Introduction

1.1 NEED FOR ASSESSING CLIMATE CHANGE IMPACTS ON URBAN DRAINAGE

For more than a century, large-scale separate and combined sewer systems have been constructed across many cities worldwide. As the name suggests, combined sewer systems convey both urban runoff and sewage in the same (combined) pipe drainage system. This is the most common type of urban drainage system in Western Europe and North American regions. The alternative solution is a separate system, which consists of parallel sewers for storm and waste water (e.g. Burian et al. 1999; Butler & Davies, 2010). Separate systems are widely used in many countries in Asia, Australia, Europe and North America for newly developed urban areas. In separate sytems, sewage is conveyed in smaller diameter pipe systems while urban runoff is conveyed separately, usually in either open channels or street pipe drainage systems. They are built to reduce the pollution effect of urban drainage on receiving waters, and to enhance the efficiency of the wastewater treatment plant (less diluted wastewater). For instance, in Japan separate systems are only constructed since the 1980s and currently about 20% of the sewer systems are of the combined type. This percentage of combined systems is much higher in Europe, for example about 70% in the UK (Butler & Davies, 2010). For clarity in this book, both combined and separate systems will henceforth be referred to as urban drainage systems.

In general, these urban drainage systems have reduced the vulnerability of the cities to the health risks since they are often built as part of municipal sanitation programs. However, the installation of these systems could make them more vulnerable to rainfall extremes, partly due to the lack of consideration to what occurs when the design criteria are exceeded. In particular, urban land use is constantly changing in response to the continuous changes in demographic and socio-economic conditions of the population (O'Loughlin et al. 1995). As a consequence of these environmental changes, designers and managers must now cope with the increase in surface imperviousness and the shorter response time of urban catchments, which boost stormwater runoff volumes and velocities beyond the capacity of existing drainage systems.

For most cities, it is expected that these trends will continue over the coming decades. At the same time, many highly developed regions already realise that their urban design and planning processes urgently need to incorporate more sustainable approaches. Many urban water systems are particularly vulnerable to rapid population growth and climate change (Semadeni-Davies et al. 2008). In the presence of climate change induced uncertainty, urban water systems need to be more resilient and multi-sourced. This is partly because of decreasing volumetric rainfall trends in many parts of the world, which might have severe effects on reservoir yields and operational practices. In addition, severe intensity rainfall events can cause failure of drainage system capacity and subsequent urban flood inundation problems (Beecham & Chowdhury, 2012).

Besides this increased vulnerability, there is also strong evidence that the probabilities and risks of urban flooding and sewer surcharge are changing due to the increasing trends of some climatic parameters such as precipitation and temperature extremes (Stone et al. 2000; Alexander et al. 2006; Allan & Soden, 2008). In particular, in their Fourth Assessment Report (AR4) the Intergovernmental Panel on Climate Change (IPCC) of the World Meteorological Organization and the United Nations Environment Program reports for the late 20th century a worldwide increase in the frequency of extreme rain storms as most likely a result of global warming (IPCC, 2007a; WMO, 2009a; Giorgi et al. 2011). Extremes were by the IPCC (2007a) defined as events that are relevant from a disaster risk management perspective, for example urban flood disasters. The increase in rainfall extremes is most pronounced in the period of anthropogenic greenhouse gas (GHG) induced twentieth-century warming (approximately 0.5 deg. C worldwide in the period 1976–2000) after the so-called climate shift (IPCC, 2007a). The study by Min et al. (2011) revealed that human-induced increases in GHG have contributed to the observed intensification of heavy rainfall events over approximately two-thirds of the data-covered parts of the Northern Hemisphere land areas. Based on climate model simulations with different future GHG emission scenarios, IPCC (2007a) furthermore concluded that it is very likely that this trend will continue in the 21st century. The consequences of these changes have to be assessed in a perspective of sustainable development. Water managers have to anticipate these changes in order to limit flood risks for communities. Also the insurance industry, as well as the various water users and policy makers, need quantification of these risks so as to develop and adapt policies.

