Fig. 1: Sensitivity of earth system models to initial conditions and implications for multi-sector impact assessment. | npj Climate and Atmospheric Science

Fig. 1: Sensitivity of earth system models to initial conditions and implications for multi-sector impact assessment.

From: Non-linear dynamical approaches for characterizing multi-sector climate impacts under irreducible uncertainty

Fig. 1

The variability in the spatial pattens of average decadal changes in a temperature over Australia, b precipitation in the United States, and c freshwater availability (calculated as the difference between precipitation and evaporation) in South America in the 2040s (2040–2049), relative to the 2020s (2020–2029) from two CMIP6 climate models (CESM 2, MRI-ESM2-0) and 3 initial conditions under SSP585. The role of internal climate variability in impact assessments is shown for d projected population density of Macrolophus pygmaeus (where a decline to zero indicates local extinction), e IDF curves for a 24-h precipitation event at the Earth system model grid point corresponding to Boston, and f decadal changes in mean runoff in the Amazon Basin (analysis for runoff is shown only for CESM2). These results highlight the strong influence of choice of initial conditions on impact assessments and the need to account for internal climate variability in adaptation planning. (See Methodology and Supplementary Information for details).

Back to article page