A framework for modeling uncertainty in regional climate change
- 2.8k Downloads
In this study, we present a new modeling framework and a large ensemble of climate projections to investigate the uncertainty in regional climate change over the United States (US) associated with four dimensions of uncertainty. The sources of uncertainty considered in this framework are the emissions projections, global climate system parameters, natural variability and model structural uncertainty. The modeling framework revolves around the Massachusetts Institute of Technology (MIT) Integrated Global System Model (IGSM), an integrated assessment model with an Earth System Model of Intermediate Complexity (EMIC) (with a two-dimensional zonal-mean atmosphere). Regional climate change over the US is obtained through a two-pronged approach. First, we use the IGSM-CAM framework, which links the IGSM to the National Center for Atmospheric Research (NCAR) Community Atmosphere Model (CAM). Second, we use a pattern-scaling method that extends the IGSM zonal mean based on climate change patterns from various climate models. Results show that the range of annual mean temperature changes are mainly driven by policy choices and the range of climate sensitivity considered. Meanwhile, the four sources of uncertainty contribute more equally to end-of-century precipitation changes, with natural variability dominating until 2050. For the set of scenarios used in this study, the choice of policy is the largest driver of uncertainty, defined as the range of warming and changes in precipitation, in future projections of climate change over the US.
KeywordsClimate Sensitivity Natural Variability Precipitation Change Regional Climate Change Community Climate System Model Version
It is well established that the uncertainty in climate system parameters and projected emissions are important drivers of uncertainty in global climate change (Sokolov et al. 2009; Webster et al. 2012). In our modeling system, the climate response to given emissions is essentially controlled by three climate parameters, i.e., the climate sensitivity, the strength of aerosol forcing and the ocean heat uptake rate (Forest et al. 2001, 2008). Future emissions are driven by future economic activity and technological pathways influenced by climate policies, and population growth. Other sources of uncertainty in future climate projections, in particular at the regional level, include natural variability and structural uncertainty associated with differences in parameterization in existing climate models. It is well known that year-to-year variability in the climate system is large, in particular at high latitudes, making the emergence of significant climate change slow and signal-to-noise detection difficult (Mahlstein et al. 2011, 2012; Hawkins and Sutton 2012). At the same time, climate projections are heavily influenced by the characteristics of the chosen climate model and global climate models remain inconsistent in capturing regional precipitation changes and other atmospheric processes. Quantifying the likelihood of future regional climate change, such as the potential range of future warming, might be insightful to policy makers and impact modeling research groups who investigate climate change and its societal impacts at the regional level, including agriculture productivity, water resources and energy demand (Reilly et al. 2013).
In this study, we introduce a new modeling framework to investigate the uncertainty in regional climate change over the United States (US) associated with four sources of uncertainty, namely: (i) uncertainty in the emissions projections; (ii) uncertainty in the response of the global climate system to changes in greenhouse gases and aerosols concentrations; (iii) natural variability; and (iv) model structural uncertainty. The modeling framework is built around the Massachusetts Institute of Technology (MIT) Integrated Global System Model (IGSM) (Sokolov et al. 2005, 2009), an integrated assessment model with an Earth System Model of Intermediate Complexity (EMIC) (with a two-dimensional zonal-mean atmosphere). Regional climate change over the US is obtained through a two-pronged approach. First, we use the IGSM-CAM framework, which links the IGSM to the National Center for Atmospheric Research (NCAR) Community Atmosphere Model (CAM) (Monier et al. 2013a). Secondly, we use a pattern-scaling method that extends the IGSM zonal mean based on climate change patterns from various climate models (Schlosser et al. 2013). This two-pronged approach has been used to compute probabilistic projections of 21st century climate change over Northern Eurasia (Monier et al. 2013b).
In this paper, we present a description of the framework for modeling uncertainty in regional climate change. We then give a description of the matrix of simulations and present results of regional climate change over the US. We place a particular emphasis on quantifying the range of uncertainty and identifying the contributions of different sources of uncertainty considered in this study. The simulations presented here are part of a multi-model project to achieve consistent evaluation of climate change impacts in the US (Waldhoff et al. 2013).