Consequently, the number of hydrological impact studies of climate change has increased greatly in recent years. These studies, however, most often focus on river discharge extremes and low flow risks. The number of climate change studies dealing with urban drainage impacts is still rather limited, partly because they require a specific focus on small urban catchment scales (normally on a scale of 1–10 km2) and short duration precipitation extremes (normally less than 1 hour). This is because of the small characteristic time scales of the processes involved in the hydrological cycle within urban areas. These processes react very quickly to rainfall.

Despite a significant increase in computational power in recent years, the spatial resolution of climate models still remains relatively coarse and they are therefore unable to resolve significant climate features relevant at the fine scales of urban drainage systems. They also have limitations in the accuracy with which they describe precipitation extremes (e.g. high-intensity convective storms leading to urban flooding). This is due to an incomplete knowledge and inadequate description of the complex nonlinear and dynamical phenomena during a convective storm leading to the most extreme events on a local scale. As such, the climate model results cannot be used directly for providing an adequate assessment of the impacts of future climate change on urban hydrological processes, which is usually undertaken through simulation with urban hydrological and sewer system models. This poses strong challenges to the urban drainage impact modeller.

1.2 OVERVIEW OF CLIMATE CHANGE IMPACT ASSESSMENT FOR URBAN DRAINAGE

Evaluating regional impacts on urban drainage from possible future climate change requires a methodology to estimate extreme and short-duration rainfall statistics for the time period and the geographical region of interest. In general, two physical systems are involved: the climate system and the urban drainage system (Figure 1.1). Climate models can simulate the effects of climate forcing scenarios such as changes in GHG emissions or GHG concentrations in the atmosphere, for example due to anthropogenic activities, on the climate system. Various types of climate models – global (GCM) and/or regional (RCM) – can be used, providing climate system outputs (climatic variables including extreme rainfall). As GCMs and RCMs are effectively deterministic models of atmospheric processes, they calculate a single value of climate variables at each time step and for each grid cell.

Also urban drainage impact models can take several forms and these can require different rainfall inputs. Most common is the use of simulation models in which rainfall inputs are translated into discharge (either long-term rainfall series or design storms) (Butler & Davies, 2010). Other models are (semi-)probabilistic where probability distributions of urban runoff discharge are calculated based on the rainfall input distribution (e.g. Bacchi et al. 2008). Some urban drainage models account for evapotranspiration (mainly important where vegetated areas are considered) and temperature (mainly for snow melt calculations), but these inputs are generally of secondary importance in comparison with the rainfall input.

For historical periods, the results of the climate models can be validated based on historical observations. Also based on historical records, climate change effects can be investigated by analysing trends in available series (i.e. long-term rainfall series). This is termed "empirical analysis" in Figure 1.1. For future conditions, simulation models for both the climate system and the urban drainage system are needed. Changes simulated in the climate system output (rainfall) due to (anthropogenic GHG) climate forcing need to be transferred to changes in the urban drainage model inputs.

The changes imposed by the climate forcing should be compared to the inherent natural variation of precipitation. The rainfall generating processes occur over temporal scales ranging from multi-decadal to sub-minute resolutions with corresponding changes in spatial scales. Therefore great care should be taken when analysing outcomes of historical and simulated precipitation series, as trends in time series of 20 to 40 years can be due to natural variation rather than a change in precipitation patterns. As such the assumption of inter-annual independence is clearly violated.

Another important feature of GCMs and RCMs is their spatial and temporal resolution. Section 1.1 already highlighted that this resolution is too coarse for urban drainage applications. Consequently all applications of impacts of climate change in urban drainage must make assumptions about how anticipated future precipitation patterns will impact at the urban catchment temporal and spatial scales. These scales are related to the area of the urban catchment (which is typically limited to the size of a town, city or district). Due to the limited area, the relevant temporal scales are generally short and are controlled by the concentration time of the urban drainage system (the time the rainwater needs to move from the most remote location in the urban catchment to the impact location of interest; Chow, 1964; Chow et al. 1988). This means that rainfall information is needed with time steps smaller or equal to the smallest concentration time in the system. To bridge the gap between the climate model scales and the local urban drainage scales and to account for the inaccuracies in describing precipitation extremes, downscaling techniques and bias correction methods are required. GCM projections can be downscaled by using a higher resolution RCM nested within a GCM, called dynamical downscaling. Statistical downscaling relates large-scale climate variables to local scale climate using empirical-statistical relationships. Traditionally, statistical downscaling of GCM projections has been considered, but in recent years statistical downscaling methods that optimally combine dynamic and statistical downscaling have been developed.

