Fig. 5
From: When optimization for governing human-environment tipping elements is neither sustainable nor safe
Human-environment systems in paradigms classification. For risky (a) and (b) cautious policy here shown in model parameter space of collapse probability δ versus recovery probability ρ. Color indicates the paradigms combination similarly as in Fig. 3. Here, additional gradual changes between the color regimes indicate the probability of being in a certain paradigms combinations regime under parameter uncertainty ranges. Remaining parameters where chosen linearly within the range of 0.95 ≤ γ ≤ 0.99, 0.3 ≤ rl/rh ≤ 0.7, 0.1 ≤ rmin/rh ≤ 0.5. The approx. transition probabilities δ and ρ were assigned to the human-environment systems climate, fisheries and farming agriculture according to the timescale of the average time spent in one state (see Methods). For farming, a risky policy is likely to be optimal but neither sustainable nor safe. The parameter uncertainty of the other parameters does not allow a clear statement in which parameter regime fisheries are likely to fall. The climate system may lie at the edge of the sweet spot, where all paradigms yield the cautious policy. However, for smaller collapse probability δ optimization is more likely to yield the risky policy, which becomes also neither sustainable nor safe at this point. This suggests the use of other paradigms for climate policy making
