Analysis of uncertainties in future climate projections for South America: comparison of WCRP-CMIP3 and WCRP-CMIP5 models

Abstract

This paper identifies two sources of uncertainties in model projections of temperature and precipitation: internal and inter-model variability. Eight models of WCRP-CMIP3 and WCRP-CMIP5 were compared to identify improvements in the reliability of projections from new generation models. While no significant differences are observed between both datasets, some improvements were found in the new generation models. For example, in summer CMIP5 inter-model variability of temperature was lower over northeastern Argentina, Paraguay and northern Brazil, in the last decades of the 21st century. Reliability of temperature projections from both sets of models is high, with signal to noise ratio greater than 1 over most of the study region. Although no major differences were observed in both precipitation datasets, CMIP5 inter-model variability was lower over northern and eastern Brazil in summer (especially at the end of the 21st century). Reliability of precipitation projections was low in both datasets. However, the signal to noise ratio in new generation models was close to 1, and even greater than 1 over eastern Argentina, Uruguay and southern Brazil in some seasons.

Appendix

The new generation models (WCRP-CMIP5) present some differences regarding the previous generation (WCRP-CMIP3). Some of them are listed below:

• CanESM2 model (Chylek et al. 2011): presents a new algorithm of radiation (k-correlated distribution, Li and Barker 2005): an improvement in the aerosol optical property parameterization, direct and indirect radiative effects in aerosols using a prognostic bulk aerosol scheme, a new shallow convection scheme (von Salzen and McFarlane 2002).

• CSIRO-Mk3.6 model (Rotstayn et al. 2010): this version includes an interactive aerosol scheme (sulfate, dust, sea salt and carbonaceous aerosol); an updated in the radiation scheme; a modification in the boundaru-layer treatment (non-local scheme); some minor changes are introduced in convection and clouds schemes (including direct and indirect effects of aerosols and aerosols from volcanic eruptions).

• INM-CM4 model (Volodin et al. 2010): the time step has reduced from 12 to 5 min; the horizontal resolution is increased; some physical parameterizations have slightly changed: radiation, shallow and deep convections, orographic and nonorographic gravity waves resistances and processes related to vegetation and soil. IPSL-CM5A-LR (http://icmc.ipsl.fr/images/internal/ipsl-esm-20100228-1.pdf): improved horizontal resolution, inclusion of the carbon cycle; connection between tropospheric chemistry and layer.

• MIROC5 model (Watanabe et al. 2010): the vertical coordinate σ was changed to the hybrid σ-p (Arakawa and Konor 1996); the radiative scheme was updated (improvements in the line absorption and continuum absorption with an increase in the number of absorption bands from 18 to 29); the cumulus parameterization was changed (Chikira and Sugiyama 2009): it is an entraining-plume model, where the lateral entrainment rate varies vertically depending on the surrounding environment; in order to better represent cloud and cloud-radiative feedback, it was developed a prognostic LSC scheme (Watanabe et al. 2009) and a bulk microphysical scheme was implemented (Wilson and Ballard 1999); a aerosol module (SPRINTARS, Takemura et al. 2005, 2009) was coupled with the radiation and cloud microphysics schemes to calculate the direct and indirect effects of the aerosols.

• MIROC5-ESM model (Watanabe et al. 2011): based on the global climate model MIROC includes an atmospheric chemistry component (CHASER), a nutrient-phytoplankton-zooplankton-detritus (NPZD) type 10 ocean ecosystem component, and a terrestrial ecosystem component dealing with dynamic vegetation (SEIB-DGVM). This model has the fully resolved stratosphere and mesosphere (Watanabe et al. 2008). The hybrid terrain-following (sigma) pressure vertical coordinate system is used, and there are 80 vertical layers between the surface and about 0.003 hPa. In order to obtain the spontaneously generated equatorial quasi-biennial oscillation (QBO), a fine vertical resolution of about 680 m is used in the lower stratosphere. Some parameterizations were modified, especially those who are crucial for the representation of the large-scale dynamical and thermal structures in the stratosphere and mesosphere.

• MRI-CGCM3 model (Yukimoto et al. 2011): some settings were changed. For example, in the cumulus scheme (Arakawa and Schubert 1974) a prognostic mass flux method is used to express the condition of away from quasi-equilibrium (Randall and Pan 1993; Pan and Randall 1993, 1998). The cumulus downdraft is also considered (Cheng and Arakawa 1997). A mid-level convection scheme is included to express the mid-latitude convection accompanying the front and synoptic-scale disturbances. In the radiative scheme, the short and long wave radiation are treated separately. The long-wave (short-wave) region of the spectrum is divided into 9 (22) bands. The radiative flux is calculated in each band. In the cloud scheme, the model can recognize the spatial and temporal distribution of aerosols due to coupling between the aerosol model and the chemistry. It was implemented a new soil model (HAL; Hosaka 2011).

• HadGEM2-ES model (Martin et al. 2011, Collins et al. 2009): it was implemented an adaptive detrainment parameterization in the convection scheme, which improves the simulation of tropical convection and leads to a much reduced (and more realistic) wind stress over the tropical Pacific, and a package of changes including an alteration to the treatment of excess water from super-saturated soil surfaces and improved representation of the lifetime of convective cloud. Several changes and additions to the representation of aerosol have been carried out. Improvements include changes to existing aerosol species, such as sulphate and biomass-burning aerosols, and representation of additional species, such as mineral dust, fossil-fuel organic carbon, and secondary organic aerosol from biogenic terpene emissions. New Earth System components include the terrestrial and oceanic ecosystems and tropospheric chemistry. The ecosystem components are introduced principally to allow simulation of the carbon cycle and its interactions with climate.