Abstracts

Detection of trends in hydrological extremes for Canadian watersheds

Donald H. Burn, Mohammed Sharif and Kan Zhang

Abstract: The potential impacts of climate change can alter the risk to critical infrastructure resulting from changes to the frequency and magnitude of extreme events. As well, the natural environment is affected by the hydrologic regime, and changes in high flows or low flows can have negative impacts on ecosystems. This article examines the detection of trends in extreme hydrological events, both high and low flow events, for streamflow gauging stations in Canada. The trend analysis involves the application of the Mann–Kendall non-parametric test. A bootstrap resampling process has been used to determine the field significance of the trend results. A total of 68 gauging stations having a nominal record length of at least 50 years are analysed for two analysis periods of 50 and 40 years. The database of Canadian rivers investigated represents a diversity of hydrological conditions encompassing different extreme flow generating processes and reflects a national scale analysis of trends. The results reveal more trends than would be expected to occur by chance for most of the measures of extreme flow characteristics. Annual and spring maximum flows show decreasing trends in flow magnitude and decreasing trends in event timing (earlier events). Low flow magnitudes exhibit both decreasing and increasing trends.

Probabilistic design of a riverine early warning source water monitoring system1

Heather P. Sim, Donald H. Burn, and Bryan A. Tolson

 Abstract: Source water protection involves safeguarding water supplies from contamination and depletion. Despite best efforts, spills cannot always be prevented from entering a source water body. However, many spills can be prevented from entering a drinking water treatment plant if an early warning source water monitoring station is used. These stations provide downstream water utilities with advanced notification of spills so the utilities have time to implement their responses. This paper addresses the design of an early warning monitoring station for a riverine source of drinking water. Riverine water supplies face many threats related to accidental spills, which are inherently uncertain in nature. Therefore, designing a monitoring station for the detection of these events requires a probabilistic modelling approach. The design objectives include maximizing the probabilities of detection and of having a threshold amount of warning time. The methodology is applied to a water supply intake on the Grand River in southern Ontario.

Homogeneity testing: How homogeneous do heterogeneous cross-correlated regions seem?

A. Castellarin, D.H. Burn and A. Brath

Abstract: The homogeneity of the flood frequency regime for a given pooling-group of sites is a fundamental assumption for many regional flood frequency analysis techniques. Assessing regional homogeneity is a critical step, which may be complicated by the presence of cross-correlation among flood sequences. The scientific literature proposes a number of statistical homogeneity tests and documents that inter-site correlation of floods is normally not negligible, but does not specifically address the impact of cross-correlation on such statistical tests. This paper analyzes the effectiveness of a well-known homogeneity test proposed in the scientific literature in the presence of inter-site cross-correlation through a series of Monte Carlo experiments. The numerical experiments enable us to comment on a possible theoretical correction for the test and to identify an empirical tool that accounts for the impact of inter-site cross-correlation of floods.

The Processes, Patterns and Impacts of Low Flows Across Canada

Donald H. Burn, James M. Buttle, Daniel Caissie, Greg MacCulloch, Chris Spence and Kerstin Stahl

Abstract: This paper provides an overview of low flow characteristics for six regions of Canada: the Arctic; the Mountains; the Prairies; southern Ontario; the Canadian Shield and the Atlantic. Processes that influence low flows are contrasted between the six regions examined. Data from a common analysis period for 51 gauging stations are used to evaluate flow duration curves and to explore the relationship between low flows and drainage area. The results reveal a diversity of processes influencing low flows and illustrate important regional differences in low flow characteristics and the impacts associated with low flows.

Identification and quantification of streamflow trends on the Canadian Prairies

Donald H. Burn, Lin Fan & Gordon Bell

Abstract: Trend analysis was performed on streamflow data for a collection of stations on the Canadian Prairies, in terms of spring and summer runoff volumes, peak flow rates and peak flow occurrences, as well as an annual volume measure, for analysis periods of 1966–2005, 1971–2005, and 1976–2005. The Mann-Kendall statistical test for trend and bootstrap resampling were used to identify the trends and to determine the field significance of the trends. Partial correlation analysis was used to identify relationships between hydrological variables that exhibit a significant trend and meteorological variables that exhibit a significant trend. Noteworthy results include decreasing trends in the spring snowmelt runoff event volume and peak flow, decreasing trends (earlier occurrence) in the spring snowmelt runoff event peak date and decreasing trends in the seasonal (1 March–31 October) runoff volume. These trends can be attributed to a combination of reductions in snowfall and increases in temperatures during the winter months.