2.1 Modeling framework
In this study, the core simulations use the MIT IGSM version 2.3 (Dutkiewicz et al. 2005; Sokolov et al. 2005), an integrated assessment model that couples an Earth System Model of Intermediate Complexity (EMIC), with a two-dimensional zonal-mean atmosphere, to a human activity model. The IGSM includes a representation of terrestrial water, energy, and ecosystem processes, global scale and urban chemistry including 33 chemical species, carbon and nitrogen cycle, thermodynamical sea ice, and a three-dimensional dynamical ocean component based on the MIT ocean general circulation model (Marshall et al. 1997a, b). Finally, the human systems component of the IGSM is the MIT Emissions Predictions and Policy Analysis (EPPA) model (Paltsev et al. 2005), which provides projections of world economic development and emissions over 16 global regions along with an analysis of proposed emissions control measures.
Since the IGSM atmospheric component is two-dimensional (zonally averaged), heat and freshwater fluxes are anomaly coupled in order to simulate a realistic ocean state. In order to more realistically capture surface wind forcing over the ocean, six-hourly National Centers for Environmental Prediction (NCEP) reanalysis 1 (Kalnay et al. 1996) surface 10-meter wind speed from 1948–2007 is used to formulate wind stress. The data are detrended through the analysis of the changes in zonal mean over the ocean (by month) across the full 60-year period; this has little impact except over the Southern Ocean, where the trend is quite significant (Thompson and Solomon 2002). For any given model calendar year, a random calendar year of wind stress data is applied to the ocean. This approach ensures that both short-term and interannual variability are represented in the ocean’s surface forcing. Different random sampling can be applied to simulate different natural variability, augmenting the traditional approach of specifying perturbations in initial conditions.
Because the IGSM includes a human activity model, it is possible to analyze uncertainties in emissions resulting from both uncertainty in model parameters and uncertainty in future climate policy decisions. Another major feature is the flexibility to vary key climate parameters controlling the climate response: climate sensitivity, strength of the aerosol forcing and ocean heat uptake rate. Because the IGSM has a two-dimensional zonal mean atmosphere, it cannot be directly used to simulate regional climate change. To simulate climate change over the US, we use a two-pronged method.
On the one hand, the MIT IGSM-CAM framework (Monier et al. 2013a) links the IGSM to the National Center for Atmospheric Research (NCAR) Community Atmosphere Model (CAM) version 3.1 (Collins et al. 2006b), with new modules developed and implemented in CAM to allow climate parameters to be changed to match those of the IGSM. In particular, the climate sensitivity of CAM is changed using a cloud radiative adjustment method (Sokolov and Monier 2012). In the IGSM-CAM framework, CAM is driven by greenhouse gas concentrations and aerosol loading computed by the IGSM model as well as IGSM sea surface temperature (SST) anomalies from a control simulation corresponding to pre-industrial forcing superposed on an observed monthly climatology from the merged Hadley-OI SST, a surface boundary dataset designed for uncoupled simulations with CAM (Hurrell et al. 2008). More details on the IGSM-CAM framework can be found in Monier et al. (2013a).
The IGSM-CAM simulations provide daily output at a resolution of 2° × 2.5° while the IGSM-pattern scaling method provides monthly output at the same 2° × 2.5° horizontal resolution.
2.2 Description of the simulations
To investigate the uncertainty in projections of future climate change, a core of 12 simulations with the IGSM is conducted with four values of climate sensitivity and three emissions scenarios. The three emissions scenarios are (i) a reference scenario with unconstrained emissions after 2012 (REF), with a total radiative forcing of 10.0 W m−2 by 2100; (ii) a stabilization scenario (POL4.5), with a total radiative forcing of 4.5 W m−2 by 2100; and (iii) a more stringent stabilization scenario (POL3.7), with a total radiative forcing of 3.7 W m−2 by 2100. More details on the ! emissions scenarios and economic implications, along with how they rel ate to the Representative Concentration Pathway (RCP) scenarios are given in Paltsev et al. 2013). To limit the number of simulations in this study we only consider one version of the three-dimensional dynamical ocean, with one particular ocean heat uptake rate (see Monier et al. 2013a). The four values of climate sensitivity (CS) considered are 2.0, 3.0, 4.5 and 6.0 °C, which represent respectively the lower bound (CS2.0), best estimate (CS3.0) and upper bound (CS4.5) of climate sensitivity based on the Fourth Assessment Report of the Intergovernmental Panel on Climate Change (Pachauri and Reisinger 2007), and a low probability/high risk climate sensitivity (CS6.0). Because the correlation between net aerosol forcing and climate sensitivity are conditional on the ocean heat uptake rate, we rely on the bivariate (climate sensitivity-net aerosol forcing) probability distribution for the particular value of ocean heat uptake rate used in this study to estimate the net aerosol forcing that provides the best agreement to reproduce historical climate change (Monier et al. 2013a). The values for the net aerosol forcing are −0.25 W/m 2, −0.70 W/m 2, −0.85 W/m 2 and -0.95 W/m 2 respectively, for CS = 2.0 °C, CS = 3.0 °C, CS = 4.5 °C and CS = 6.0 °C.
