Evaluation of the AR5 Conclusion of Anthropogenic Climate Change

The IPCC fifth assessment (AR5) was released in its entirety in November of 2014 demonstrating that human-induced climate change was occurring. The findings were formed on unprecedented advances in climate modelling and several predictive climate scenarios (Cubasch, 2013). The models considered the complexities of earth systems science and the intricacy socio-political policy and human behaviour has on earth’s climate response (Cubasch, 2013). The AR5 concluded that long-term and near-term mitigation and adaption measures were needed and that inherent uncertainties and climate science knowledge gaps, were becoming a significant challenging for governments in implementing mitigation and adaption measures (Collins, 2013).

Get Help With Your Essay
If you need assistance with writing your essay, our professional essay writing service is here to help!
Essay Writing Service

This report will first examine the science that forms the basis for the AR5 conclusion of anthropogenic climate change. Extrapolating climate data to find climate trends involves deliberate assumptions resulting in uncertainties. This paper explores the uncertainty in climate change forecast between global climate models, emission scenarios, climate sensitivity, and feedbacks, which leads to how uncertainties are becoming a challenge for governments and organisations to implement mitigation and adaptions measures. Finally, this report also looks at Australia’s alpine region to demonstrate the forecast, impacts, and risks for snow dependent regions.

To surmise climate forecasts, past and present climates must be understood. Therefore, the earth’s complex climate systems must be observed, and data from the atmosphere, oceans, and terrestrial areas collected. In doing so, several technologies have been deployed, ranging from ground-based instruments, ships, buoys, ocean profilers, balloons, aircraft, satellites (Cubasch, 2013). New generation Satellites equipped with high spectral resolution infrared technology provides observations of terrestrial climate variables (Cubasch, 2013). Remote sensing allows terrestrial data to be systematically collected instead of relying on in-situ infrastructure (Cubasch, 2013). Ocean observation technologies such as the Argo global array profiling floats system (Cubasch, 2013). Provide comprehensive coverage through profile floating and surface drifting buoys, allowing the oceans to be systematically monitored and recorded (Cubasch, 2013). These technologies form the Global Climate Observation System, along with the tools to analyse and process the data collected. The data provides a compressive picture of the earth’s climate. As a consequence of advances in data collection and monitoring, so too had climate modelling (Flato, 2013).

The systematic evaluation of climate data allows climate models to simulate past, present, and future climates (Cubasch, 2013). Designed to investigate the response additional forcings have on climate systems and make predictions about the environment on seasonal and decadal time scales. There are over 35 simulation models from various countries, combined and collectively called CMIP5, including both CSIRO simulation models from Australia CSIRO-Mk3.6.0 and Australian Community Climate and Earth-System Simulator 1.0 and 1.3 (Collins, 2013).

Each model is evaluated through model simulations called Organised Model Intercomparison Projects (MIP) (Flato, 2013). The simulations are benchmark experiments that test the models’ ability to simulate observed climate. The latest MIP is the Coupled model intercomparison Project 5 (CMIP5), which includes a more comprehensive suite of models than CMIP3 (predecessor to CIMP5) (Flato, 2013). The CMIP5 includes decadal length predictions of Atmosphere-Ocean General Circulation Models (AOGCMs) designed to simulate the atmosphere, land surface, oceans, and Sea-Ice. Also, the Earth System Model (ESM) expands on the AOGCM by simulating various biochemical cycles, such as the carbon cycle, sulphur cycle, and ozone (Flato, 2013).

The simulations are tested against the models’ ability to simulate observed climate, providing confidence in performance (Cubasch, 2013). Importantly, the models simulate future climate scenario emissions. They forecast emitted Green House Gases (GHG) and the consequential radiative forcing, which approximates the severity of atmospheric and ocean temperature increases (Flato, 2013). The fourth IPCC assessment report (AR4), used the CMIP3 climate model, coupled with the Socio-Economic Driven scenarios (SRES) (Flato, 2013). The SRES focused on socio-economic predictions about future scenarios but was limited by the climate data inputs (Flato, 2013). For AR5, the SRES scenarios were extended to explore different technologies, socio-economic, and policy futures incorporated in Integrated Assessment Models (IAM) (Collins, 2013).