Climatic influences on streamflow timing in the headwaters of the Mackenzie River Basin

Donald H. Burn  

Abstract: The Mann–Kendall non-parametric test for trend is used to explore the trend behaviour of nine measures of the timing of runoff. The relationship between trends in timing measures and trends in meteorological variables are investigated using partial correlation analysis. The relationships between six climate indices and trends in the timing measures are also examined. The analysis is conducted for 26 streamflow gauging stations from three sub-watersheds of the Mackenzie River Basin in northern Canada. The results reveal that for several of the timing measures, many more trends are identified than can be expected to occur by chance. The spring freshet is observed to be occurring earlier with this timing shift appearing particularly strong in headwater catchments. Based on the partial correlation analysis, it is plausible to attribute some of the observed trends to trends in meteorological variables. Timing of runoff is affected by the Pacific Decadal Oscillation (PDO), the North Atlantic Oscillation (NAO), the North Pacific (NP) index and the Atlantic Multidecadal Oscillation (AMO) but not by the El Nin˜o-Southern Oscillation (ENSO) or the Arctic Oscillation (AO).

Trends in evaporation for the Canadian Prairies

Donald H. Burn and Nicole M. Hesch 

Abstract: Trend analyses of evaporation data were conducted for 48 sites on the Canadian Prairies for three analysis periods. Significant trends were identified using the Mann–Kendall statistical test for trend and a bootstrap resampling technique. Trends in calculated evaporation were compared with trends in input variables used to calculate evaporation for all sites and with trends in pan evaporation for four sites. June, July, August, October and warm season evaporation revealed significant trends that were mainly decreasing. The longest analysis period identified an increasing trend in April. Increasing trends were typically in the more northern regions and decreasing trends in the more southern regions. Examining causal mechanisms for evaporation revealed that wind speed had more of an influence on decreasing trends and vapour pressure deficit had more of an influence on increasing trends.

Improved K-Nearest Neighbour Weather Generating Model

Mohammed Sharif and Donald H. Burn

Abstract: A major limitation of K-nearest neighbor based weather generators is that they do not produce new values but merely reshuffle the historical data to generate realistic weather sequences. In this paper, a modified approach is developed that allows nearest neighbour resampling with perturbation of the historic data. A strategy is introduced that resamples the historical data with perturbations while preserving the prominent statistical characteristics, including the interstation correlations. The approach is similar in spirit to traditional autoregressive models except that the new values are obtained by adding a random component to the individual resampled data points. An advantage of the approach is that unprecedented precipitation amounts are generated that are important for the simulation of extreme events. The approach is demonstrated through application to the Upper Thames River Basin in Ontario. Daily weather variables _maximum temperature, minimum temperature, and precipitation_ were simulated at multiple stations in and around the basin. Analysis of the simulated data demonstrated the ability of the model to reproduce important statistical parameters of the observed data series while allowing perturbations to the observed data points. Additionally, no site-specific assumptions regarding the probability distribution of variables are required.

A comparison of trends in potential and pan evaporation for the Canadian prairies.

Donald H. Burn and Nicole M. Hesch

Abstract: Comparisons are made between potential evaporation trends and pan evaporation trends for a collection of sites on the Canadian Prairies. Trends are analyzed using the non-parametric Mann-Kendall test. There are many similar trend results, but most of these indicate no trend for either measure of evaporation. Seven cases (out of 53 in total) show matching significant trends, six of which are decreasing trends. One case shows opposing significant trends with an increasing pan evaporation trend and a decreasing potential evaporation trend. Despite the opposing trends, time series plots of pan and potential evaporation show similar timing of maximum and minimum values. Wind speed exerting an influence on potential evaporation and not on pan evaporation was the most common explanation for discrepancies between pan and potential evaporation trends.

Simulating climate change scenarios using an improved K-nearest neighbor model

Mohammed Sharif, Donald H Burn

Abstract: An improved weather-generating model that allows nearest neighbour resampling with perturbation of the historic data is applied to generate weather data based upon plausible climate scenarios. The intent is to create an ensemble of climate scenarios that can be used for the assessment of the vulnerability of a watershed to extreme events, including both floods and droughts. Analysis of the results clearly indicates that the model adequately simulated extreme unprecedented events for five different climate change scenarios. Based on the simulation results, the increasing precipitation scenario is identified as the critical scenario for the assessment of risks associated with the occurrence of floods in the basin. The increasing temperature scenario appears to be the critical scenario for the analysis of droughts in the basin. Frequency analysis was carried out to determine the impact of potential climatic change on the occurrence of storm depths of any given magnitude. A promising potential application of the model is in rainfall-runoff modelling where the storms depths could be related to the occurrence of extreme events in the basin. The proposed model, in conjunction with a rainfall-runoff model, has the potential of providing valuable aid in developing efficient management strategies for a watershed. The model produces spatially correlated data, which is crucial for accurate runoff estimation. Although the model is applied to the Upper Thames River Basin in the Canadian province of Ontario, it is generic and transportable to any other watershed with minimal changes.