For each set of emissions scenario and climate sensitivity, a five-member ensemble of IGSM-CAM simulations is run with different random wind sampling and initial conditions, referred to as simply initial conditions in the remainder of the article, in order to account for the uncertainty in natural variability (Monier et al. 2013a), and the IGSM-pattern scaling is applied to four different patterns of regional climate change. Three IPCC AR4 climate models are chosen along with the IPCC AR4 multi-model ensemble mean. First, the NCAR Community Climate System Model version 3 (CCSM3.0) (Collins et al. 2006a) is chosen to compare with the IGSM-CAM results since both modeling systems have the same atmospheric and land components and they show similar biases over land (Monier et al. 2013a). However, because they have different ocean component, simulations with the IGSM-CAM and the IGSM-pattern scaling with CCSM3.0 are not necessarily expected to be identical. Nonetheless, this provides an opportunity to examine if the relative simple pattern-scaling scheme is sufficiently effective to replicate what can be represented in a more sophisticated three-dimensional climate model. The two additional models chosen are the models with the largest and smallest projected increases in precipitation over the US, respectively, the Bjerknes Centre for Climate Research Bergen Climate Model version 2.0 (BBCR_BCM2.0, Otterâ et al. 2009) and the Model for Interdisciplinary Research on Climate version 3.2 medium resolution (MIROC3.2_medres, Hasumi and Emori 2004). Finally, the multi-model ensemble pattern of regional climate change is obtained from the 17 IPCC AR4 climate models (see Schlosser et al. 2013).
In the remainder of this article, we refer to present-day as the mean over the 1991–2010 period and to 2100 as the mean over the 2091–2110 period.
3.1 Time series of US mean temperature and precipitation
3.2 Regional patterns of change
4 Summary and Conclusion
As part of a multi-model project to achieve consistent evaluation of climate change impacts in the US (Waldhoff et al. 2013), we use a series of 12 core simulations with the MIT IGSM with three different emissions scenarios and four values of climate sensitivity (Paltsev et al. 2013). We obtain regional climate change over the US using a two-pronged approach. On the one hand, we use the MIT IGSM-CAM framework, which links the IGSM to the CAM model, and run each of the 12 core simulations with five different initial conditions to account for uncertainty in natural variability. On the other hand, we apply a pattern scaling method to extend the latitudinal projections of the IGSM 2D zonal-mean atmosphere by applying longitudinally resolved patterns from 3 IPCC AR4 climate models and the multi-model mean based on 17 IPCC AR4 models. The three models chosen are the NCAR_CCSM3.0, which shares the same atmospheric model as the IGSM-CAM, BCCR_BCM2.0, which projects the largest increases in precipitation over the US, and MIROC3.2_medres, which predicts the least amount of precipitation increases over the US. This new framework for modeling uncertainty in regional climate change covers four dimensions of uncertainty: projected emissions, the global climate system response to changes in greenhouse gases and aerosols concentrations, natural variability and model structural uncertainty. Altogether, these simulations provide an efficient matrix of future climate projections to study climate impacts under uncertainty.
The simulations display a large range of US mean temperature and precipitation changes, and different regional patterns of change. In addition, the two different methods have very different treatments of natural variability. The IGSM-CAM physically simulates changes in both mean climate and extreme events, but relies on one particular model. The pattern scaling approach allows the spatial patterns of regional climate change of different climate models to be considered, but significantly underestimates year-to-year variability and cannot simulate extreme events or their potential changes under climate change. Together, these two methods provide complementary skills and an efficient framework to investigate uncertainty in future projections of regional climate change. The limitations of each methodology should be carefully accounted for when using this framework to drive impact models. In particular, researchers using climate simulations to drive impact models should always use individual model simulations and not ensemble mean simulations in order to account for natural variability. That is because natural variability is a driver for extreme climate and weather events, which can dominate impacts, and would not be accounted in ensemble mean simulations, as illustrated in Mills et al. (2004).
