1. Introduction
Great uncertainty persists in future projections of the hydrological cycle response to global warming, which is caused by anthropogenic emissions of greenhouse gases and aerosols (Meehl et al. 2007). One of the major sources of uncertainty is the large range of potential climate sensitivities of surface air temperature to radiative forcing (Gregory et al. 2002; Forest et al. 2002, 2006; Knutti et al. 2003; Murphy et al. 2004; Stainforth et al. 2005; Meehl et al. 2007). Larger changes in temperature would lead to higher atmospheric moisture loading and hence contribute to larger changes in both evaporation and precipitation (Boer 1993; Allen and Ingram 2002; Held and Soden 2006). However, even if precipitation projections are standardized by the global mean temperature changes, uncertainties remain with regard to the precipitation sensitivity to 1 K of global warming. The largest source of uncertainty in the precipitation sensitivity is due to the model’s uncertainty (Boer 1993; Meehl et al. 2007; Shiogama et al. 2010, hereafter S10). However, S10 suggested that the emission scenario dependency of precipitation sensitivity is also important. According to climate change experiments that were carried out by multicoupled atmosphere–ocean general circulation models (AOGCMs), which were used in the Intergovernmental Panel on Climate Change (IPCC) Fourth Assessment Report (AR4), the precipitation sensitivities are smaller in emission scenarios with a larger emission of anthropogenic aerosols (also see the supplementary table S10.2 of Meehl et al. 2007).
The dependency of precipitation sensitivity on the emission scenario is important for impact, adaptation, and mitigation studies. These studies scale the precipitation sensitivities from outputs of AOGCMs to the global mean temperature projections from a simple climate model in order to obtain the precipitation climate scenarios under varied conditions of greenhouse gas concentrations and aerosol emissions; this is called the pattern scaling method (Santer et al. 1990; Takahashi et al. 1998; Wigley et al. 2000; Hulme et al. 2000; Huntingford and Cox 2000; Wigley 2003). The dependency of precipitation sensitivity on the emission scenario can bias the climate scenarios based on the pattern scaling method as well as the corresponding water resource impact assessments (S10).
There are three possible causes of precipitation sensitivity that differ within different emission scenarios (Mitchell et al. 1999; Mitchell 2003):
(i) The dependency of climate change patterns on the rate of global warming (Mitchell et al. 1999). For example, due to different heat capacities of land and ocean, the regional contrasts of temperature warming between continents and oceans are larger in faster warming scenarios (Mitchell 2003).
(ii) Nonlinearity stemming from the combination of greenhouse gas responses and other forcing responses (Feichter et al. 2004; Lohmann and Feichter 2005).
(iii) Different precipitation responses to different external forcing agents (Roeckner et al. 1999, 2006; Allen and Ingram 2002; Liepert et al. 2004; Ramanathan et al. 2001; Hansen et al. 2005; Jones et al. 2007; Yoshimori and Broccoli 2008; Ramanathan and Carmichael 2008; Lambert and Allen 2009; Liepert and Previdi 2009).
The aim of this study is to investigate which of these three factors determines the dependency of precipitation sensitivity on the emission scenario. Using a state-of-the-art AOGCM called Model for Interdisciplinary Research on Climate 3.2 (MIROC3.2) (K-1 Model Developers 2004), we perform ensembles with all of the combined forcing and separated individual forcing under two different emission scenarios. To examine the importance of factors (i), (ii), and (iii), we compare the following:
the well-mixed greenhouse gas (GHGS) runs under the different emission scenarios;
the all-forcing run and the sum of the individual forcing runs;
the precipitation sensitivities in the different forcing runs.
Using these analyses, we will show that different precipitation sensitivities between GHGS and carbonaceous aerosols are important contributors to the emission scenario dependency. Moreover, we examine the reasons for different precipitation sensitivities due to GHGS versus carbonaceous aerosols.
This paper is organized as follows: section 2 introduces the MIROC3.2 model and describes the experimental setup; section 3 contains the primary results; and section 4 presents the discussion and conclusions.
2. Model simulations
We use the MIROC3.2 AOGCM, which includes a simple aerosol transport–radiation model called the Spectral Radiation-Transport Model for Aerosol Species (SPRINTARS) (Takemura et al. 2000, 2002, 2005). This model calculates three-dimensional distributions of five types of aerosols: sulfate, black carbon, organic carbon, sea salt, and soil dust.
We performed future climate projections for 2001–2100 that are forced by changes in GHGS, black and organic carbon aerosols, sulfate aerosols, and tropospheric and stratospheric ozone under the A2 and B1 types of the IPCC Special Report on Emissions Scenarios (SRES) (Nakicenovic et al. 2000). Figure 1 shows the changes in the external forcing agents for both the A2 and B1 scenarios. Increases in CO2 concentrations are larger in A2 than in B1 (Fig. 1a). Changes in other GHGS (CH4, N2O, and 16 species of organic halogens) were shown by Nozawa et al. (2007). Column loading of black and organic carbon aerosols monotonically increases in A2, while changes in the carbonaceous aerosols are smaller in B1. We consider four major sources of black and organic carbon aerosols: fossil fuel combustion, domestic fuel wood consumption, agricultural waste burning, and wildfires. Emissions of carbonaceous aerosols were provided by T. Nozawa (2005, unpublished data), brief descriptions of which were given by Nozawa et al. (2007). In both of the scenarios, sulfate aerosols peak in the early half of the twenty-first century and subsequently decrease. We assume that the ozone mixing ratio in the stratosphere recovers in this century. Tropospheric ozone increases in A2, while in B1 it decreases in the latter half of this century.