Long-lead probabilistic forecasting of streamflow using ocean-atmospheric and hydrological predictors

Shahab Araghinejad, Donald H. Burn, and Mohammad Karamouz

Absract: A geostatistically based approach with a local regression method is used to predict the magnitude of seasonal streamflow using ocean-atmospheric signals and the hydrological condition of a basin as predictors. The model characterizes the stochastic behavior of a forecast variable by generating a conditional distribution of the predicted value for different hydroclimatic conditions. The correlation structure between dependent and independent variables is represented by the variography of the predicted values in which the distance variable in the variogram is determined by measuring the distance between the predictors. This variogram in a virtual field constructed from the predictors makes it possible to predict variables as unmeasured points while considering historic information as measurement points of the field. Different types of kriging, as well as a generalized linear model regression, are used to predict data in interpolation and extrapolation modes. The forecast skill is evaluated using a linear error in probability space score for different combinations of predictors and different kriging methods. The method is applied to a case study of the Zayandeh-rud River in Isfahan, Iran. The utility of the method is demonstrated for forecasting autumn-winter and spring streamflow using the Southern Oscillation Index, the North Atlantic Oscillation, serial correlation between seasonal streamflow series, and the snow budget. The study analyzes the application of the proposed method in comparison with a K-nearest neighbor regression method. The results of this study show that the proposed method can significantly improve the long-lead probabilistic forecast skill for a nonlinear relationship between hydroclimatic predictors and streamflow in a region.

Switching the pooling similarity distances: Mahalanobis for Euclidean

Juraj M. Cunderlik and Donald H. Burn

Abstract: In recent years, catchment similarity measures based on flood seasonality have become popular alternatives for identifying hydrologically homogeneous pooling groups used in regional flood frequency analysis. Generally, flood seasonality pooling measures are less prone to errors and are more robust than measures based on flood magnitude data. However, they are also subject to estimation uncertainty resulting from sampling variability. Because of sampling variability, catchment similarity in flood seasonality can significantly deviate from the true similarity. Therefore sampling variability should be directly incorporated in the pooling algorithm to decrease the level of pooling uncertainty. This paper develops a new pooling approach that takes into consideration the sampling variability of flood seasonality measures used as pooling variables. A nonparametric resampling technique is used to estimate the sampling variability for the target site, as well as for every site that is a potential member of the pooling group for the target site. The variability is quantified by Mahalanobis distance ellipses. The similarity between the target site and the potential site is then assessed by finding the minimum confidence interval at which their Mahalanobis ellipses intersect. The confidence intervals can be related to regional homogeneity, which allows the target degree of regional homogeneity to be set in advance. The approach is applied to a large set of catchments from Great Britain, and its performance is compared with the performance of a previously used pooling technique based on the Euclidean distance. The results demonstrate that the proposed approach outperforms the previously used approach in terms of the overall homogeneity of delineated pooling groups in the study area.

A Case Study of Climate Change Impacts on Navigation on the Mackenzie River

Rudy Y.-J. Sung, Donald H. Burn and Eric D. Soulis

Abstract: The impacts of climate change on the Mackenzie River are examined through a case study that considers navigational requirements for barge transportation. Climate change scenarios are used to estimate projected changes in streamflow, and potential changes in water levels, through the use of a hydrological model and a hydraulic model. The climate change scenarios suggest an earlier spring freshet and increased fall flow while some scenarios suggest lower flows in the summer period. From a navigational perspective, the climate change scenarios are generally favourable, indicating an extended navigation season, in comparison to the base conditions.

Trends and variability in the hydrological regime of the Mackenzie River Basin

Omar I. Abdul Aziz, Donald H. Burn

Abstract: Trends and variability in the hydrological regime were analyzed for the Mackenzie River Basin in northern Canada. The procedure utilized the Mann–Kendall non-parametric test to detect trends, the Trend Free Pre-Whitening (TFPW) approach for correcting time-series data for autocorrelation and a bootstrap resampling method to account for the cross-correlation structure of the data. A total of 19 hydrological and six meteorological variables were selected for the study. Analysis was conducted on hydrological data from a network of 54 hydrometric stations and meteorological data from a network of 10 stations. The results indicated that several hydrological variables exhibit a greater number of significant trends than are expected to occur by chance. Noteworthy were strong increasing trends over the winter month flows of December to April as well as in the annual minimum flow and weak decreasing trends in the early summer and late fall flows as well as in the annual mean flow. An earlier onset of the spring freshet is noted over the basin. The results are expected to assist water resources managers and policy makers in making better planning decisions in the Mackenzie River Basin.