The CMIP5 model included IAM and Representative Concentration Pathways (RCPs) comprised of land use, GHG emissions and concentrations, gas and aerosol emissions, and ozone to forecast emission scenarios (Cubasch, 2013). The RCP’s are do not necessarily predict the future any better than the SRES but do consider new climate systems. Four RCP’s represent the different emission output RCP 2.6, 4.5, 6.0, and 8.5. For example, RCP 2.5 has the least amount of emissions and the most ambitious targets for 2100 (Flato, 2013). While RCP 8.5 is an emission prediction based on the current trajectory and is, therefore, the worst case for 2100 (Collins, 2013). The suffix numerals highlight the estimated radiative forcing at 2100, such that RCP 2.6 would result in forcings of 2.6 W m2 and RCP8.5 forcings of 8.5 M m2 by 2100 (Cubasch, 2013).

The RCP’s, IAM, and CMIP5 are comprised to form Global Climate Models (GCM). In simple terms, the models represent the observed climate as mathematical representations of the climate system based on the laws of physics, conservation of mass, energy, and momentum (Stocker, 2013). The GCM is a three-dimensional representation of the atmosphere and oceans, representing large scale synoptic features of the atmosphere and ocean currents and other earth systems. However, because the RCPs and IAM are parameters for future conditions, uncertainty and emission scenarios, climate sensitivities and global models are a significant factor (Intergovernmental Panel on Climate Change, 2014).

The RCPs are designed to predict the future of anthropogenic GHG and aerosol emissions and by virtue, radiative forcing. Inherent uncertainties exist, which will impact the RCP outcomes such as population, economic growth, technological developments, and political and social changes (Collins, 2013). The four RCPs demonstrates the uncertainty inherent in predicting the impact emissions will have the earth’s climate, ranging from low impact (RCP2.5) to high impact (RCP8.5) (van Vuuren , et al., 2011). The estimated amount of GHG is determinate by several factors such as technologies, socio-economic pressures resulting in policies that work either towards increasing or decreasing emissions. Meeting RCP2.6 targets requires dramatic changes that need to be made to the socio-economic landscape by 2020, reducing our carbon emissions to 490Gt by 2100 (van Vuuren , et al., 2011).

Additionally, uncertainties exist around how mechanisms of the earths system will respond to increased GHG emission and warming. Instances that have already emerged were components of the observed earth system have responded unexpectedly, demonstrated through divergences from the observed climate data to the simulated climate data (Flato, 2013). For example, from 1998 to 2001, the observed global mean surface temperature had shown a reduced linear trend than the observed climate trend over the past 30 to 60 years. As a consequence, the models, in particular, model HadCRUT4 simulated a climate different from the observed climate (Flato, 2013). Examples such as these make it difficult to qualify the uncertainties, particularly as a consequence of climate sensibility ad feedbacks.

Climate change response can be challenging to detect in climate models owing to the sensitivity and feedbacks. For example, forcings can be amplified by feedbacks, which may be present before human-induced warming (Cubasch, 2013). Equilibrium climate sensitivity (ECS), which is a determinate of water vapour/lapse rate albedo and cloud feedbacks have different levels of certainty. Of all the feedback systems, the cloud component has the greatest amount of uncertainty. Key uncertainties exist around the understanding of shallow cumulus convection, micro physicals process in ice clouds and partial cloudiness from small-scale variations of cloud producing and cloud dissipation (Collins, 2013).

Furthermore, uncertainties of climate sensitivity and feedbacks exist around the AOGCM estimate ranged at 2 to 4.5 degrees Celsius (CSIRO, 2015). However, when coupled with CMIP5 the mean is 3.2 degrees Celsius, similar to CMIP3, demonstrating high sensitivities in some parameter ensemble models (Collins, 2013). Recent comparisons of ensembles compared with observed climate found that models with ECS values between 3 and 4 degrees Celsius had the smallest errors. Therefore, suggesting that models with mean ECS below 2 degrees Celsius are unreliable for showing mean climate (Flato, 2013).

There is some comparative difference in some climate model sensitivity based on warm and cold climates owing to the effects of cloud feedbacks (Cubasch, 2013). These estimates of climate sensitivity include slow feedbacks, such as ice sheets and vegetation. The temperature variability increases the complexity of relating the sensitivity to climate. ECS estimates from paleoclimate data inherently have uncertainty as a consequence of missing data, heterogeneous spatial coverage of data, and structural limitations in models (Flato, 2013).