We also computed ensembles forced by changes in individual external forcing agents (the other forcing agents are held constant at the year 2000 conditions): well-mixed greenhouse gases (the corresponding ensembles are referred to as A2GHG and B1GHG for the A2 and B1 scenarios, respectively), black and organic carbon aerosols (A2CRB and B1CRB), sulfate aerosols (A2SUL and B1SUL), and tropospheric and stratospheric ozone (A2OZN and B1OZN). The year 2000 commitment runs (Y2K) were also performed, wherein all of the external forcing factors are stabilized in the year 2000 (Wigley 2005; Meehl et al. 2005, 2007). Each ensemble includes 3 or 10 members. Their initial conditions in the year 2000 were taken from the 10-member ensemble of the 1850–2000 historical simulations (Nozawa et al. 2005; Shiogama et al. 2007, 2008). Details of the experimental setup are summarized in Table 1. S10 also analyzed A2, B1, A2GHG, B1GHG, and Y2K. Here, we examine the ensemble averages of the 10-yr mean anomalies relative to the 1981–2000 climatology. A stable 3600-yr preindustrial control run (CTL) is also used to estimate the possible influences of natural variability on the results.
3. Results
Figure 2 shows the decadal mean anomalies of global mean surface air temperature (ΔT) and precipitation (ΔP/P; percent change) in A2, B1, and Y2K relative to the 1981–2000 climatology. Although ΔT in A2 is significantly larger than that in B1, ΔP/P is similar in both scenarios, which indicates that the precipitation sensitivity (ΔP/P/ΔT) is smaller in A2 than B1 (S10).
To understand the sources of the ΔP/P/ΔT dependency on the emission scenarios, we examine ΔT and ΔP/P in the individual forcing runs (Fig. 3). To compare the sum of the individual forcing runs with the all forcing run, it is necessary to remove the influence of climate change commitment in the twenty-first century due to external forcing by the year 2000 (Wetherald et al. 2001; Wigley 2005; Meehl et al. 2005, 2007), which is included in all of the runs. Therefore, we hereafter remove the climate responses in Y2K (Fig. 1) from all of the runs. In the 2091–2100 period, both ΔT and ΔP/P in A2GHG are about twice those in the B1GHG. Table 2 shows the linear trends of ΔP/P divided by the linear trends of ΔT for different ensembles. Since the ensemble sizes of some runs are too small, only three, we did not use standard deviations of ensemble members in order to estimate error bars. Here error bars 〈,〉 indicate the minimum–maximum ranges of adding thirty-six 100-yr segments from the control run to the forced runs. To take into account the effect of ensemble mean of the forced runs, the control run was inflated by
The sums of the individual forcing runs (blue lines of Fig. 3) are very close to the climate responses in the all forcing runs (black lines) for both ΔT and ΔP/P (see also Table 2). Thus, factor (ii) of section 1 is not the reason for the emission scenario dependency. This is inconsistent with the results of Feichter et al. (2004), who showed that the linear additivity of the temperature responses between the GHGS and the aerosols was not restored in an atmosphere general circulation model (AGCM) with a mixed layer ocean (called ECHAM4), which included the first and second indirect effects. By contrast, Jones et al. (2007) suggested that another AGCM [Hadley Centre Global Environmental Model version 1 (HadGEM1)] that also included the first and second indirect effects actually held the linear additivity of the temperature and the precipitation between the sulfate and carbon aerosols. Further testing of the additivity is clearly warranted (see also the review by Forster et al. 2007).
It is clear that the suppression of ΔP/P in A2 can be attributed to the large reduction in ΔP/P in A2CRB (Fig. 3c). Increases in the carbonaceous aerosol induce little cooling in A2CRB. Although ΔT in A2CRB is much smaller than in A2GHG, decreases in ΔP/P in A2CRB counteract about half of the ΔP/P in A2GHG. By contrast, ΔT and ΔP/P in B1CRB are not very large in the 2091–2100 period since changes in the carbon aerosol forcing are small in the B1 scenario. These results mean that the dependency of ΔP/P/ΔT on the emission scenario is mainly caused by the larger ΔP/P/ΔT to the carbon aerosol forcing than that to the GHGS forcing (Table 2), and by different carbon aerosol emissions (Fig. 1).
The significant effects of carbonaceous aerosol forcing in the A2 scenario are also found in the geographical distributions of precipitation changes (Fig. 4). Differences between A2 and A2GHG are found in some areas, such as the tropics, the northern Pacific Ocean, North Africa, and North America. These differences are similar to the precipitation changes in the A2CRB but with some exceptions such as the Indian Ocean. It is suggested that carbonaceous aerosol forcing in the A2 scenario suppressed precipitation in the tropics and in some midlatitude regions, which mainly led to the differences in the precipitation trend patterns between A2 and A2GHG. The aforementioned discrepancy between A2 minus A2GHG and A2CRB seems to be largely attributable to A2OZN. Although the influence of ozone on regional precipitation is interesting, a full understanding will require future work.
The larger sensitivities of the hydrological cycle to the aerosol forcing than to the GHGS forcing have been shown by previous studies (Roeckner et al. 1999, 2006; Liepert et al. 2004; Ramanathan et al. 2001; Denman et al. 2007; Ramanathan and Carmichael 2008; Liepert and Previdi 2009). These studies suggested that dimming at the surface due to the interception of solar radiation by aerosols effectively spins down the hydrological cycle. In the global mean and the long-term mean, changes in precipitation are controlled by perturbations in evaporation (ΔEVAP) at the surface (e.g., Boer 1993). There is a balance at the surface between ΔEVAP, shortwave radiation (ΔSW), longwave radiation (ΔLW), sensible heat flux (ΔSENS), and ocean heat uptake (Boer 1993). Figure 5 indicates ΔP/|ΔT|, ΔEVAP/|ΔT|, ΔSENS/|ΔT|, ΔSW/|ΔT|, ΔLW/|ΔT|, and the residual (which should be the ocean heat uptake), where |ΔT| is the absolute value of ΔT for each run. There is an obvious balance between ΔP/|ΔT| and ΔEVAP/|ΔT| in all of the runs. The directions of ΔSW/|ΔT| are different between A2 and B1. Shortwave radiation is found to be dimmed in A2 but brightened in B1. These differences in ΔSW/|ΔT| are mainly balanced by smaller increases in ΔEVAP/|ΔT| for A2 than B1. Changes in radiation and heat fluxes are similar in both A2GHG and B1GHG. The GHGS forcing enhances downward trends in ΔSW, ΔLW, and ΔSENS while enhancing upward trends in ΔEVAP. Carbonaceous aerosols reduce the shortwave radiation that reaches the surface in A2CRB and A2. The reduction of available energy at the surface due to this large dimming effect in A2CRB is compensated for by significant negative ΔEVAP/|ΔT|(=ΔP/|ΔT|), which gives rise to a larger ΔP/P/ΔT in A2CRB than in A2GHG.