Site-focused nonparametric test of regional homogeneity based on flood regime

Juraj M. Cunderlik , Donald H. Burn

Abstract: In regional flood frequency analysis, the proper identification of homogeneous pooling groups is a key factor for reliable quantile estimation. A new test of regional homogeneity based on flood regime is proposed in this paper. In contrast to other tests, this test does not use flood magnitude data and is focused on a site of interest and a target return period. The test relies on a nonparametric comparison of the regional variability of local flood regimes within a given pooling group with the variability that could be expected in a theoretically homogeneous group with the same hydrological properties as the original group. The performance of the test in the detection of regional heterogeneity is evaluated in a set of simulation experiments, each with different pooling group size, record length, and degree of regional heterogeneity. The test is then applied to a large set of real data, and the pooling measures and quantile errors are compared with the results obtained from the Hosking and Wallis’s (HW) homogeneity test included in the same pooling algorithm. The simulation results demonstrated that the power of the test rapidly improves as the record length, pooling size, and degree of regional heterogeneity increase. In our study area the new test rejected fewer similar sites and led to similar quantile errors in comparison with the approach based on the HWtest. The test is a sound alternative to the existing tests of regional homogeneity and is particularly convenient for site-focused pooling algorithms that require homogeneity testing at each step of the pooling.

Use of tree ring reconstructed streamflows to assess drought

David V. Bonin and Donald H. Burn

Abstract: The reconstruction of past streamflow events is of great interest to the water resources engineer to obtain the best possible estimates of extreme flow conditions for investment, decision making, and design. The tree ring data offer a unique way of addressing this problem. The pattern of growth rings of a tree reflects the environmental conditions experienced during each year. Tree rings are produced annually and can be precisely and reliably linked to climatic variations, which makes them ideal for correlation with annual climatic records. This paper demonstrates the utility of using the methods of dendroclimatology, the study of climate through tree rings, to extend streamflow records. The techniques developed were applied to the Athabasca River at Athabasca. The results reveal considerable benefits from the reconstruction through more precise, and more extreme, estimates of drought quantiles.

Probabilistic forecasting of hydrological events using geostatistical analysis

Shahab Araghinejad & Donald H. Burn

Abstract A method is introduced for probabilistic forecasting of hydrological events based on geostatistical analysis. In this method, the predictors of a hydrological variable define a virtual field such that, in this field, the observed dependent variables are considered as measurement points. Variography of the measurement points enables the use of the system of kriging equations to estimate the value of the variable at non-measured locations of the field. Non-measured points are the forecasts associated with specific predictors. Calculation of the estimation variance facilitates probabilistic analysis of the forecast variables. The method is applied to case studies of the Red River in Manitoba, Canada and Karoon River in Khoozestan, Iran. The study analyses the advantages and limitations of the proposed method in comparison with a K-nearest neighbour approach and linear and nonlinear multiple regression. The utility of the proposed method for forecasting hydrological variables with a conditional probability distribution is demonstrated.

A Comparison of Trends in Hydrological Variables for Two Watersheds in the Mackenzie River Basin†

Donald H. Burn, Omar I. Abdul Aziz and Alain Pietroniro

Abstract: A study of trends and variability of hydrological variables was conducted for natural streamflow gauging stations within two watersheds that are important sources of flow within the Mackenzie River Basin. A comparison was made between trend results for the Liard River Basin and for the Athabasca River Basin. These basins represent a north-south transect of high elevation headwater basins within the Mackenzie River system and are significant since they produce 34% of the annual flow, while occupying only 24% of the total drainage area. Trend analysis was conducted using the Mann-Kendall test with an approach that corrects for serial correlation. The global (or field) significance of the results for each watershed was evaluated using a bootstrap resampling approach. The relationships between trends in hydrological variables and trends in meteorological variables were investigated using partial correlation analysis. The results reveal more trends in some hydrological variables than are expected to occur by chance. In general, both basins exhibit an increase in winter flows and some increase in spring runoff. These increased flows are somewhat offset by decreases (not field significant) in summertime flow. Almost 50% of the stations used in the analysis show an increasing trend in annual minimum flows. Other differences in trend responses are noted for the two watersheds and possible explanations for the differences are hypothesized.

An integrated approach to the estimation of streamflow drought quantiles

Donald H. Burn, Jeremy Wychreschuk & David V, Bonin

Abstract: An approach was developed for combining streamflow drought information from synthetic (generated) data with data reconstructed based on palaeoclimatic information (tree ring widths). The tree ring data were used to reconstruct streamflow in periods when no streamflow data were collected. The reconstructed data were then used as a source of historical data for estimating drought severity quantiles. The generated data were obtained using a nearest neighbour resampling method while the tree ring reconstruction was accomplished using a regression model. The application of the approach was to data from the Athabasca River in Alberta, Canada. The results demonstrate the feasibility and the utility of the approach for obtaining more accurate and precise estimates of extreme drought severity quantiles.