The changes in downscaled local short-duration rainfall extremes then can be assessed and transferred to changes in the inputs for urban hydrological impact models. The models were calibrated based on historical rainfall data, which usually take the form of design rain storms or full rainfall time series. These rainfall data will be changed according to the results obtained from the downscaled climate model projections. Finally, the changes in impact results between the today's climate and the climate change scenarios are to be assessed. If long time series of observations are available for the impact variables (i.e. sewer runoff flows, flood frequencies, sewer overflow frequencies), impact assessment can also be done after trend analysis on the series ("empirical analysis" in Figure 1.1). However, this trend analysis cannot go beyond the period covered by the historical observations. While the model-based impact assessment can look into future trends.

It is important to be aware of the uncertainties introduced at each stage of the process. When attempting to make a future projection, as opposed to a hypothetical scenario or numerical experiment, the uncertainty begins with the need to arbitrarily choose a climate forcing scenario, and this initial uncertainty is then compounded by further modelling variability all the way to the final (urban) catchment-scale projection. The uncertainty introduced at each step comes from several sources, such as natural variability, physical parameterisations of the models, and the lack of process descriptions (known or unknown) that are important for modelling climate change.

The above methodology outlines the impact assessment of climate change to urban drainage, focusing on its main driver, namely the changes in short-duration rainfall statistics. However, many other drivers affect the performance of an urban drainage system, particularly urbanization and changes in urban drainage management and planning. This book focuses on estimation of climate change impacts on urban drainage, but the reader has to be aware that these other drivers might be as important. For example, urbanization and associated increasing population can lead to a significant increase in water use and increased impermeable areas. Urban areas might also be affected by other types of climate change impacts such as sea level rise and increase in river flood frequency. When combined, these changes could have impacts that are more significant than those caused by changes in short-duration rainfall extremes only and/or due to climate change only. Also note that urban water management practices are likely to improve into the future, and this might potentially offset some or all of the negative impacts.

1.3 SCOPE AND LIMITATIONS

This book aims to present a state-of-the-art review of climate change impact assessment in the field of urban drainage. More specifically, the objectives of the book are:

– To give an overview of current practices with respect to rainfall analysis and modelling for urban drainage simulations;

– To review trend analyses in historical urban rainfall extremes;

– To introduce the basic concepts of atmospheric modelling;

– To describe the fundamentals of dynamical and statistical downscaling of rainfall;

– To review evaluations of downscaled rainfall;

– To review expected future changes in urban rainfall extremes and the corresponding impacts on urban drainage;

– To give an overview of adaptation issues, principles and methods; and

– To provide practical tools and instructions.

The book provides on the one hand a review of methods and difficulties concerning the assessment of climate change impacts on urban rainfall extremes and urban drainage systems. On the other hand, it provides a practical and useful guide on these methods. The audience of the book is therefore not only scientists, but also practitioners (urban drainage engineers, urban planners) and students.

While the book aims to give a representative overview of current knowledge, practices and challenges associated with climate change impact investigations in the field of urban drainage and rainfall extremes, the authors are aware of some limitations:

– Because of its focus on urban drainage, this book mainly focuses on extreme rainfall at small (sub-daily) time scales. However, some references to investigations using daily time scales are also provided.

– The book presents many case studies based mainly on European conditions, but also from other continents, such as North America (USA and Canada), Asia and Australia.

– While the scope of this book is extensive, it has not been possible to cover all investigations and research papers. We have relied on what we view as important scientific contributions but we would appreciate any feedback regarding significant omissions for potential inclusion in future editions.

– The authors are aware that climate science evolves very rapidly, which means that new knowledge and methodologies might have become available after the date that the book manuscript was delivered to the publisher. The authors therefore recognise that future updates will be required. Readers are invited to send their additions and comments to the authors. Text updates will be provided together with the electronic supplement through the IWA Water Wiki that accompanies this book.

1.4 BOOK OUTLINE

The nine chapters of this book discuss the various aspects and steps involved in climate change impact investigations in the field of urban hydrology, as outlined in Section 1.2.