Changes in the atmospheric moisture (the precipitable water, ΔQ/Q) would be proportional to ΔT in all of the runs, as expected from the Clausius–Clapeyron arguments under a constant relative humidity (e.g., Allen and Ingram 2002). Thus, differences in ΔP/P/ΔT between the GHGS and carbon aerosol forcing involve differences in the conversion rates from Q to P and different atmospheric moisture residence times. Figure 6 compares ΔQ/Q and ΔP/P for the GHGS runs and the carbon aerosol runs. In A2GHG and B1GHG, although ΔQ/Q increases as the climate warms, ΔP/P is about 20% of ΔQ/Q, which indicates a decrease in the conversion rate (Trenberth 1998; Douville et al. 2002; Liepert et al. 2004). By contrast, in A2CRB and B1CRB, ΔQ/Q and ΔP/P decrease at similar rates as the climate cools. Feichter et al. (2004) also reported a similarly small change in the conversion rate in the carbon and sulfate aerosol forcing runs. In terms of the hydrological cycle in the troposphere, these different responses of the conversion rate from Q to P to ΔT contribute to the difference in ΔP/P/ΔT between the GHGS runs and the carbon aerosol runs. A recent study (O’Gorman and Schneider 2009) suggested that surface water vapor is more relevant to changes in extreme heavy precipitation than Q. However, it is not clear whether surface water vapor is more relevant to light and/or mean precipitation. Thus, we do not analyze surface water vapor in this paper.
Recent studies suggested that the decrease of the conversion rate from Q to P in the transient response to the GHGS forcing is attributable to a weakening in atmospheric vertical circulation (Held and Soden 2006; Vecchi and Soden 2007). As the dry static stability increases at a faster rate than the radiative cooling of the troposphere in a warmer climate, the subsidence rate decreases, which results in a weakening of the upward motion in the convection regions (Knutson and Manabe 1995). Following Vecchi and Soden (2007), we examine a vertical circulation index ω↑, which is defined as a globally integrated upward vertical pressure velocity at the 500-hPa surface from the monthly mean model outputs. The top panels of Fig. 7 show the percent change in vertical circulation, Δω↑/ω↑. In A2GHG and B1GHG, Δω↑/ω↑ decreases as the climate warms; in contrast, Δω↑/ω↑ increases as the climate cools in A2CRB and B1CRB. These changes in vertical circulation would offset parts of the effect of ΔQ/Q on ΔP/P. It should be noted that Δω↑/ω↑/ΔT of A2CRB (the linear trend of Δω↑/ω↑ divided by the linear trend of ΔT is −0.95% K−1) has a smaller magnitude than that of A2GHG (−2.19% K−1). This difference of Δω↑/ω↑/ΔT would contribute to the larger ΔP/P/ΔT of A2CRB than that of A2GHG. By absorbing shortwave, black carbon aerosols induce heating of the aerosol layer, leading to an increase of static stability (e.g., Hansen et al. 1997). This so-called semidirect effect may cause the small amplitude of vertical velocity changes per 1 K temperature changes in A2CRB.
Organic carbon aerosols involve the second indirect effect that directly influences the conversion rate from Q to P. The bottom panels of Fig. 7 are the percent changes in cloud droplet number concentration (ΔNc/Nc). In A2CRB, ΔNc/Nc monotonically increases as expected from changes in the column loading of the organic carbon aerosol (Fig. 1b). Therefore, the second indirect effect of the organic carbon aerosol would contribute to the suppression of precipitation (increases in ΔP/P/ΔT) in A2CRB. In B1CRB, changes in Nc are small. Small reductions in ΔNc/Nc are found in the GHGS runs. Decreases in sea salt emissions, which are internally computed in the model, seem to be the primary cause of these small reductions (not shown).
Figure 8 shows the corresponding estimations of ΔQ/Q, αΔω↑/ω↑, and βΔNc/Nc (note that the total least squares regression algorithm reduced climate noises in Δω↑/ω↑ and ΔNc/Nc). The primary causes of ΔP/P are ΔQ/Q in all the runs. The influences of ΔQ/Q on ΔP/P are counteracted by αΔω↑/ω↑. In the 2091–2100 period, (αΔω↑/ω↑)/(ΔQ/Q) are −0.80, −0.79, −0.31, and −0.37 for A2GHG, B1GHG, A2CRB, and B1CRB, respectively. In the GHGS runs, αΔω↑/ω↑ cancels large fractions of ΔQ/Q (about 80%), which results in the small ΔP/P/ΔT (Held and Soden 2006; Vecchi and Soden 2007). By contrast, in A2CRB and B1CRB, the magnitudes of positive αΔω↑/ω↑ are much smaller than those of negative ΔQ/Q. Furthermore, the negative βΔNc/Nc offsets the positive αΔω↑/ω↑ in A2CRB and B1CRB. These factors result in a ΔP/P close to ΔQ/Q.
