Uncertainty quantification of socio-economic outcomes can be combined with scenario discovery techniques to explore a full range of outcomes and provide insight into associated likelihoods while also identifying individual scenarios of interest. This unique approach quantifies multi-sector risks, which can aid decision-making about energy and technology choices and sectoral strategies.
This study combines two well-known computer modeling techniques to scope out the energy and technology choices needed over the coming decades to achieve desired environmental and economic outcomes. The first technique, Monte Carlo analysis, quantifies uncertainty levels for dozens of energy and economic indicators; feeds that information into a model of the world economy that captures the cross-sectoral impacts of energy transitions; and runs that model hundreds of times to estimate the likelihood of different outcomes. The second technique, scenario discovery, uses machine learning tools to screen databases of model simulations in order to identify outcomes of interest and their conditions for occurring. In a novel approach, this study combines these tools with the Monte Carlo analysis to explore how different outcomes are related to one another. This approach can also identify individual scenarios that result in specific combinations of outcomes of interest, and provide insight into their likelihood and the conditions needed for them to occur.
This unique approach shows that there are many pathways to a successful energy transition that benefit both the environment and economy, and quantifies multi-sector risks associated with those pathways. It can thus be used to guide decision-makers in government and industry to make sound energy and technology choices and avoid biases in perceptions of what “needs” to happen to achieve certain outcomes. It can also be used to home in on scenarios of interest for use in particular studies.
Simulation models are often used to explore future development pathways and their impacts on energy, emissions, economies and the environment. This requires making assumptions about various socio-economic conditions, such as how fast populations and economies will grow, the cost of technology options, or the amount of fossil fuels available. Different assumptions have significant impacts on model results, yet analyses typically only test a few alternatives. In this study, researchers at the MIT Joint Program on the Science and Policy of Global Change develop and use probability distributions to capture this uncertainty. They draw samples from these distributions, run an energy-economic model hundreds of times, and quantify the resulting uncertainty in model outcomes, providing insight into their likelihood. They focus on results related to emissions and output from different economic sectors, as well as energy and electricity technologies. They also apply machine learning approaches to find scenarios of interest from within the database of scenarios. They find that many patterns of energy and technology development are possible under a given long-term environmental pathway or economic outcome.
BER PM Contact
Jennifer Morris (email@example.com)
Research Scientist, MIT Joint Program on the Science and Policy of Global Change
The study was funded by the U.S. Department of Energy (DOE) Office of Science under the grant DE-FG02-94ER61937. The scenario discovery component of the work was supported by DOE-funded collaborative projects between MIT and PNNL on Multi-Sector, Multi-Resource Interactions with Multiple Forcers (Contract #: 517296) and Uncertainty Characterization and Scenario Discovery in GCIMS (Contract #: 547784)
Morris, J.F., J.M. Reilly, S. Paltsev, A. Sokolov and K. Cox, Representing Socio-Economic Uncertainty in Human System Models, Earth’s Future Vol. 10, Issue 4 (April 2022). (This article also appears in Earth’s Future Special Issue Modeling MultiSector Dynamics to Inform Adaptive Pathways.)
PM Recommendation for SC Web Publication
Additional Information: Image by Flickr user jmenj