Finally, an analysis of the contribution of the four different sources of uncertainty reveals that the choice of policy and the value of the climate sensitivity have the largest impact on surface air temperature changes (the choice of policy being the dominant contributor), while the contributions from natural variability and structural uncertainty are small. On the other hand, the contributions of the four sources of uncertainty are more equal for changes in precipitation over the US but show large spatial heterogeneity. The role of natural variability on precipitation changes is substantial, in particular until 2050 when it dominates. This result is consistent with previous studies highlighting the important role of natural variability on regional climate projections (Deser et al. 2012; Monier et al. 2013a, b). Between 2050 and 2080, the four sources of uncertainty contribute equally to changes in precipitation over the US. After that, the choice of policy dominates similarly to previous studies (Hawkins and Sutton 2009, 2011). Since the pattern scaling method is based on a singlev realization of a climate model, the differences between models likely include contribution from natural variability associated with using different initial conditions. This is supported by the work of Deser et al. (2014) and reinforces the important role that natural variability plays in regional climate projections.
In light of these new results, it appears clear that the largest source of uncertainty in end-of-the-century projections of climate change over the US is also the only source that society has a control over: the emissions scenario. This should reflect the need to seriously consider implementing a global climate policy aiming at stabilizing greenhouse gases concentrations in the atmosphere.
While this study considers major sources of uncertainty in regional climate projections, it does not consider all possible sources. For example, Li et al. (2012) investigate the uncertainty in high-resolution temperature scenarios for North America from dynamical downscaling using different regional climate models and statistical downscaling. In addition, the contribution of each source of uncertainty depends strongly on the particular samples and choices made in this study. The implementation of only moderate policies or the choice of only low values of climate sensitivity would certainly decrease the estimates of their contribution to the overall changes. It should be reemphasized that we exclude the 6 °C climate sensitivity scenario, because of its low probability, when estimating the range of temperature and precipitation changes for each source of uncertainty considered in this study. In addition, sampling the ocean heat uptake rate would likely change the regional projections and the estimates of the impact of the uncertainty in the global climate response on changes in temperature and precipitation over the US. Nonetheless, this analysis demonstrates the relevance of modeling each source of uncertainty. It further demonstrates the need of new and more complete frameworks for modeling uncertainty in regional climate change.
This work was partially funded by the US Environmental Protection Agency’s Climate Change Division, under Cooperative Agreement XA-83600001 and by the US Department of Energy, Office of Biological and Environmental Research, under grant DEFG02-94ER61937. The Joint Program on the Science and Policy of Global Change is funded by a number of federal agencies and a consortium of 40 industrial and foundation sponsors. (For the complete list see http://globalchange.mit.edu/sponsors/all). This research used the Evergreen computing cluster at the Pacific Northwest National Laboratory. Evergreen is supported by the Office of Science of the US Department of Energy under Contract No. (DE-AC05-76RL01830). 20th Century Reanalysis V2 data provided by the NOAA/OAR/ESRL PSD, Boulder, Colorado, USA, from their Web site at http://www.esrl.noaa.gov/psd/.