We also derived a method for diagnosing the instantaneous second indirect effects of aerosols on precipitation. First, for each calendar month, we calculated a ratio (γ) of the 21-yr running averaged Nc between Y2K and A2CRB (Y2K/A2CRB) at each three-dimensional grid. We then computed the A2CRB run another time but with the large-scale condensation scheme considered twice. In the first call, Nc in Eq. (2) was scaled by γ. In the second call, Nc in Eq. (2) was not scaled by γ. After both calls, the increments from the second call were then applied to update the model state; the precipitation from large-scale condensation was stored for both. The instantaneous second indirect effect on precipitation was then computed as the difference in precipitation for the large-scale condensation between the two calls. This method can evaluate precipitation responses to low-frequency changes in Nc without any influence on the climate state. The black dotted lines in Figs. 8c and 8d indicate the instantaneous second indirect effect on precipitation in A2CRB and B1CRB. They are very close to the blue lines, which indicates that the estimations of the second indirect effect from the regression model are accurate. It should be noted that, when we computed γ, we used low-frequency changes in Nc. Also note that the total least squares regression algorithm reduced high-frequency climate noises in βΔNc/Nc. These smoothing processes also contributed to the resemblance between βΔNc/Nc and the instantaneous second indirect effect.
4. Summary and discussion
We investigated the reasons for the dependency of precipitation sensitivity on the emission scenario using the MIROC3.2 model. The precipitation sensitivity to 1 K of global warming is smaller in A2 than in B1. This difference of precipitation sensitivity is due to different compositions of emissions. In the emission scenario with high (low) emissions of carbon aerosols, one gets low (high) precipitation sensitivity. Future changes in precipitation amount, and associated climate change impacts (S10), depend not only on the amount of temperature raise but also on the composition of forcing agents.
The magnitude of precipitation change per 1 K of temperature change in the carbon aerosol runs is much larger than that of the GHGS runs. The interception of shortwave radiation due to the direct and indirect effects of carbon aerosols leads to surface dimming and hence to a reduction in evaporation. As a consequence, the carbonaceous aerosol forcing effectively spins down the hydrological cycle (largely decreases precipitation), whereas cooling of temperature is small (e.g., Roeckner et al. 1999; Liepert et al. 2004; Ramanathan and Carmichael 2008). By contrast, the GHGS forcing is primarily felt in the troposphere and indirectly changes the surface heat balance, which causes the precipitation sensitivity to be smaller than when aerosols are considered (Boer 1993; Allen and Ingram 2002; Liepert et al. 2004).
In the GHGS runs and the carbon aerosol runs, atmospheric moisture increases and decreases, mainly causing an increase and reduction of precipitation, respectively. There is a difference in the conversion rate of atmospheric moisture to precipitation between the GHGS runs and the carbon aerosol runs. Although the conversion rate decreases as the climate warms in the GHGS runs, changes in the conversion rate are small in the carbon aerosol runs. Whereas changes in vertical circulation lead to the large difference between perturbations in the atmospheric moisture and precipitation in the GHGS runs, changes in vertical circulation are small in the carbon aerosol runs. Additionally, in the carbon aerosol runs, the second indirect effect of the organic carbon aerosol cancels the effect of the vertical circulation changes. These factors cause the small change in conversion rate and hence the large precipitation sensitivity in the carbon aerosol runs.
We analyzed the precipitation response from the perspectives of surface heat budget and dynamical argument including the second indirect effect of aerosols. However, we did not examine the tropospheric energy budget, which is also important for the global precipitation response (Mitchell et al. 1987; Allen and Ingram 2002; Yang et al. 2003; Sugi and Yoshimura 2004; Stephens and Ellis 2008; Lambert and Allen 2009; Andrews 2009; Andrews et al. 2009). Changes in precipitation (and evaporation) are determined by the atmospheric radiative–convective energy balance. Increases of CO2 concentration lead to a surface-temperature-independent heating in the troposphere, which is compensated for by decreases in the tropospheric latent heating, that is, reductions of precipitation (e.g., Allen and Ingram 2002; Lambert and Allen 2009; Andrews 2009; Andrews et al. 2009). Shortwave forcing agents such as solar irradiance and direct effects of nonabsorbing aerosols directly affect the surface, causing larger sensitivities of precipitation and evaporation (e.g., Andrews et al. 2009). Carbon aerosols lead to a heating in the troposphere through shortwave absorption but a cooling at the surface through shortwave dimming. It is uncertain what fraction of the heating due to shortwave absorption in the troposphere is compensated for by a reduction in the tropospheric latent heating. Further analysis is necessary to understand the relationships between the different perspectives of precipitation sensitivity.
Using 20-yr satellite observations from the Special Sensor Microwave Imager (SSM/I), Wentz et al. (2007) suggested that precipitation have changed at a rate of 7% per 1 K of surface temperature change, which is much larger than precipitation sensitivities in the GHGS runs of AOGCMs, over the past two decades. Do current AOGCMs underestimate the precipitation response to anthropogenic surface temperature warming? There are some cautions in declaring that observed and AOGGCM precipitation sensitivities differ significantly. Uncertainties of satellite observations are not negligible (e.g., Allan and Soden 2007). Since the record length of satellite observations is short, uncertainties due to the natural variability are large (Previdi and Liepert 2008; Liepert and Previdi 2009). Although the record length of land precipitation observations is relatively long (about 100 years), the tropospheric energy budget does not constrain changes in the land precipitation (Lambert and Allen 2009). Natural sulfate aerosols due to large volcanic activity affected the past changes in land precipitation (Lambert et al. 2004; Gillett et al. 2004). The present paper showed that precipitation sensitivity depends on emissions of anthropogenic aerosols. Anthropogenic aerosols can have a large influence on the observed precipitation sensitivity over the past two decades (Lambert et al. 2008; Lambert and Allen 2009; Liepert and Previdi 2009; Andrews 2009; Andrews et al. 2009). Emissions of carbonaceous aerosols, especially from biomass burning, are highly uncertain in past data as well as in the future (Ramanathan and Carmichael 2008). Improvements in carbon aerosol emission estimations could have a great impact on detection-attribution as well as on projections of precipitation changes.
