CRCM: a promising tool for studying the hydrological cycle in a
changing climate
by Biljana Music
Ouranos
Various approaches of hydroclimatic analysis at watershed scales
The traditional method to simulate the hydrological regime of a watershed is to use a hydrological model,
driven by observed meteorological data, which is calibrated by optimizing an ensemble of parameters to fit the
observed flow as closely as possible to the simulated regime. The perturbation or “delta” technique is often used
to assess the future hydrological regime of a watershed. A series of observed meteorological variables are
perturbed by the projected changes of the climate models, which are then used as inputs in hydrological models.
It is also possible to use a more direct method where outputs from the climate model (debiased or not) are used
as inputs in the hydrological model. However, these techniques do not account for the feedback processes that
operate at the surface-atmospheric interface via exchanges of various fluxes between the atmospheric and the
terrestrial branches of the water cycle.
A promising alternative for estimating watershed hydrological responses to global warming is to use
hydrological variables directly from the land-surface scheme that are embedded in a regional climate model
(RCM). This approach allows modellers to assess the future hydrology of a watershed while taking into
consideration the important interactions between the land-surface and the atmosphere. As RCMs operate based
on the concept of water and energy conservation, an internal consistency of the simulated climatic variables is
assured, which is a major advantage of this approach. The spatial resolution of RCMs is continually increasing,
resulting in a better representation of small scale processes. Thus, a more detailed description of the spatial
variability of soil and vegetation properties – which have significant effects on the hydrological cycle – becomes
possible. At the same time, physical parameterizations designed to capture the complex phenomena of the
subgrid-scale processes are improving continuously.
The Canadian Regional Climate Model (CRCM4),
developed by Ouranos, is one of the best performing
models among the numerous climate models
worldwide. In recent years, the Ouranos Climate
Simulations’ team has produced a great number of
climate simulations at a spatial resolution of 45 km.
Data from this vast databank proves to be very useful
for not only purely scientific studies, but also for
concrete applications. In 2008, Ouranos and the
Pacific Climate Impacts Consortium (PCIC) launched
a collaborative study with the general objective of
assessing the impact of climate change on the water
resources of several British Columbia watersheds by
using the raw outputs of the CRCM. At the same
time, PCIC initiated complementary studies using a
hydrological model (VIC), calibrated on the same
watersheds. The systematic conceptualization for the
overall project shown in Figure 1, which summarizes
Figure 1. Schema of the different methods for evaluating the
the methods that are now available to assess the
hydrological response of a watershed to global warming (Music et
al. 2010).
effects of global warming on runoff, has been
developed at Ouranos (Music et al. 2008; Music et al.
2010). Note that runoff is one of the principal components of the hydrological cycle that can be considered as a
measurement of water availability. On the same track, a collaboration agreement between Ouranos and Ontario
Power Generation (OPG) was concluded at the beginning of this year. The results of these studies were
integrated in several research reports (Rodenhuis et al. 2011; Music & Sykes, 2011; Shrestha et al. 2011).
RCM projections and associated uncertainty
In Rodenhuis et al. (2011) and Music & Sykes (2011), the results from the application of method “b” (Figure
1) are presented. Changes in the hydrological regimes of the investigated watersheds, due to global warming,
were assessed using the raw outputs from the
CRCM. Figure 2 shows the hydrological
change signal for the 2050 time horizon over
four watersheds (Upper Peace, Fraser,
Columbia et Campbell) located on West North
America, derived from the operational version
of the CRCM (V4.2) as a response to a
forcing following the SRES-A2 green gas
emissions’ scenario (Nakicenovic and Swart,
2000).
Uncertainty
associated
with
future
greenhouse gas emissions, climate model
imperfections, and variability and nonlinearity
of the climate system requires that the
complex issue related to the estimation of
uncertainties associated with the projections
be addressed. First, the effects of the internal
variability of the Canadian Global Climate
Model (GCM) on hydrological variables at the
watershed scale were assessed. This internal
variability can be used as an estimate of the
Figure 2. Changes projected in precipitation (P), runoff (R) and snow water
climate system’s natural variability resulting
equivalent (SWE) under the SRES-A2 scenario from an ensemble of the
from nonlinear interactions between its
CRCM simulations (between the period 1961-1990 and 2041-2070; in %) and
various components. Then, the regional
the uncertainties related to the natural variability. The bar graphs associated
repercussions of the internal variability on
with each watershed (Upper Peace, Fraser, Columbia and Campbell)
hydrological variables were estimated using
illustrate (Rodenhuis et al., 2011) the balance in projected changes (in
mm/day) of evapotranspiration (EVAP), runoff (RUNOFF) and precipitation
the CRCM.
(PCP).
To obtain a preliminary estimate of the effect of the GCM selection, which imposes the lateral boundary
conditions in regional model simulations, several GCMs were used to drive the CRCM (Figure 3). These GCMs
differ in their physics and dynamics as well as in their spatial and temporal resolutions. The effects of the
internal variability of the CRCM and its physical parameterization were also studied. This analysis was
accompanied by an evaluation of the ability of three different versions of the CRCM (V3.6, V3.7, V4.2) to
adequately simulate each of the water budget components of the investigated watersheds.
Newsletter – OURANOS – September 2011
2
Figure 3. Comparison of the projections for one of the priority Ontario Power Generation watersheds (under the SRES-A2
scenario) derived from the CRCM (symbols starting with CRCM), from the GCMs used to drive the CRCM (symbols
starting with CGCM and ECHAM), and from the other CMIP3 (Coupled Model Intercomparison Project phase 3) GCMs.
The box plots refer to the CMIP3 GCM projections, where the red solid/dashed lines show respectively the ensemble
median/mean (Music and Sykes, 2011).
References
Rodenhuis, D., B. Music, M. Braun, and D. Caya. 2011. Climate Diagnostics of Future Water Resource in BC Watersheds, Pacific Climate
Impacts Consortium, University of Victoria, 74 pp.
Music B. and C. Sykes. 2011. CRCM Diagnostics for Future Water Resources in OPG Priority Watersheds, Ouranos/Ontario Power
Generation, 48 pp.
Music, B., D. Caya, A. Frigon, M. Slivitzky, A. Musy, R. Roy and D. Rodenhuis. 2010. Canadian Regional Climate Model (CRCM) as a Tool
for Assessing Hydrological Impacts of Climate Change at the Watershed Scale, Proceedings of the Serbian Academy of Science
and Arts: MilutinMilankovitch 130th Anniversary Symposium, Belgrade, Serbia.
Music B., A. Musy and R. Roy. 2008. The Impact of Climate Change on Hydro-Electricity Generation, CEATI/Ouranos, 96 pp.
Nakicenovic, N., and R. Swart (Eds.). 2000. Special Report on Emissions Scenarios (SRES), Intergovernmental Panel on Climate Change
(IPCC), Cambridge University Press, 570 p.
Shrestha, R.R., A.J. Berland, M.A. Schnorbus, A.T. Werner. 2011. Climate Change Impacts on Hydro-Climatic Regimes in the Peace and
Columbia Watersheds, British Columbia, Canada. Pacific Climate Impacts Consortium, University of Victoria, Victoria, BC, 37 pp.
(pdf available).
OURANOS
550 Sherbrooke West Street, 19e floor, Montreal (Quebec) H3A 1B9
Telephone : 514-282-6464/ Fax : 514-282-7131
www.ouranos.ca
Newsletter – OURANOS – September 2011
3