Increased energy use for adaptation significantly impacts mitigation pathways

Abstract

Climate adaptation actions can be energy-intensive, but how adaptation feeds back into the energy system and the environment is absent in nearly all up-to-date energy scenarios. Here we quantify the impacts of adaptation actions entailing direct changes in final energy use on energy investments and costs, greenhouse gas emissions, and air pollution. We find that energy needs for adaptation increase considerably over time and with warming. The resulting addition in capacity for power generation leads to higher greenhouse gas emissions, local air pollutants, and energy system costs. In the short to medium term, much of the added capacity for power generation is fossil-fuel based. We show that mitigation pathways accounting for the adaptation-energy feedback would require a higher global carbon price, between 5% and 30% higher. Because of the benefits in terms of reduced adaptation needs, energy system costs in ambitious mitigation scenarios would be lower than previous estimates, and they would turn negative in well-below-2-degree scenarios, pointing at net gains in terms of power system costs.

Fig. 1. Integrated approach to the adaptation-energy feedback loop.

The circle represents the integrated framework of the World Induced Technical Change Hybrid (WITCH) model, linking the economy, the energy system, and the climate system. Red lines indicate the components modified with new equations to model the adaptation-energy feedback loop. Top-left panel: eight clusters characterize the heterogeneity in the relationship between the Extreme Temperature Indicators (ETIs) and annual average temperature across world regions. Top-central panel: semi-elasticities as estimated in ref. representing the percentage change in energy demand for one additional day with average daily temperature (T) in the upper (T > 27. 5 °C)/lower (T < 12. 5 °C) bin, see ref. . A detailed description of each step and of the methodological advancements is presented in the Methods and Supplementary Methods. The WITCH model version used for the analysis (WITCH 5.0) is described in detail in ref. .

Fig. 2. Future changes in the frequency of warm and cold days.

a Difference (Δ) between future (2090–2100) and historical (2005) annual number of days with average daily temperature (T) > 27.5 °C and T < 12. 5 °C. b Regional count of total days with T > 27. 5 °C and T < 12. 5 °C in 2005 and in 2100 by policy scenario. Temperature indicators are constructed with population-weighted daily temperatures. Scenarios: Current policies (C.Pol) and well-below 2 °C (W.b. 2 °C).

Fig. 3. Projected electricity and fuel (including liquids and gases) demand for adaptation under Shared Socioeconomic Pathway (SSP) 2 assumptions.

a Annual global average demand from 2020 to 2100 across the different scenarios excluding (dotted) and including (solid) the adaptation-energy feedback under the SSP2. b Regional final energy demand in 2100. Light bars show the value excluding the adaptation-energy feedback, while dark stacked bars show the positive or negative variation in energy demand induced by the adaptation-energy feedback. Labels in panel b show the regional percentage increase. Scenarios: Current policies (C.Pol), 2. 5 °C and well-below 2 °C (W.b. 2 °C).

Fig. 4. Additional power generation capacity.

a Technology mix of the additional average annual capacity to fulfill the additional energy for adaptation. b Additional fossil-based new capacity installed cumulatively with (solid lines) and without (dotted lines) the adaptation-energy feedback. The additional new capacity installed cumulatively including also renewable sources is presented in Supplementary Fig. 6. The technologies unaffected by the adaptation-energy feedback are not included. Scenarios: Current policies (C.Pol), 2.5 °C and well-below 2 °C (W.b. 2 °C).

Fig. 5. Annual electricity system costs by scenario.

a Total electricity system costs in trillion $, 2005 Purchasing Power Parity (PPP). b Additional electricity system costs in the mitigation scenarios with respect to the current policy, in trillion $(2005, PPP). c Variation in the cumulative electricity system costs associated to the more ambitious mitigation policy scenarios with respect to the current policy, in trillion $(2005, PPP). All projections are presented alternatively for the case with (solid lines) or without (dotted lines) adaptation. Operative fuel expenses for fossil-based power generation are included in the electricity system costs. Scenarios: Current policies (C.Pol), 2.5 °C and well-below 2 °C (W.b. 2 °C). Results presented in panel a and b for the scenario 2 °C are not shown to avoid clutter and can be fund in the Supplementary Material.

Fig. 6. Regional variation in greenhouse gas (GHG) emissions and air-pollutants.

a Variation in air-pollutants by region. Average total annual increase between 2020 and 2100 in the following air-pollutants: black carbon (BC), nitrogen oxides (NO_x), carbon monoxide (CO), sulphur dioxide (SO₂), organic compounds (OC), volatile organic compounds (VOC). b Additional cumulative GHG emissions for adaptation in 2100, total (left) and per capita (right). Scenarios: Current policies (C.Pol), 2.5 °C and well-below 2 °C (W.b. 2 °C).