Aerosols may also affect events of light precipitation. Qian et al. (2009) suggested that significant increases of aerosols are partly responsible for decreases of light precipitation events observed in China over the past 50 years. In other words, aerosols can have implications for changes in drought. It is often assumed that changes in extremely heavy precipitation are constrained by changes in the atmospheric moisture (e.g., Allen and Ingram 2002; Emori and Brown 2005; Pall et al. 2007), implying that the sensitivity of extremely heavy precipitation may not depend on the emission scenarios. However, recent studies suggested that changes in daily heavy precipitation can exceed increases of the atmospheric moisture in the tropics (Sugiyama et al. 2010) and hourly heavy precipitation in the midlatitudes (Lenderink and van Meijgaard 2008). For daily heavy precipitation in the tropics, changes in the vertical profile and amplitude of upward velocity are important (O’Gorman and Schneider 2009; Sugiyama et al. 2010). It is not clear whether aerosols affect vertical velocity when extremely heavy precipitation events occur. Since extremely heavy precipitation washes out aerosols, changes in extremely heavy precipitation may not depend on the emission scenarios. It remains for future research to test whether aerosols have significant impacts on extremely heavy precipitation.
The dependency of precipitation sensitivity on the emission scenario has important implications for climate change impact assessments and adaptation studies (S10). Owing to limited computing resources, climate modeling centers can perform AOGCM simulations only for a limited number of emission scenarios. The pattern scaling method will therefore continue to be an important tool for constructing climate scenarios for a range of emission scenarios and stabilization pathways in the next assessment report of IPCC (Hibbard et al. 2007; Cox and Stephenson 2007; Moss et al. 2008). If we have only a B1 AOGCM simulation, we should overestimate precipitation changes in the A2 “extrapolated” projection (S10).
The Model for the Assessment of Greenhouse-Gas Induced Climate Change/Scenario Generator (MAGICC/SCENGEN) scales greenhouse gases and sulfate aerosol patterns to obtain climate scenarios (Schlesinger et al. 1997, 2000; Wigley et al. 2000; Hulme et al. 2000; Wigley 2003). This “separated pattern” approach is useful to overcome the influences of the emission scenario dependencies; however, this approach requires care in its use. The linear additivity of the responses of different forcing factors should be tested. A practical difficulty is the poor signal-to-noise ratio of the AOGCM responses to small forcing factors, such as aerosols. Although the large ensemble size of the AOGCM simulation enables us to improve the signal-to-noise ratio of the ensemble mean response, it requires extensive computing resources. Equilibrium simulations by AGCMs are low computing cost alternatives to transient AOGCM simulations. In this case, we should correct differences between the responses from AGCM and AOGCM, and between equilibrium and transient simulations (Mitchell et al. 1999; Mitchell 2003; Harris et al. 2006). The large aerosol forcing simulations of AOGCM could possibly improve the signal-to-noise ratio of the aerosol scaling pattern. However, there are issues regarding nonlinearity.
The results of the present study are based on the particular AOGCM and the particular assumption of carbonaceous aerosol emissions. Since the model’s uncertainties for the aerosol forcing and responses are large (e.g., Forster et al. 2007; Denman et al. 2007), multi-AOGCM studies are worthwhile. S10 showed that the smaller precipitation sensitivity in A2 and not B1 is a common feature in current AOGCMs, which contributed to AR4. However, most of the models have not considered carbonaceous aerosols. Although sulfate aerosols caused the emission scenario dependency of precipitation sensitivities in those models without carbonaceous aerosols, the emission scenario dependency was smaller in those models than that in a few models with carbonaceous aerosols (S10). In those models without carbonaceous aerosols, inclusions of carbonaceous aerosols will probably enhance the emission scenario dependencies of precipitation sensitivities (S10).
Acknowledgments
We thank M. Sugiyama, N. T. Edit, A. J. Broccoli, and the reviewers for their useful comments. This work was supported by the Global Environment Research Fund (S-5) of the Ministry of the Environment of Japan, by the Innovative Program of Climate Change Projection for the twenty-first century and by Grant-in-Aid 20740274 for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology of Japan. The Earth Simulator at JAMSTEC and an NEC SX-8R at NIES were employed to perform the AOGCM simulations.
REFERENCES
Albrecht, B. A., 1989: Aerosols, cloud microphysics, and fractional cloudiness. Science, 245 , 1227–1230.
Allan, R. P., and B. J. Soden, 2007: Large discrepancy between observed and simulated precipitation trends in the ascending and descending branches of the tropical circulation. Geophys. Res. Lett., 34 , L18705. doi:10.1029/2007GL031460.
Allen, M. R., and W. J. Ingram, 2002: Constraints on future changes in climate and the hydrologic cycle. Nature, 419 , 224–232.
Allen, M. R., and P. A. Stott, 2003: Estimating signal amplitudes in optimal fingerprinting. Part I: Theory. Climate Dyn., 21 , 477–491.
Andrews, T., 2009: Forcing and response in simulated 20th and 21st century surface energy and precipitation trends. J. Geophys. Res., 114 , D17110. doi:10.1029/2009JD011749.
Andrews, T., P. M. Forster, and J. M. Gregory, 2009: A surface energy perspective on climate change. J. Climate, 22 , 2557–2570.
Berry, E. X., 1967: Cloud droplet growth by collection. J. Atmos. Sci., 24 , 688–701.
Boer, G. J., 1993: Climate change and the regulation of the surface moisture and energy budgets. Climate Dyn., 8 , 225–239.
Cox, P., and D. Stephenson, 2007: A changing climate for prediction. Science, 317 , 207–208. doi:10.1126/science.1145956.
Denman, K. L., and Coauthors, 2007: Coupling between changes in the climate system and biogeochemistry. Climate Change 2007: The Physical Science Basis, S. Solomon et al., Eds., Cambridge University Press, 499–588.
Douville, H., F. Chauvin, S. Planton, J-F. Royer, D. Salas, D. Mélia, and S. Tyteca, 2002: Sensitivity of the hydrological cycles to increasing amounts of greenhouse gases and aerosols. Climate Dyn., 20 , 45–68.
Emori, S., and S. J. Brown, 2005: Dynamic and thermodynamic changes in mean and extreme precipitation under changed climate. Geophys. Res. Lett., 32 , L17706. doi:10.1029/2005GL023272.
