Which downscaled rainfall data for climate change impact studies in urban areas? Review of current approaches and trends

Abstract

Changes in extreme precipitation should be one of the primary impacts of climate change (CC) in urban areas. To assess these impacts, rainfall data from climate models are commonly used. The main goal of this paper is to report on the state of knowledge and recent works on the study of CC impacts with a focus on urban areas, in order to produce an integrated review of various approaches to which future studies can then be compared or constructed. Model output statistics (MOS) methods are increasingly used in the literature to study the impacts of CC in urban settings. A review of previous works highlights the non-stationarity nature of future climate data, underscoring the need to revise urban drainage system design criteria. A comparison of these studies is made difficult, however, by the numerous sources of uncertainty arising from a plethora of assumptions, scenarios, and modeling options. All the methods used do, however, predict increased extreme precipitation in the future, suggesting potential risks of combined sewer overflow frequencies, flooding, and back-up in existing sewer systems in urban areas. Future studies must quantify more accurately the different sources of uncertainty by improving downscaling and correction methods. New research is necessary to improve the data validation process, an aspect that is seldom reported in the literature. Finally, the potential application of non-stationarity conditions into generalized extreme value (GEV) distribution should be assessed more closely, which will require close collaboration between engineers, hydrologists, statisticians, and climatologists, thus contributing to the ongoing reflection on this issue of social concern.

Full text

ORIGINAL PAPER
Eustache Gooré Bi
1,2
&Philippe Gachon
3
&Mathieu Vrac
4
&Frédéric Monette
1
Received: 12 October 2014 /Accepted: 6 October 2015 /Published online: 27 October 2015
#Springer-Verlag Wien 2015
Abstract Changes in extreme precipitation should be one of
the primary impacts of climate change (CC) in urban areas. To
assess these impacts, rainfall data from climate models are
commonly used. The main goal of this paper is to report on
the state of knowledge and recent works on the study of CC
impacts with a focus on urban areas, in order to produce an
integrated review of various approaches to which future stud-
ies can then be compared or constructed. Model output statis-
tics (MOS) methods are increasingly used in the literature to
study the impacts of CC in urban settings. A review of previ-
ous works highlights the non-stationarity nature of future cli-
mate data, underscoring the need to revise urban drainage
system design criteria. A comparison of these studies is made
difficult, however, by the numerous sources of uncertainty
arising from a plethora of assumptions, scenarios, and model-
ing options. All the methods used do, however, predict in-
creased extreme precipitation in the future, suggesting poten-
tial risks of combined sewer overflow frequencies, flooding,
and back-up in existing sewer systems in urban areas. Future
studies must quantify more accurately the different sources of
uncertainty by improving downscaling and correction
methods. New research is necessary to improve the data val-
idation process, an aspect that is seldom reported in the liter-
ature. Finally, the potential application of non-stationarity con-
ditions into generalized extreme value (GEV) distribution
should be assessed more closely, which will require close
collaboration between engineers, hydrologists, statisticians,
and climatologists, thus contributing to the ongoing reflection
on this issue of social concern.
1 Introduction
It is now widely recognized, based on numerous studies, that
climate change (CC) will lead to higher likelihood and inten-
sity of weather phenomena that have the potential to cause
injury and loss of life to humans, property damage, social
and economic upheaval, and environmental degradation
(Berggren et al. 2011;GooréBi2015;Hayhoe2007;IPCC
2013; Langeveld et al. 2013; Semadeni-Davies et al. 2008;
Sunyer et al. 2014; Willems 2013). Optimal management of
rainwater in urban settings requires as clear as possible an
understanding of the response of existing drainage infrastruc-
tures to climate change. Since current urban drainage system
design is based on the fundamental assumption that historical
rain events are stationary, flooding in urban areas resulting
from future events that exceed the capacity of existing systems
could occur more frequently (Denault et al. 2006;GooréBi
et al. 2015b; Langeveld et al. 2013; Mailhot and Duchesne
*Eustache Gooré Bi
ba-eustache.goore-bi.1@ens.etsmtl.ca;
eustache.goorebi@ville.longueuil.qc.ca
Philippe Gachon
gachon.philippe@uqam.ca
Mathieu Vrac
mathieu.vrac@lsce.ipsl.fr
Frédéric Monette
frederic.monette@etsmtl.ca
1
Department of Construction Engineering, École de technologie
supérieure, Université du Québec, 1100 Notre-Dame Street West,
Montréal, Québec H3C 1K3, Canada
2
Department of Civil Engineering, City of Longueuil, 4250, chemin
de la Savane, Longueuil, Québec J3Y 9G4, Canada
3
Centre pour l’Étude et la Simulation du Climat à l’Échelle Régionale
(ESCER), Université du Québec à Montréal (UQAM),
Montréal, Québec, Canada
4
Laboratoire des Sciences du Climat et de l’Environnement-IPSL,
CNRS/CEA/UVSQ, Orme des Merisiers,
91191 Gif-sur-Yvette, France
Theor Appl Climatol (2017) 127:685–699
DOI 10.1007/s00704-015-1656-y
Which downscaled rainfall data for climate change impact studies
in urban areas? Review of current approaches and trends
Content courtesy of Springer Nature, terms of use apply. Rights reserved.