Furthermore, statistical methods for shaping ECS or TCS are sensitive to assumed distribution. In principle, the results can narrow the uncertainty by combining observed warming trends, volcanic eruption, model climatology, and paleoclimate, yet there are no general methods on how this should be applied uniformly. Therefore it is sensitive and mostly reliant on expert judgment, and therefore there can be differences in results between global climate models (GCMs) (Flato, 2013).

GCMs represent the atmosphere and ocean as three-dimensional grids, with a typical atmospheric resolution of around 200 km and 20 to 50 levels in the vertical, explicitly representing large scale synoptic features of the atmosphere such as large scale oceanic currents and overturning (Flato, 2013). However, what these models do not pick up is the finer spatial scale such as radiation and precipitation process, cloud formation and atmospheric and oceanic turbulence. The GCMs include such mechanisms through parameterisation; that is, the approximated effects converted into a metric, developed through theoretical and observational studies (Collins, 2013).

Uncertainty in models can arise as a consequence of the detail and timing of climate change. How to represent the process in models and express confidence is more significant for some variables. Such as temperature and rainfall (Flato, 2013). No model is perfect; most models perform well within strict parameters. Often different models are better at predicting different mechanisms. Therefore the type of model is dependent on the question (Flato, 2013). For example, to compare climate impacts of various emissions, climate parameters need to be identified for which to predict and measure, that is either radiative forcings or temperature response. Each parameter will result in a required modelling framework and inherently produced uncertainty in climate forecasts (Flato, 2013).

The uncertainty of climate change provides challengers for adapting and mitigating climate change (CSIRO, 2015). The near-term implication to climate change is less challenging than long term implications only by the mere fact that near-term responses are present, and the scientific certainty is more accurate than that of the long-term (Intergovernmental Panel on Climate Change, 2014).

Implementation of adaption and mitigation in high-density human areas is particularly challenging at local and community levels because of the lack of consistent information about projected impacts (Stocker, 2013). For example, challengers involve the assessment and execution of policy due to limited human and financial resource, integration of the various levels of government, lack of guidance around principles and priorities and the different human responses to climate change in addition to the different values placed on natural resources (Stocker, 2013).

The natural ecosystems in Australia have been subject to recent climate trends with medium to high confidence. Hitherto, uncertainties remain around non-climatic drivers, like changes in atmospheric CO2, fire management, land use, and grazing (Intergovernmental Panel on Climate Change, 2014). In freshwater systems, it is challenging to identify climate change signatures over impacts from over-allocation of water, pollution, sedimentation, invasive species, and natural climate variability. The alpine ecosystems in New Zealand are equally as difficult because tree lines remain stable and unaffected for several hundred years despite a 0.9 degree Celsius warming over the past century (Flato, 2013).

Similarly, modelling suggests that the forestry industry will be impacted by the reduction of rainfall and increases in temperature. Nevertheless, existing climate variability and other confounding factors have prevented the identification of climate change from having an impact on forests (Stocker, 2013). The greatest challenge to adaption and mitigation measures of forestry is the incomplete knowledge of how plants will respond to increasing CO2 (Reisinger, 2014).

Likewise, based on the RCP climate scenarios agriculture is forecast to see a 4% reduction in gross value as a consequence of decreased rainfall and increased temperatures (Stocker, 2013). Nonetheless, challenges in adaption and mitigations measures emerge from uncertainties in the impact for elevated CO2 on the nutrient cycle, forage production, quality, water availability (Stocker, 2013). The IPCC report asserts that the incremental mitigation and adaptions measures initiated by Australia in response to the uncertainty of the RCP scenarios will not be enough to ensure the continued function of natural and human systems (Intergovernmental Panel on Climate Change, 2014).

Forecast precipitation and temperature scenarios are the most significant uncertainty preventing major industries such as agriculture, forestry, water, fire, and natural ecosystems from implementing necessary adaptive and mitigations measures (Reisinger, 2014). The scientific uncertainty is compounded by significant knowledge gaps of how earth system mechanisms will respond to climate change, such as ecological thresholds, rising CO2 concentrations on pests, natural climate variabilities such as ENSO and native and managed ecosystems (Reisinger, 2014).