Feichter, J., E. Roeckner, U. Lohmann, and B. Liepert, 2004: Nonlinear aspects of the climate response to greenhouse gas and aerosol forcing. J. Climate, 17 , 2384–2398.
Forest, C. E., P. H. Stone, A. P. Sokolov, M. R. Allen, and M. D. Webster, 2002: Quantifying uncertainties in climate system properties with the use of recent climate observations. Science, 295 , 113–117.
Forest, C. E., P. H. Stone, and A. P. Sokolov, 2006: Estimated PDFs of climate system properties including natural and anthropogenic forcings. Geophys. Res. Lett., 33 , L01705. doi:10.1029/2005GL023977.
Forster, P., and Coauthors, 2007: Changes in atmospheric constituents and in radiative forcing. Climate Change 2007: The Physical Science Basis, S. Solomon et al., Eds., Cambridge University Press, 129–234.
Gillett, N. P., A. J. Weaver, F. W. Zwiers, and M. F. Wehner, 2004: Detection of volcanic influence on global precipitation. Geophys. Res. Lett., 31 , L12217. doi:10.1029/2004GL020044.
Gregory, J., R. Stouffer, S. Raper, P. Stott, and N. Rayner, 2002: An observationally based estimate of the climate sensitivity. J. Climate, 15 , 3117–3121.
Hansen, J., M. Sato, and R. Ruedy, 1997: Radiative forcing and climate response. J. Geophys. Res., 102 , 6831–6864.
Hansen, J., and Coauthors, 2005: Efficacy of climate forcings. J. Geophys. Res., 110 , D18104. doi:10.1029/2005JD005776.
Harris, G. R., D. M. H. Sexton, B. B. B. Booth, M. Collins, J. M. Murphy, and M. J. Webb, 2006: Frequency distributions of transient regional climate change from perturbed physics ensembles of general circulation model simulations. Climate Dyn., 27 , 357–375. doi:10.1007/s00382-006-0142-8.
Held, I. M., and B. J. Soden, 2006: Robust responses of the hydrological cycle to global warming. J. Climate, 19 , 5686–5699.
Hibbard, K. A., G. A. Meehl, P. M. Cox, and P. Friedlingstein, 2007: A strategy for climate change stabilization experiments. Eos, Trans. Amer. Geophys. Union, 88 , 217–219.
Hulme, M., T. M. L. Wigley, E. M. Barrow, S. C. B. Raper, A. Centella, S. J. Smith, and A. C. Chipanshi, 2000: Using a climate scenario generator for vulnerability and adaptation assessments: MAGICC and SCENGEN Version 2.4 Workbook. Climatic Research Unit, 52 pp.
Huntingford, C., and P. M. Cox, 2000: An analogue model to derive additional climate change scenarios from existing GCM simulations. Climate Dyn., 16 , 575–586.
Jones, A., J. M. Haywood, and O. Boucher, 2007: Aerosol forcing, climate response and climate sensitivity in the Hadley Centre climate model. J. Geophys. Res., 112 , D20211. doi:10.1029/2007JD008688.
K-1 Model Developers, 2004: K-1 coupled GCM (MIROC) description. K-1 Tech. Rep. 1, H. Hasumi and S. Emori, Eds., Center for Climate Systems Research, University of Tokyo, 34 pp.
Knutson, T. R., and S. Manabe, 1995: Time-mean response over the tropical Pacific to increased CO2 in a coupled ocean–atmosphere model. J. Climate, 8 , 2181–2199.
Knutti, R., T. F. Stocker, F. Joos, and G. Plattner, 2003: Probabilistic climate change projections using neural networks. Climate Dyn., 21 , 257–272.
Lambert, F. H., and M. R. Allen, 2009: Are changes in global precipitation constrained by the tropospheric energy budget? J. Climate, 22 , 499–517.
Lambert, F. H., P. A. Stott, M. R. Allen, and M. A. Palmer, 2004: Detection and attribution of changes in 20th century land precipitation. Geophys. Res. Lett., 31 , L10203. doi:10.1029/2004GL019545.
Lambert, F. H., A. Stine, N. Y. Krakauer, and J. C. H. Chiang, 2008: How much will precipitation increase with global warming? Eos, Trans. Amer. Geophys. Union, 89 , 193–194.
Lenderink, G., and E. van Meijgaard, 2008: Increases in hourly precipitation extremes beyond expectations from temperature changes. Nat. Geosci., 1 , 511–514.
Liepert, B. G., and M. Previdi, 2009: Do models and observations disagree on the rainfall response to global warming? J. Climate, 22 , 3156–3166.
Liepert, B. G., J. Feichter, U. Lohmann, and E. Roeckner, 2004: Can aerosols spin down the water cycle in a warmer and moister world? Geophys. Res. Lett., 31 , L06207. doi:10.1029/2003GL019060.
Lohmann, U., and J. Feichter, 2005: Global indirect aerosol effects: A review. Atmos. Chem. Phys., 5 , 715–737.
Meehl, G. A., W. M. Washington, W. D. Collins, J. M. Arblaster, A. Hu, L. E. Buja, W. G. Strand, and H. Teng, 2005: How much more global warming and sea level rise? Science, 307 , 1769–1772.
Meehl, G. A., and Coauthors, 2007: Global climate projections. Climate Change 2007: The Physical Science Basis, S. Solomon et al., Eds., Cambridge University Press, 747–846.
Mitchell, J. F. B., C. A. Wilson, and W. M. Cunnington, 1987: On CO2 climate sensitivity and model dependence of results. Quart. J. Roy. Meteor. Soc., 113 , 293–332.
Mitchell, J. F. B., T. C. Johns, M. Eagles, W. J. Ingram, and R. A. Davis, 1999: Towards the construction of climate change scenarios. Climatic Change, 41 , 547–581.
Mitchell, T. D., 2003: Pattern scaling. An examination of the accuracy of the technique for describing future climates. Climatic Change, 60 , 217–242.