Artificial neural network ensembles and their application in pooled flood frequency analysis

Chang Shu and Donald H. Burn

Abstract: Recent theoretical and empirical studies show that the generalization ability of artificial neural networks can be improved by combining several artificial neural networks in redundant ensembles. In this paper, a review is given of popular ensemble methods. Six approaches for creating artificial neural network ensembles are applied in pooled flood frequency analysis for estimating the index flood and the 10-year flood quantile. The results show that artificial neural network ensembles generate improved flood estimates and are less sensitive to the choice of initial parameters when compared with a single artificial neural network. Factors that may affect the generalization of an artificial neural network ensemble are analyzed. In terms of the methods for creating ensemble members, the model diversity introduced by varying the initial conditions of the base artificial neural networks to reduce the prediction error is comparable with more sophisticated methods, such as bagging and boosting. When the same method for creating ensemble members is used, combining member networks using stacking is generally better than using simple averaging. An ensemble size of at least 10 artificial neural networks is suggested to achieve sufficient generalization ability. In comparison with parametric regression methods, properly designed artificial neural network ensembles can significantly reduce the prediction error.

Wavelet analysis of variability in annual Canadian streamflows

Paulin Coulibaly and Donald H. Burn

Absract:  Wavelet analysis is used to identify and describe variability in annual Canadian streamflows and to gain insights into the dynamical link between the streamflows and the dominant modes of climate variability in the Northern Hemisphere. Results from applying continuous wavelet transform to mean annual streamflows from 79 rivers selected from the Canadian Reference Hydrometric Basin Network (RHBN) reveal striking climaterelated features before the 1950s and after the 1970s. The span of available observations, 1911–1999, allows for depicting variance for periods up to 12 years. Scale-averaged wavelet power spectra are used to simultaneously assess the interannual and spatial variability in 79 annual streamflow time series. The most striking feature, in the 2–3 year band and in the 3–6 year band (the 6–12 year band is dominated by white noise (since 1950) and is not considered further) is a net distinction between the timing and intensity of the interannual variability in western, central, and eastern streamflows, which is shown to be linked to the regional climatology. It is found that for the 2–3 year band, the Canadian streamflows are depicted mainly by the Pacific North America (PNA) during 1950–1999, and the Northern Hemisphere Annular Mode (NAM) only prior to 1950, and the North Atlantic Oscillation (NAO) after 1970. Similarly, in the 3–6 year band, the streamflows are depicted mostly by the NAO, the sea surface temperature anomalies over the Nin˜o-3 region (ENSO3) and the PNA during the period 1950–1999, and the NAM prior to 1950. Furthermore, strong local correlations between teleconnection patterns and western, central, and eastern streamflows are also revealed in both the 2–3 and 3–6 year bands with striking changes around 1950 and 1970. The correlation analysis in the 2–3 year and 3–6 year bands revealed the presence of two change points in the west and east streamflows occurring around 1950 and 1970.

Linkages between Regional Trends in Monthly Maximum Flows and Selected Climatic Variables

Juraj M. Cunderlik and Donald H. Burn 

Abstract: The potential impact of climate change on the hydrologic regime is a crucial question for water resources management. This study explores regional trends in monthly maximum flows and their possible linkages to trends in selected climatic variables in a hydroclimatologically homogeneous region. Trends are identified using the Mann-Kendall nonparametric test, with a modification for autocorrelated data. The regional significance of trends identified at the local scale is evaluated by means of a regional bootstrap algorithm. A trend significance index that accounts for both local and regional significance levels is proposed as a convenient tool for quantification and visual comparison of different trend results. The index is also used for identifying potential linkages between trends in hydroclimatic records. The plausibility of identified linkages is explored by means of cross-correlation analysis applied on residuals that are obtained from the original records after subtracting all serially dependent components. An uncertainty in regional trend analysis resulting from different observation periods is presented and quantified by calculating trend significance indices for several scenarios of different locations and lengths of a common observation period shifted on a timescale. The results show significant changes in the intraannual flood regime in the case study area of southern British Columbia. A regionally, strongly significant increase in the spring air temperature shifts the timing of the snowmelt process, resulting in a significant increase in early spring maximum flows and a significant decrease in late spring maximum flows. An autumn decrease in flows is related to increasing air temperature in the preceding summer months, which tends to dry out catchments more intensively, and is also related to precipitation activity in the previous months. The regional trend results are highly sensitive to the location and length of a given regional observation period on a timescale. Possible sources of the uncertainty in a low frequency climatic variability such as the Pacific Decadal Oscillation are discussed.

Hydrological trends and variability in the Liard River basin

Donald H. Burn, Juraj M. Cunderlik and Alain Pietroniro

Abstract: A study of the trends and variability of hydrological variables was conducted for natural streamflow gauging stations within the Liard River basin in northern Canada. Trends were investigated using the Mann-Kendall test, with an approach that corrects for serial correlation. The field significance of the results was evaluated using a bootstrap resampling approach. The relationships between trends in hydrological variables and both meteorological variables and a large-scale oceanic and atmospheric process were investigated using correlation analysis. The results reveal more trends in some hydrological variables than are expected to occur by chance. The observed trends are related to both trends in meteorological variables and a large-scale oceanic and atmospheric process.

