Extended Data Fig. 8: Comparison of our probabilistic diffusion model results with IAM climate mitigation scenarios, targets, and further studies.

In both the EU and globally, IAM scenarios tend to use (far) less green hydrogen than envisaged by policy targets or other studies such as by Hydrogen Europe, the Hydrogen Council, the IEA, or IRENA. This is especially true for the global IPCC SR1.5 scenarios, which contain hardly any hydrogen produced via electrolysis. Most of all, this comparison demonstrates that the awareness of hydrogen’s critical importance for climate change mitigation and potential for technological learning is just emerging and already acknowledged by the IEA and IRENA. However, this awareness has not yet penetrated into the IAM community, even though electrolysis capacity deployment levels in climate mitigation scenarios have increased from the IPCC SR1.5 to the IPCC AR6 globally, as well as from the IPCC AR6 to the more recent EU-focused INNOPATHS scenarios. We believe that our analysis can help to parameterise plausible expansion pathways of green hydrogen for climate change mitigation scenarios in IAMs. Note that in the long run the technology diffusion pathways are asymptotically constrained by the exogenously assumed final market volume (see Table 1). The long-run capacity levels achieved in reality may exceed this level. The box plots show the median (50% quantile) as the centre, the 25% and 75% quantile as the bounds of box, and the whiskers as the minima and maxima of the full data set. Data sources: refs. ^{9,10,63,64,65,66,67}.