Moss, R., and Coauthors, 2008: Towards new scenarios for analysis of emissions, climate change, impacts, and response strategies. IPCC Expert Meeting Rep., 169 pp. [Available online at http://www.ipcc.ch/meetings/session28/doc8.pdf].
Murphy, J. M., D. M. H. Sexton, D. N. Barnett, G. S. Jones, M. I. Webb, M. Collins, M. R. Allen, and D. J. Stainforth, 2004: Quantifying uncertainties in climate change using a large ensemble of general circulation model predictions. Nature, 430 , 768–772.
Nakicenovic, N., and Coauthors, 2000: Special Report on Emissions Scenarios. Cambridge University Press, 599 pp.
Nozawa, T., T. Nagashima, H. Shiogama, and S. A. Crooks, 2005: Detecting natural influence on surface air temperature change in the early twentieth century. Geophys. Res. Lett., 32 , L20719. doi:10.1029/2005GL023540.
Nozawa, T., T. Nagashima, T. Ogura, T. Yokohata, N. Okada, and H. Shiogama, 2007: Climate change simulations with a coupled ocean–atmosphere GCM called the Model for Interdisciplinary Research on Climate: MIROC. CGER’S Supercomputer Monograph Rep. 12, 77 pp.
O’Gorman, P. A., and T. Schneider, 2009: The physical basis for increases in precipitation extremes in simulations of 21st-century climate change. Proc. Natl. Acad. Sci. USA, 106 , 14773–14777.
Pall, P., M. R. Allen, and D. A. Stone, 2007: Testing the Clausius–Clapeyron constraint on changes in extreme precipitation under CO2 warming. Climate Dyn., 28 , 351–363.
Previdi, M., and B. G. Liepert, 2008: Interdecadal variability in the precipitation response to global warming. Eos, Trans. Amer. Geophys. Union, 89 , 193–195.
Qian, Y., D. Gong, J. Fan, L. R. Leung, R. Bennartz, D. Chen, and W. Wang, 2009: Heavy pollution suppresses light rain in China: Observations and modeling. J. Geophys. Res., 114 , D00K02. doi:10.1029/2008JD011575.
Ramanathan, V., and G. Carmichael, 2008: Global and regional climate changes due to black carbon. Nat. Geosci., 1 , 221–227.
Ramanathan, V., P. J. Crutzen, J. T. Kiehl, and D. Rosenfeld, 2001: Aerosols, climate, and the hydrological cycle. Science, 294 , 2119–2124.
Roeckner, E., L. Bengtsson, J. Feichter, J. Lelieveld, and H. Rodhe, 1999: Transient climate change simulations with a coupled atmosphere–ocean GCM including the tropospheric sulfur cycle. J. Climate, 12 , 3004–3032.
Roeckner, E., P. Stier, J. Feichter, S. Kloster, M. Esch, and I. Fischer-Bruns, 2006: Impact of carbonaceous aerosol emissions on regional climate change. Climate Dyn., 27 , 553–571. doi:10.1007/s00382-006-0147-3.
Santer, B. D., T. M. L. Wigley, M. E. Schlesinger, and J. F. B. Mitchell, 1990: Developing climate scenarios from equilibrium GCM results. Max-Planck-Institut für Meteorologie Rep. 47, Hamburg, Germany, 29 pp.
Schlesinger, M. E., N. G. Andronova, A. Ghanem, S. Malyshev, T. Reichler, E. V. Rozanov, W. Wang, and F. Yang, 1997: Geographical scenarios of greenhouse-gas and anthropogenic-sulfate-aerosol induced climate changes. Rep. of the Climate Research Group, Department of Atmospheric Sciences, University of Illinois at Urbana–Champaign, 85 pp.
Schlesinger, M. E., and Coauthors, 2000: Geographical distributions of temperature change for scenarios of greenhouse gas and sulfur dioxide emissions. Technol. Forecasting Soc. Change, 65 , 167–193.
Shiogama, H., T. Nozawa, and S. Emori, 2007: Robustness of climate change signals in near term predictions up to the year 2030: Changes in the frequency of temperature extremes. Geophys. Res. Lett., 34 , L12714. doi:10.1029/2007GL029318.
Shiogama, H., A. Hasegawa, T. Nozawa, and S. Emori, 2008: Changes in mean and extreme precipitation in near-term predictions up to the year 2030. SOLA, 4 , 17–20. doi:10.2151/sola.2008-005.
Shiogama, H., and Coauthors, 2010: Emission scenario dependencies in climate change assessments of the hydrological cycle. Climatic Change, 99 , 321–329. doi:10.1007/s10584-009-9765-1.
Stainforth, D., and Coauthors, 2005: Uncertainty in predictions of the climate response to rising levels of greenhouse gases. Nature, 433 , 403–406.
Stephens, G. L., and T. D. Ellis, 2008: Controls of global-mean precipitation increases in global warming GCM experiments. J. Climate, 21 , 6141–6155.
Sugi, M., and J. Yoshimura, 2004: Mechanism of tropical precipitation change due to CO2 increase. J. Climate, 17 , 238–243.
Sugiyama, M., H. Shiogama, and S. Emori, 2010: Precipitation extreme changes exceeding moisture content increases in MIROC and IPCC climate models. Proc. Natl. Acad. Sci. USA, 107 , 571–575.
Takahashi, K., Y. Matsuoka, and H. Harasawa, 1998: Impacts of climate change on water resources, crop production and natural ecosystem in the Asia and Pacific region. J. Global Environ. Eng., 4 , 91–103.
Takemura, T., H. Okamoto, Y. Maruyama, A. Numaguti, A. Higurashi, and T. Nakajima, 2000: Global three-dimensional simulation of aerosol optical thickness distribution of various origins. J. Geophys. Res., 105 , (D14). 17853–17873.
Takemura, T., T. Nakajima, O. Dubovik, B. N. Holben, and S. Kinne, 2002: Single-scattering albedo and radiative forcing of various aerosol species with a global three-dimensional model. J. Climate, 15 , 333–352.