The use of resampling for estimating confidence intervals for single site and pooled frequency analysis

Donald H. Burn

Abstract: A balanced resampling approach is presented for estimating confidence intervals for extreme flow quantiles determined from data at a single site. The approach is also adapted to provide resampled estimates for confidence intervals for extreme flow quantiles obtained from pooled frequency analysis. The balanced resampling approach does not require assumptions, in contrast to conventional approaches that are typically based on an asymptotic formula and require a distributional assumption. The approach is demonstrated to provide useful information in the context of both single site and pooled frequency analysis. The application of the approach also demonstrates the benefits of employing a pooled frequency analysis approach for estimating extreme flow quantiles.

A comparison of index flood estimation procedures for ungauged catchments

Patrick L. Grover, Donald H. Burn, and Juraj M. Cunderlik

Abstract: Flood frequency analysis is used by water resources professionals to estimate the probability of exceedence associated with a flood of a given magnitude. The estimation of flood frequencies is important because they are used in the planning and design of hydraulic structures, in flood-plain management, and in reservoir operation. The index flood method is commonly used to develop a flood frequency curve that relates flood magnitude to flood rarity. This method involves scaling a dimensionless flood frequency curve by the index flood. The index flood is a middle-sized flood for which the mean or median of the flood data series is typically used. When the catchment of interest is ungauged, statistical models, such as multiple regression, are often used to relate the index flood to catchment descriptors. In this study six different parameter estimation techniques and three regionalization techniques are compared in terms of ability to predict the index flood for an ungauged catchment. A case study employing a split-sample experiment with data from catchments in Ontario, Canada, was used to evaluate the approaches. The models were assessed using three performance indices to evaluate the capability to predict the index flood for 20 stations. The dimensionless nonlinear model outperformed all of the other parameter estimation techniques for each of the three indices selected. The performance was improved through the use of geostatistical residual mapping, however, the improvement was small. The residual mapping was found to greatly improve the estimates obtained using ordinary least-squares regression.

Local and Regional Trends in Monthly Maximum Flows in Southern British Columbia

Juraj M. Cunderlik and Donald H. Burn

Abstract:  The potential impact of climate change on the hydrologic regime is an important topic in contemporary hydrology. A number of studies have been conducted to identify climate change signals in annual maximum flood records. However less attention has been paid to changes in the regime of flood events of smaller return periods. The overall trend observed in annual maximum flood records gives limited information about intra–annual changes in the hydrological regime of extreme flows. The trends in monthly maximum flows at local and regional scales in a hydro–climatologically homogeneous region from a 40–year common observation period are explored in this study. Trends are identified using the Mann–Kendall non–parametric test with a modification for auto–correlated data. The regional significance of trends identified at the local scale is evaluated by means of a regional bootstrap algorithm that preserves the regional cross–correlation structure. Regional average magnitudes of the trend in monthly maximum flows from a common observation period are estimated over the selected region by means of the ordinary block kriging technique and are compared with results obtained from annual maximum flood records. The results show significant changes in the intra– annual monthly maximum flow regime in southern British Columbia. 

Analysis of the linkage between rain and flood regime and its application to regional flood frequency estimation

Juraj M. Cunderlik and Donald H. Burn

Abstract:  Improving techniques of flood frequency estimation at ungauged sites is one of the foremost goals of contemporary hydrology. River flood regime is a resultant reflection of a composite catchment hydrologic response to flood producing processes. In this sense the process of identifying homogeneous pooling groups can be plausibly based on catchment similarity in flood regime. Unfortunately the application of any pooling approach that is based on flood regime is restricted to gauged sites. Because flood regime can be markedly determined by rainfall regime, catchment similarity in rainfall regime can be an alternative option for identifying flood frequency pooling groups. An advantage of such a pooling approach is that rainfall data are usually spatially and temporary more abundant than flood data and the approach can also be applied at ungauged sites. Therefore in this study we have quantified the linkage between rainfall and flood regime and explored the appropriateness of substituting rainfall regime for flood regime in regional pooling schemes. Two different approaches to describing rainfall regime similarity using tools of directional statistics have been tested and used for evaluation of the potential of rainfall regime for identification of hydrologically homogeneous pooling groups. The outputs were compared to an existing pooling framework adopted in the Flood Estimation Handbook. The results demonstrate that regional pooling based on rainfall regime information leads to a high number of initially homogeneous groups and seems to be a sound pooling alternative for catchments with a close linkage between rain and flood regimes.