- Adler R, Huffman G, Chang A, Ferraro R, Xie P, Janowiak J, Rudolf B, Schneider U, Curtis S, Bolvin D, et al. (2003) The version-2 global precipitation climatology project (GPCP) monthly precipitation analysis (1979-present). J Hydrometeor 4(6):1147–1167. doi: 10.1175/1525-7541(2003)004<1147:TVGPCP>2.0.CO;2 CrossRefGoogle Scholar
- Collins W, Bitz C, Blackmon M, Bonan G, Bretherton C, Carton J, Chang P, Doney S, Hack J, Henderson T, et al. (2006a) The community climate system model version 3 (CCSM3). J Clim 19(11):2122–2143. doi: 10.1175/JCLI3761.1
- Dutkiewicz S, Sokolov A, Scott J, Stone P (2005) A three-dimensional ocean-seaice-carbon cycle model and its coupling to a two-dimensional atmospheric model: uses in climate change studies. Tech. Rep. 122, MIT Joint Program on the Science and Policy of Global ChangeGoogle Scholar
- Hasumi H, Emori S (2004) K-1 coupled gcm (MIROC) description. Tech. Rep. Center for Climate System Research, University of Tokyo, TokyoGoogle Scholar
- Kalnay E, Kanamitsu M, Kistler R, Collins W, Deaven D, Gandin L, Iredell M, Saha S, White G, Woollen J, Zhu Y, Chelliah M, Ebisuzaki W, Higgins W, Janowiak J, Mo K C, Ropelewski C, Wang J, Leetmaa A, Reynolds R, Jenne R, Joseph D (1996) The NCEP/NCAR 40-year reanalysis project. Bull Am Meteorol Soc 77(3):437–471. doi: 10.1175/1520-0477(1996)077<0437:TNYRP>2.0.CO;2 CrossRefGoogle Scholar
- Meehl G, Stocker T, Collins W, Friedlingstein P, Gaye A, Gregory J, Kitoh A, Knutti R, Murphy J, Noda A, Raper S, Watterson I, Weaver A, Zhao Z C (2007) Global climate projections. In: Solomon S, Qin D, Manning M, Chen Z, Marquis M, Averyt K B, Tignor M, Miller H L (eds) Climate change 2007: the physical science basis.Contribution of working group I to the fourth assessment report of the intergovernmental panel on climate change, vol 8. Cambridge University Press, Cambridge, pp 747–845Google Scholar
- Mills D, Jones R, Carney K, St Juliana A, Ready R, Crimmins A, Martinich J, Shouse K, DeAngelo B, Monier E (2014) Quantifying and Monetizing Potential Climate Change Policy Impacts on Terrestrial Ecosystem Carbon Storage and Wildfires in the United States. Climatic Change, in press. doi: 10.1007/s10584-014-1118-z.
- Pachauri R, Reisinger A (2007) Climate change 2007 Synthesis report. contribution of working groups I, II and III to the fourth assessment report of the intergovernmental panel on climate change. Intergovernmental Panel on Climate ChangeGoogle Scholar
- Paltsev S, Reilly J, Jacoby H, Eckaus R, McFarland J, Sarofim M, Asadoorian M, Babiker M (2005) The MIT emissions prediction and policy analysis (EPPA) model: version 4. Tech. Rep. 125, MIT Joint Program on the Science and Policy of Global ChangeGoogle Scholar
- Paltsev S, Monier E, Scott JR, Sokolov AP, Reilly JM (2013) Integrated Economic and Climate Projections for Impact Assessment. Climatic Change. doi: 10.1007/s10584-013-0892-3
- Randall D A, WR A, Bony S, Colman R, Fichefet T, Fyfe J, Kattsov V, Pitman A, Shukla J, Srinivasan J, Stouffer R J, Sumi A, Taylor K E (2007) Climate models and their evaluation. In: Solomon S, Qin D, Manning M, Chen Z, Marquis M, Averyt K B, Tignor M, Miller H L (eds) Climate Change 2007: The physical science basis. Contribution of working group I to the fourth assessment report of the intergovernmental panel on climate change, vol 10. Cambridge University Press, Cambridge, pp 589–662Google Scholar
- Sokolov A, Schlosser C, Dutkiewicz S, Paltsev S, Kicklighter D, Jacoby H, Prinn R, Forest C, Reilly J, Wang C, et al. (2005) MIT integrated global system model (IGSM) version 2: model description and baseline evaluation. Tech. Rep. 124, MIT Joint Program on the Science and Policy of Global ChangeGoogle Scholar
- Sokolov A, Stone P, Forest C, Prinn R, Sarofim M, Webster M, Paltsev S, Schlosser C, Kicklighter D, Dutkiewicz S, et al. (2009) Probabilistic forecast for twenty-first-century climate based on uncertainties in emissions (without policy) and climate parameters. J Clim 22(19):5175–5204. doi: 10.1175/2009JCLI2863.1 CrossRefGoogle Scholar
- Waldhoff S, Martinich J, Sarofim M, DeAngelo B, McFarland J, Jantarasami L, Shouse K, Crimmins A, Li J (2013) Overview of the Special Issue: A multi-model framework to achieve consistent evaluation of climate change impacts in the United States. Climatic Change. (submitted, this issue)Google Scholar