some aspects of peaks-over-threshold modeling of floods under nonstationarity using climate covariates , 2015, Stochastic Environmental Research and Risk Assessment.

[55]  T. Phillips,et al.  Non-stationary return levels of CMIP5 multi-model temperature extremes , 2015, Climate Dynamics.

[56]  Elena Volpi,et al.  Sworn testimony of the model evidence: Gaussian Mixture Importance (GAME) sampling , 2017 .

[57]  S. Kanae,et al.  Global flood risk under climate change , 2013 .

[58]  J. Stedinger,et al.  Getting From Here to Where? Flood Frequency Analysis and Climate 1 , 2011 .

[59]  David R. Easterling,et al.  Probable maximum precipitation and climate change , 2013 .

[60]  Bruno Merz,et al.  A Probabilistic Modelling System for Assessing Flood Risks , 2006 .

[61]  Upmanu Lall,et al.  A hierarchical Bayesian regional model for nonstationary precipitation extremes in Northern California conditioned on tropical moisture exports , 2015 .

[62]  Jan Picek,et al.  Estimating extremes in climate change simulations using the peaks-over-threshold method with a non-stationary threshold , 2010 .

[63]  W. E. Watt,et al.  Assessment of fitting techniques for the log Pearson type 3 distribution using Monte Carlo Simulation , 1979 .

[64]  Bradley P. Carlin,et al.  Bayesian measures of model complexity and fit , 2002 .

[65]  W. Kirby,et al.  Computer‐oriented Wilson‐Hilferty transformation that preserves the first three moments and the lower bound of the Pearson type 3 distribution , 1972 .

[66]  J. Stedinger,et al.  Generalized maximum‐likelihood generalized extreme‐value quantile estimators for hydrologic data , 2000 .

[67]  G. Hegerl,et al.  Human contribution to more-intense precipitation extremes , 2011, Nature.

[68]  D. Gamerman,et al.  Bayesian analysis of extreme events with threshold estimation , 2004 .

[69]  T. Karl,et al.  Global climate change impacts in the United States. , 2009 .

[70]  Jery R. Stedinger,et al.  Incorporating Climate Change and Variability into Bulletin 17B LP3 Model , 2007 .

[71]  Demetris Koutsoyiannis,et al.  Modeling and mitigating natural hazards: Stationarity is immortal! , 2014 .

[72]  Jasper A. Vrugt,et al.  High‐dimensional posterior exploration of hydrologic models using multiple‐try DREAM(ZS) and high‐performance computing , 2012 .

[73]  D. Moorhead,et al.  Increasing risk of great floods in a changing climate , 2002, Nature.

[74]  D. Rubin,et al.  Inference from Iterative Simulation Using Multiple Sequences , 1992 .

[75]  B. Renard,et al.  An application of Bayesian analysis and Markov chain Monte Carlo methods to the estimation of a regional trend in annual maxima , 2006 .

[76]  Upmanu Lall,et al.  Floods in a changing climate: Does the past represent the future? , 2001, Water Resources Research.

[77]  J. M. Landwehr,et al.  Hydro-climatic data network (HCDN); a U.S. Geological Survey streamflow data set for the United States for the study of climate variations, 1874-1988 , 1992 .

[78]  R. Katz,et al.  Non-stationary extreme value analysis in a changing climate , 2014, Climatic Change.

[79]  Bruno Merz,et al.  Projecting flood hazard under climate change: an alternative approach to model chains , 2013 .