The use of flood regime information in regional flood frequency analysis

Juraj M. Cunderlik & Donald H. Burn

Abstract: Understanding the hydro-climatological controls on floods is fundamental for estimating flood frequency. The river flood regime is a reflection of a complex catchment hydrological response to flood producing processes. Hence, the catchment similarity in a flood regime is a feasible basis for identifying flood frequency pooling groups used in regional estimation of design events. This study describes a focused pooling approach that is based on flood regime information. A flood regime descriptor that is sensitive to the modality of the underlying temporal distribution of flood occurrences, and depicts both flood seasonal pattern and flood regularity, was developed and tested. The approach was applied to peaks-over-threshold data from a number of essentially rural sites using a site-focused pooling framework. The relative performance of this approach was evaluated and compared with the performance of a pooling approach based on a previously used flood seasonality measure, using a regional bootstrap resampling technique. The regional bootstrap model was further used for quantifying the sensitivity of the proposed flood regime descriptor to the record length and the length of overlapping period. The results demonstrate that pooling based on the regime index proposed in this study out-performed the pooling based on the previously used seasonality measure in terms of both bias and RMSE of estimated flow quantiles. A detailed description of flood regime captured in the proposed index provides sufficient information for effective regional estimation of extreme flow quantiles for the study area.

Detection of hydrologic trends and variability

Donald H. Burn  and Mohamed A. Hag Elnur

Abstract: This paper describes the development and application of a procedure that identifies trends in hydrologic variables. The procedure utilizes the Mann–Kendall non-parametric test to detect trends, a permutation approach to estimate the test distribution, and accounts for the correlation structure in the data in determining the significance level of the test results. The research investigates 18 hydrologic variables that reflect different parts of the hydrologic cycle. The hydrologic variables are analyzed for a network of 248 Canadian catchments that are considered to reflect natural conditions. A selection of catchments identified to have trends in hydrologic variables is studied further to investigate the presence of trends in meteorological variables and the relationship between the hydrologic and the meteorological response to climatic change. It is concluded that a greater number of trends are observed than are expected to occur by chance. There are differences in the geographic location of significant trends in the hydrologic variables investigated implying that impacts are not spatially uniform.

Flood frequency analysis for the Red River at Winnipeg

Donald H. Burn and N.K. Goel 

Abstract: This paper reviews the flood frequency characteristics of the Red River at Winnipeg. The impacts of persistence in the flood series on estimates of flood quantiles and their associated confidence intervals are examined. This is done by generating a large number of data sequences using a mixed noise model that preserves the short-term an  long-term correlation structures of the observed flood series. The results reveal that persistence in the data series can lead to a slight increase in the expected flood magnitude for a given return period. More importantly, persistence is shown to dramatically increase the uncertainty associated with estimated flood quantiles. The 117-year flood series for the Red River at Winnipeg is demonstrated to be equivalent to roughly 45 years of independent data.

Waste-Load Allocation Using Genetic Algorithms

Donald H. Burn, and Jeanne S. Yulianti 

Abstract: This paper explores the capabilities of genetic algorithms for identifying solutions to waste-load allocation problems. Three optimization model formulations are developed for examining different waste-load allocation problems. Two of the model formulations address problems arising in the planning context while the third model addresses waste-load allocation decisions of use when developing an operational strategy for a river basin. The models can all be used to develop a trade-off relationship for a multiobjective optimization problem that would be of use to a decision maker when selecting a solution for implementation. The models developed are applied to a case study based on the Willamette River in Oregon and the resulting trade-off relationships are discussed both in objective space and in decision space.

Assessing the effectiveness of hydrological similarity measures for Flood frequency analysis

A. Castellarin, D.H. Burn, and A. Brath

Abstract: This paper evaluates the relative performance of four hydrological similarity measures that are used to form homogeneous pooling groups for regional frequency analysis. One pair of similarity measures is based on seasonality indexes that reflect the timing of extreme events. A further pair of measures considers a characterisation, at the basin scale, of the frequency distribution of rainfall extremes and the extent of the impervious portion of the catchment. The measures are applied to a case study encompassing a large area in Northern-Central Italy. The similarity measures are examined in the context of a pooling scheme that is designed to identify hierarchical, focused pooling groups. The performance of the similarity measures is quantified using a Monte Carlo experiment. The results demonstrate that similarity measures based on seasonality indexes are effective for estimating extreme flow quantiles for the study area. For ungauged catchments, a similarity measure incorporating both rainfall statistics and permeability information is most effective.

The formation of groups for regional flood frequency analysis

Donald H. Burn and N.K. Goel

Abstract:  A new technique is developed for identifying groups for regional flood frequency analysis. The technique uses a clustering algorithm as a starting point for partitioning the collection of catchments. The groups formed using the clustering algorithm are subsequently revised to improve the regional characteristics based on three requirements that are defined for effective groups. The result is overlapping groups that can be used to estimate extreme flow quantiles for gauged or ungauged catchments. The technique is applied to a collection of catchments from India and the results indicate that regions with the desired characteristics can be identified using the technique. The use of the groups for estimating extreme flow quantiles is demonstrated for three example sites.

Perceptions of flood risk: A case study of the Red River flood of 1997

Donald H. Burn

Abstract: Risk perception is examined in the context of the Red River flood of 1997. The role that experience with prior flood events plays in risk perception is highlighted as well as the impacts of experience on the mitigation actions selected by individuals. The Red River flood of 1997 demonstrated that individuals with different prior flood Experience could be expected to behave differently during a flood event. This implies that flood warnings should be tailored to the characteristics of the target audience.