Chapter 2 of the IPPC report (Intergovernmental Panel on Climate Change, 2014) working group II makes the point that governmental climate policy decisions are sensitive to the uncertainty of the predicted scenarios and have categorised five broad classes of uncertainties:

– Uncertainties around climate responses to GHG emissions and their associated impacts.

– Stocks and flows and other GHGs

– Technological systems

– Market behaviours and regulatory actions

– Individual and firm perceptions

Moreover, climate science and socio-political processes form a relationship between scientists and members empowered in creating policy aimed at shaping the political landscape of climate science (Intergovernmental Panel on Climate Change, 2014). As demonstrated, there are several knowledge gaps and uncertainties within climate science that significantly affect the socio-political relationship and in many cases, impedes the progress for adaptive and mitigation measures being implemented. These knowledge gaps and uncertainties are particularly evident in climate change forecasts for the Australian alpine regions (Intergovernmental Panel on Climate Change, 2014).

Observed changes in snow cover have occurred in previous decades as a consequence of climate change. A 5% decrease in snow cover has been observed in the Northern hemisphere since 1966. Snow cover has also been decreasing since mid-1980 through Europe and America with substantial decreases in snow depth below 1800m from 1950-2000 and a shift to earlier snow runoff (Hennessy, et al., Climate change effects on snow conditions in mainland Australia and adaptation at ski resorts through snowmaking, 2008).

In the Australian context, there has been a warming of 1 degree Celcius since 1950 with rainfall showing large annual and regional variability, which includes a decline in annual precipitation since 1950. Data from the Bureau of Meteorology (BoM) from 1950 to 2001 show an increase in winter maximum temperatures in south-east Australia and very little change in the minimum temperatures ( Hennessy, et al., The impact of climate change on snow conditions in mainland Australia, 2003).

Temperature data taken from alpine regions in the south-east of Australia showed trends in minimum and maximum temperatures from June to September between the periods of 1962 and 2001, showing positive trends in nearly all months for Cabramurra, Perisher and Thredbo region. Annual average rainfall was shown to decrease over most of south-western and eastern Australia, with increases in the northwest of Australia since 1950 (Pickering , Hill, & Green, 2008).

These climate trends are likely to impact Australian snowfields negatively. However, the large annual variability in snow seasons is a challenge for detecting climate change trends (Pickering , Hill, & Green, 2008). Impacts of future climate change for Australian snow cover using CSIRO climate change scenarios from the CSIRO snow models from 1996 estimated a decline of 30 days per year in the south-east alpine region of Australia and a further 18 to 66% by 2020 and by 2070 a further 39% to 96% decline. Based on warming trends of 0.3 to 1.3 degrees Celsius and 0.6 to 3.4 degrees Celsius and rainfall changes of 0 to 8% and 0 to 20%, respectively (Pickering , Hill, & Green, 2008).

By 2020 it is predicted that snow cover will decrease by 10 to 39% based on low impact climate scenario and 22-85% by 2050 for high impact scenarios. By 2050 average snow season length will decrease by 5 to 50 days or 10 to 60% at snowfields below 1600m and 15 to 80% at snowfields above 1600m (Pickering , Hill, & Green, 2008). As a consequence, maximum snow depth will occur earlier in the season and by 2050 the total area of snow cover will decrease by 22 to 85%, resulting in an average season length decreasing by 15 to 110 days or 30 to 99% reduction, at sites below 1600m and 15 to 95% at locations above 1600m ( Hennessy, et al., The impact of climate change on snow conditions in mainland Australia, 2003). The probability of exceeding a natural snow depth of 30cm each day is likely to decline as GHG warming occurs. Specifically for Mount Hotham, the probability of snow depth exceeding 30cm will decrease by 15 to 60% by 2020 based on a high impact climate scenario (Pickering , Hill, & Green, 2008).

The snowline is also expected to rise in elevation as a result of climate change. For example, the snow line at Mount Kosciuszko is likely to rise from the current average of 1460m to 1490 to 1625 by 2020. All alpine ski areas under low impact scenario for 2020 have a low probability of impact to snow conditions, with the average season reduced by approximately five days and peak depths affected by less than 10% but larger at areas lower than 1600m (Hennessy, et al., Climate change effects on snow conditions in mainland Australia and adaptation at ski resorts through snowmaking, 2008). Resulting significant climate change impacts for the Australian alpine regions.

