Representing Socio‐Economic Uncertainty in Human System Models
Name
Earth s Future - 2022 - Morris - Representing Socio‐Economic Uncertainty in Human System Models.pdf
Description
Published version
Size
4.72 MB
Format
Adobe PDF
Checksum (MD5)
b82e4bd19906673b8203572296dec832
Author(s)
Morris, Jennifer
Reilly, John
Paltsev, Sergey
Sokolov, Andrei
Cox, Kenneth
Date Issued
March 21, 2022
Journal
Earth's Future
Publisher
American Geophysical Union
Citation
Morris, J., Reilly, J., Paltsev, S., Sokolov, A., & Cox, K. (2022). Representing socio-economic uncertainty in human system models. Earth's Future, 10, e2021EF002239.
Version
Final published version
Abstract
Socio‐economic development pathways and their implications for the environment are highly uncertain, and energy transitions will involve complex interactions among sectors. Here, traditional Monte Carlo analysis is paired with scenario discovery techniques to provide a richer portrait of these complexities. Modeled uncertain input variables include costs of advanced energy technologies, energy efficiency trends, fossil fuel resource availability, elasticities of substitution for labor, capital, and energy across economic sectors, population growth, and labor and capital productivity. The sampled values are simulated through a multi‐sector, multi‐region, recursively dynamic model of the world economy to explore a range of possible future outcomes. We find that many patterns of energy and technology development are possible for various long‐term environmental pathways and that sectoral output for most sectors is little affected through 2050 by the long‐term temperature target, but with tight constraints on emissions, emission intensities must fall much more rapidly. Scenario discovery techniques are applied to the large uncertainty ensembles to explore if there are prevailing storylines behind outcomes of interest. An illustrative investigation focused on different levels of economic growth shows many combinations of pathways and no single storyline emerging for a given economic outcome. This method can be extended to other outcomes of interest, exploring the nature of scenarios with both tail and median outcomes. Sampling from a Monte Carlo generated ensemble provides a rich set of scenarios to investigate, and potentially aids in avoiding heuristic biases in less structured scenario approaches.
MIT Department
Massachusetts Institute of Technology. Joint Program on the Science & Policy of Global Change
Terms of Use
Creative Commons Attribution
Persistent DSpace Link
DOI of Published Version
https://doi.org/10.1029/2021EF002239