Short term streamflow forecasting using artificial neural networks

Cameron M. Zealand, Donald H. Burn, Slobodan P. Simonovic

Abstract: The research described in this article investigates the utility of Artificial Neural Networks (ANNs) for short term forecasting of streamflow. The work explores the capabilities of ANNs and compares the performance of this tool to conventional approaches used to forecast streamflow. Several issues associated with the use of an ANN are examined including the type of input data and the number, and the size of hidden layer(s) to be included in the network. Perceived strengths of ANNs are the capability for representing complex, non-linear relationships as well as being able to model interaction effects. The application of the ANN approach is to a portion of the Winnipeg River system in Northwest Ontario, Canada. Forecasting was conducted on a catchment area of approximately 20 000 km2. using quarter monthly time intervals. The results were most promising. A very close fit was obtained during the calibration (training) phase and the ANNs developed consistently outperformed a conventional model during the verification (testing) phase for all of the four forecast lead-times. The average improvement in the root mean squared error (RMSE) for the 8 years of test data varied from 5 cms in the four time step ahead forecasts to 12.1 cms in the two time step ahead forecasts.

Investigating links between climatic warming and low streamflow in the prairies region of Canada

Jeanne S. Yulianti and Donald H. Burn

Absract: This paper investigates the impacts of temperature change on low streamflow conditions for 77 rivers in the Canadian Prairies. The Mann-Kendall non-parametric statistical test for trend is used to identify trends in the streamflow and temperature time series data, whose starting dates range from 1912 to 1969 with an end date of 1993. The results indicate that the magnitude of flow has a decreasing tendency, while temperature has an increasing tendency. The frequency of low flow events has an increasing tendency. A combination of a decrease in the magnitude, and an increase in the frequency of low flow results in poor water quality conditions in a river, with negative implications for aquatic life. It is realized that difficulties may occur in separating the flow variations due to temperature change from variability caused by other factors, but the results of this study provide information for predicting the variation of low flow due to changes in climatic conditions.

Climate Change Effects on the Hydrologic Regime Within the Churchill-Nelson River Basin

J.R. Westmacott and D.H. Burn

Abstract: This paper evaluates the possible effects of climate change on four hydrologic variables pertaining to the magnitude and timing of the hydrologic events within the Churchill-Nelson River Basin in west-central Canada. By using the Mann-Kendall trend test, and a regionalization procedure, the severity of climactic effects within the river basin may be quantified and used to increase awareness of future consequences for water resource systems planning and management strategies. It was found that the magnitude of hydrologic events decreased over time while snowmelt runoff events occurred earlier. The only exceptions to this behaviour were the spring mean monthly streamflow values which exhibited increasing trends due to the potential for snow melting during this period. The timing of a hydrologic event was found to be influenced to the greatest extent by changes in temperature. Geographically, the decreasing trends were concentrated in the southern regions of the river basin while the increasing trends appeared primarily in the northern regions.

Catchment similarity for regional flood frequency analysis using seasonality measures

Donald H. Burn

Abstract: A regionalization approach that uses information related to the timing of flood events is presented. The approach is applied within the region of influence (ROI) framework and has the advantage of reserving the use of information derived from flood magnitudes for the examination of the homogeneity of flood regions as opposed to first using this information to form regions. The regionalization technique is applied to a set of catchments from the Canadian prairies and is demonstrated to result in the identification of regions that are effective for the estimation of extreme flow quantiles.

Regionalization of Catchments for Regional Flood Frequency Analysis

Donald H. Burn, Zolt Zrinji and Michael Kowalchuk

Abstract: An approach to catchment regionalization is presented, in which an agglomerative hierarchical clustering algorithm is used to define homogeneous regions that can be used for regional flood frequency analysis. Catchment similarity is expressed using seasonality measures derived from the mean date of occurrence of the annual maximum flood and its associated dispersion. Regions that are largely geographically contiguous are obtained by incorporating a distance measure into the similarity metric used within the clustering algorithm. The initial regions formed using the clustering algorithm are subsequently modified in an attempt to enhance the overall regional homogeneity. The approach is demonstrated through an application to a set of 217 catchments from the Saskatchewan-Nelson River basin in west-central Canada.

Spatial characterization of drought events using synthetic hydrology

Donald H. Burn and William J. DeWit

Abstract: This paper describes an approach to spatial drought analysis based on multi-site streamflow synthesis. The approach developed can be used to assist with the quantification of the return period for the drought of record and can also be used to facilitate the identification of design drought events of a specified return period. The approach was applied to the Nelson-Churchill River basin in Manitoba, Canada. For this system, the impacts of droughts on power generation were investigated in addition to a determination of the return period for the drought of record.