Based on scenarios of high impact, the increased snowline and decreased snow cover are predicted to have a significant impact on alpine flora (Pickering , Hill, & Green, 2008). Australia’s alpine regions have high biodiversity with a large number of endemic species in alpine regions. Studies in the Snowy Mountains region have shown a trend of decreasing altitude with a decrease in species richness and an increase in altitude with an increase in species richness (Hennessy, et al., Climate change effects on snow conditions in mainland Australia and adaptation at ski resorts through snowmaking, 2008). Therefore, based on estimates of increasing altitude of snow line it is predicted that there will be a loss of species richness and a loss of specialised alpine species, particularly those that are endemic to alpine regions. Species endemic to localised alpine regions are at particular risk because unlike species of alpine regions in the northern hemisphere, Australian species are limited to altitude range with no potential for upward migration (Hennessy, et al., Climate change effects on snow conditions in mainland Australia and adaptation at ski resorts through snowmaking, 2008).

The increase in elevated temperatures is likely to alter the environment to that of a subalpine region. This not only results in the loss of endemic species but provides conditions for exotic species to be introduced, putting further pressure on the alpine ecosystem ( Hennessy, et al., The impact of climate change on snow conditions in mainland Australia, 2003).

Additional to the impact increased temperatures and reduced snow cover is likely to have on alpine biodiversity and ecosystems, is the impact increased temperatures and reduced snow cover will have on snowfields with heavily invested infrastructure (Hennessy, et al., Climate change effects on snow conditions in mainland Australia and adaptation at ski resorts through snowmaking, 2008). The decreased season length and reduced snow cover mean more snowmaking will be required, putting pressure facility costs ( Hennessy, et al., The impact of climate change on snow conditions in mainland Australia, 2003).

This problem is not unique to Australia; the Swiss ski industry released a report in 2002 noting that 85% of the countries ski resorts are likely to have a 44% decline in the coming decades, concluding that current predictions should influence operators to consider adaptation measures, which include ski slope design and artificial snowmaking ( Hennessy, et al., The impact of climate change on snow conditions in mainland Australia, 2003).

For decades snowmaking has been a significant contributor to natural snow cover in Australia, particularly for ski fields at low altitude ski runs. The quality of artificial snow in different ski areas in Australian is dependent on altitude, temperature, and humidity. Therefore the impact of increased temperature means the average number of suitable hours for making snow decreases by 2 to 7% for low impact scenarios and 17 to 54% for high impact scenarios ( Hennessy, et al., The impact of climate change on snow conditions in mainland Australia, 2003).

The impacts on snowmaking include consideration for the specific equipment. Currently, there is a mix of technology; however, with increased warming, different technology will suit different ski resorts. For example, modelling of snowmaking volumes showed that Mount Perisher had the best snowmaking capacity using a mix of technology owing to the resort’s altitude ( Hennessy, et al., The impact of climate change on snow conditions in mainland Australia, 2003). While Mount Thredbo, Mount Selwyn, Mount Buller, and Falls Creek could produce an equivalent amount of snow with specific equipment (Hennessy, et al., Climate change effects on snow conditions in mainland Australia and adaptation at ski resorts through snowmaking, 2008).

Typically a ski run in Australia is approximately 500 meters long and 40 meters wide, totalling an area of 20000m2. It is estimated that 8687 m3 of snow is required to cover a typical ski run at Mount Hotham in June. Mount Hotham, as do all the ski resorts are required to operate at a particular snow depth profile target. Based on snow modelling simulations of snow depth profile targets, June and September were the months that required the most significant amount of human-made snow. The implication and consideration of risk for Australian ski resorts will be the cost of installing new infrastructure equipped with the technology to adjust specific snowmaking parameters for different locations, increasing the cost for operators and users (Hennessy, et al., Climate change effects on snow conditions in mainland Australia and adaptation at ski resorts through snowmaking, 2008).