Takemura, T., T. Nozawa, S. Emori, T. Y. Nakajima, and T. Nakajima, 2005: Simulation of climate response to aerosol direct and indirect effects with aerosol transport-radiation model. J. Geophys. Res., 110 , D02202. doi:10.1029/2004JD005029.
Trenberth, K. E., 1998: Atmospheric moisture residence times and cycling: Implications for rainfall rates with climate change. Climatic Change, 39 , 667–694.
Twomey, S., M. Piepgrass, and T. L. Wolfe, 1984: An assessment of the impact on global cloud albedo. Tellus, 36B , 356–366.
Vecchi, G. A., and B. J. Soden, 2007: Global warming and the weakening of the tropical circulation. J. Climate, 20 , 4316–4340.
Wentz, F. J., L. Ricciardulli, K. Hilburn, and C. Mears, 2007: How much more rain will global warming bring? Science, 317 , 233–235.
Wetherald, R. T., R. J. Stouffer, and K. W. Dixon, 2001: Committed warming and its implications for climate change. Geophys. Res. Lett., 28 , 1535–1538.
Wigley, T. M. L., 2003: MAGICC/SCENGEN 4.1: Technical Manual. National Center for Atmospheric Research, 14 pp.
Wigley, T. M. L., 2005: The climate change commitment. Science, 307 , 1766–1769.
Wigley, T. M. L., S. C. B. Raper, S. Smith, and M. Hulme, 2000: The MAGICC/SCENGEN Climate Scenario Generator: Version 2.4: Technical Manual. Climatic Research Unit, University of East Anglia, 35 pp.
Yang, F., A. Kumar, M. E. Schlesinger, and W. Wang, 2003: Intensity of hydrological cycles in warmer climates. J. Climate, 16 , 2419–2423.
Yoshimori, M., and A. J. Broccoli, 2008: Equilibrium response of an atmosphere-mixed layer ocean model to different radiative forcing agents: Global and zonal mean response. J. Climate, 21 , 4399–4423.
External forcing agents in A2 (solid lines) and B1 (dashed lines). (a) Concentrations of CO2 (ppmv). Other well-mixed greenhouse gases were shown by Nozawa et al. (2007). (b) Changes in the column loading of sulfate (red), black carbon (black), and organic carbon aerosols (green) from the 1981–2000 mean. (c) Ozone mixing ratios in the stratosphere (ppmv) and the troposphere (ppbv). The ozone mixing ratios in the stratosphere are the same in A2 and B1.
Citation: Journal of Climate 23, 9; 10.1175/2009JCLI3428.1
Changes in the global mean (a) surface air temperature (K) and (b) precipitation (%). The solid, dashed, and dotted lines indicate A2, B1, and Y2K, respectively. The ensemble means are used.
Citation: Journal of Climate 23, 9; 10.1175/2009JCLI3428.1
(a) Changes in the surface air temperature (K) for A2 (black), A2GHG (red), A2SUL (orange), A2CRB (green), A2OZN (light blue), and combinations of individual forcing runs (blue). (b) As in (a) except for the B1 scenario. (c),(d) As in (a) and (b) except for the percent change in precipitation (%). The Y2K runs are removed from all runs (see text for details).
Citation: Journal of Climate 23, 9; 10.1175/2009JCLI3428.1
Geographical distribution of changes (2081–2100 minus 1981–2000; percent change) in precipitation in (a) A2, (b) A2GHG, (d) A2CRB, (e) A2SUL, and (f) A2OZN. (c) Differences in precipitation changes between A2 and A2GHG. The Y2K runs are removed from all runs.
Citation: Journal of Climate 23, 9; 10.1175/2009JCLI3428.1
Trends (2001–2100) of ΔP/|ΔT| (yellow), ΔEVAP/|ΔT| (blue), ΔSENS/|ΔT| (red), ΔSW/|ΔT| (purple), ΔLW/|ΔT| (green), and residual (light blue) (W m−2 K−1) for (a) ensembles A2, A2GHG, and A2CRB and (b) ensembles B1, B1GHG, and B1CRB. The Y2K runs are removed from all runs. The error bars are the minimum–maximum ranges of adding thirty-six 100-yr segments from the control run to the forced runs. To take into account the effect of ensemble mean of the forced runs and the Y2K runs, the control run was inflated by
Citation: Journal of Climate 23, 9; 10.1175/2009JCLI3428.1
Percent change in the atmospheric moisture (solid lines) and precipitation (dashed lines) for (a) A2GHG, (b) B1GHG, (c) A2CRB, and (d) B1CRB. The Y2K runs are removed from all runs.
Citation: Journal of Climate 23, 9; 10.1175/2009JCLI3428.1
(a) Percent change in the upward vertical velocity at the 500-hPa surface for A2GHG (solid lines) and B1GHG (dashed lines). (b) As in (a) except for A2CRB and B1CRB. (c),(d) As in (a) and (b) except for the percent change in the cloud droplet number concentration. The Y2K runs are removed from all runs.
Citation: Journal of Climate 23, 9; 10.1175/2009JCLI3428.1
The black lines are the percent change in the precipitation (%) for (a) A2GHG, (b) B1GHG, (c) A2CRB, and (d) B1CRB. The Y2K runs are removed from all of the runs. The red, green, and blue lines indicate the percent change in the precipitation attributable to changes in the atmospheric moisture, upward vertical velocity, and cloud droplet number concentration, respectively, estimated by the regression model. The black dashed lines are combinations of individual factors that are estimated by the regression model, which are close to the black solid lines. The black dotted lines in (c) and (d) show the instantaneous second indirect effects on the precipitation due to organic carbon aerosols in A2CRB and B1CRB, respectively.
Citation: Journal of Climate 23, 9; 10.1175/2009JCLI3428.1
Details of the ensembles analyzed in this study.
Precipitation sensitivities for the different ensembles (see text for details).