The risks of climate change for Australia’s alpine region is high. The level of impact for minor emission impacts (RCP2.5) is minimal. Significant changes to our human systems are required if we are to reach the RCP2.5 GHG emissions, including a significant shift in current socio-political and demographic trajectories (van Vuuren , et al., 2011). As identified in Chapter 25 of the IPCC report, Australia’s incremental shift towards climate mitigation for the long-term and climate adaption for the near-term is insufficient to maintain our natural and human systems at their current level. Based on the analysis of AR5 it is likely that not only Australia’s alpine regions but alpine regions in the Northern Hemisphere will be significantly impacted with reduced snow cover, increased temperatures, shift in precipitation trends and increased snowlines to higher altitudes. Climate change impacts to alpine regions will result in a significant change to alpine biodiversity and ecosystems, as introduced exotic species are capable of embedding into landscapes that were otherwise unavailable as landscapes shift from alpine climate to sub-alpine climates (Collins, 2013).

For Australia, this will no doubt will result in a vital loss of endemic species, unable to migrate to high altitudes because of Australia’s relatively low altitude mountains (Reisinger, 2014). Additionally, ski resorts will need to invest in precise snowmaking equipment and analysis tools to produced adequate snow for each ski fields specific latitude. Snowmaking operations will need to mature as operators find the balance between cost, quality, and environmental impact.

For climate change globally, indications of human-induced climate change are evident. The level of impact on countries worldwide encompasses a range of concerns from sea-level rise, fire management, drought, flooding. The AR5 makes the point that there will likely be a significant increase in extreme events (CSIRO, 2015). However, the ability to mitigate the risk is high provided that governments at all levels can overcome climate science uncertainty (Flato, 2013).

8.0           Conclusion

In conclusion, this report looked at the science that forms the basis for the AR5 including the technology involved in collecting the data advancements in modelling, which included RCP scenarios and IAM demonstrating the robustness of the climate science that forms the finding of the AR5. The paper examined the inherent uncertainty that exists within climate science and showed how the AR5 qualified assumptions to make confident climate projections. This paper also explored the uncertainty in climate change forecast between global climate models, emission scenarios, climate sensitivity, and feedbacks, leading to uncertainties which has led to increased challenges for governments and organisations in implement meaningful mitigation and adaptions measures. Finally, this report reviewed climate forecast and impacts on snow depth, coverage, and snow lines within the Australian alpine region, demonstrating the changing climate the forecast, impacts, and risks for snow dependent regions.

Hennessy, K., Whetton, P., Smith, I., Bathols, J., Hutchinson, M., & Sharples, S. (2003). The impact of climate change on snow conditions in mainland Australia. Aspendale, Victoria: CSIRO.

Collins, M. R.-L. (2013). Long-term Climate Change: Projections, Commitments and Irreversibility. In: Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. New York: Cambridge Univerity Press.

CSIRO. (2015). Cliamte Change in Australia Technical Report. Victoria: CSIRO.

Cubasch, U. D.-G. (2013). Introduction In: Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. New York: Cambridge Univerity Press.

Flato, G. J. (2013). Evaluation of Climate Models. In: Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. New York: Cambridge Univerity Press.

Hennessy, K., Whetton, P., Smith, I., Bathols, J., Hutchinson, M., & Sharples, S. (2008). Climate change effects on snow conditions in mainland Australia and adaptation at ski resorts through snowmaking. Climate Research, 35(2008), 255-270.

Intergovernmental Panel on Climate Change. (2014). Climate Change 2014 Mitigation of Climate Change: Working Group III Contribution to the. New York: Cambridge Univerity Press.

Pickering , C., Hill, W., & Green, K. (2008). Vascular plant diversity and climate change in the alpine zone of the Snowy Mountains, Australia. Biodiversity Conservation, 17(2008), 1627-1644.

Reisinger, A. R. (2014). Australasia. In: Climate Change 2014: Impacts, Adaptation, and Vulnerability. Part B: Regional Aspects. Contribution of Working Group II to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. New York: Cambridge Univerity Press.

Stocker, T. D.-K.-M. (2013). Technical Summary. In: Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. New York: Cambridge Univerity Press.

van Vuuren , P. D., Edmonds, J., Kainuma, M., Riahi , K., Thomson, A., Hibbard, K., . . . Rose, S. K. (2011). The representative concentration pathways: an overview. Climatic Change, 109:5 – 31.