Climate-Adaptation Readiness

What this driver is

Climate adaptation readiness is one continuing capacity mechanism: whether governments, utilities, and households can absorb heat, flood, drought, and storm without converting a physical hazard into a welfare loss. It is a mechanism, not a disaster-death score, not a COP slogan, and not a welfare index.

The driver is not climate crisis (the warming-and-harm bundle), not extreme heat exposure, not flood risk, not drought risk, not wildfire risk, and not climate-finance availability. Those measure the hazard, the money, or the warming stock. This node is readiness: plans, codes, early warning, and residual risk.

It applies globally from the Cancún adaptation framework and the Paris Agreement’s adaptation goal through the Global Goal on Adaptation (GGA) work under the UAE Framework and COP28/COP29 decisions. It is not ukraine-war and not territorial-revisionism.

The live Factrail series already uses Our World in Data “All disasters” deaths (EM-DAT lineage) as a disclosed, inverted proxy: a higher raw death year is stored as a lower readiness value. Stored points include 0.30110884 (2020), 0.32946483 (2021), 0.18629816 (2022), 0.1445205 (2023), 0.18284102 (2024), and 0.29764092 (2025). Those raw death years (15,607; 10,548; 76,244; 135,792; 79,974; 16,373) are OWID/EM-DAT figures, quoted with the chart. This page does not invent a unique Factrail death total or Contribution.size. A death series is a distant outcome proxy, not an operational definition of readiness.

Mechanism

Readiness moves welfare through named channels.

First, information. WMO’s Early Warnings for All initiative treats multi-hazard warning coverage as the first hour of adaptation. A warning that does not reach the last mile is not readiness.

Second, rules and codes. Building codes, heat-health plans, flood zoning, and drought allocation rules convert a forecast into fewer deaths and less asset loss. IPCC AR6 Working Group II is the assessment of those channels (https://www.ipcc.ch/report/ar6/wg2/ (opens in a new tab)).

Third, money that actually buys protection. Adaptation finance that stays in a pledge annex does not raise a levee.

Fourth, residual risk. Even high readiness leaves residual loss as hazards intensify under climate crisis. A later disaster-death spike can mean a worse hazard, a failed warning, or both. The inverted death proxy cannot tell those apart. That is why the proxy is disclosed as lineage, not as “readiness equals fewer deaths.”

Fifth, distribution. The same cyclone kills in an informal settlement and does not in a coded city. Age, disability, tenancy, and irrigation rights change who is ready.

Timing is hours for a warning and years for a code cycle. A later NAP, GGA indicator, or a sourced early-warning coverage series would be a better meter than inverted EM-DAT. Until then the live points stay.

Evidence base

The best evidence for the live meter is the OWID disaster-death CSV (https://ourworldindata.org/grapher/number-of-deaths-from-natural-disasters.csv (opens in a new tab)) and the stored inverted log envelope already on this driver. IPCC AR6 WG2 is the mechanism assessment. UNFCCC NAP and GGA texts are the legal neighbourhood. WMO Early Warnings for All is the operational warning programme (https://wmo.int/early-warnings-for-all (opens in a new tab)).

Evidence is strongest for a published death year and a named NAP. It is weaker for any unique “lives saved by adaptation” coefficient. It is weakest where a COP communiqué is asked to rebate 135,792 OWID deaths in 2023.

Effect on welfare indicators

This driver, when readiness actually holds, is a first-order channel into disaster mortality and a co-channel into heat and water harm.

• Under-five mortality: direction lower deaths when warning, water, and clinic protocols hold in the same population as the hazard. Horizon: days to years. Confidence: medium locally; the live meter is global EM-DAT, not under-five. No invented child-death size.
• Exposure to extreme temperature: direction lower harm when heat plans and cooling access exist. Horizon: the same heat season. Confidence: medium. Extreme heat exposure remains the hazard companion.
• Proportion of population using safely managed drinking water: direction higher access when drought and flood plans protect systems. Horizon: years. Confidence: low-medium.

Missing coefficients stay missing.

The welfare sign of genuine readiness is positive on those series. Treating a low-death year as proof of readiness, without a warning or code instrument, is the error this rewrite refuses.

Interactions and contested points

The mechanism interacts with climate crisis, climate-finance availability, extreme heat exposure, flood risk, drought risk, wildfire risk, water scarcity, and disaster-response capacity. Reinforcement is obvious when a heat plan and a warning sit in the same city. Offsetting is the claim that a death-year dip rebates a missing NAP.

The strongest competing view is that inverted EM-DAT is too remote to carry this driver at all. That objection is serious. The competing defence is only this: the node already exists, the proxy is disclosed, and an honest description is better than a template that pretends disaster deaths are readiness.

It does not. Crisis is the warming bundle. This node is the capacity to absorb it.

Established: live inverted OWID points cited above; distinction from heat/flood/drought/finance companions; IPCC WG2 as the assessment. Not established: a unique Factrail death total, a GGA numeric score, or that 2025’s higher stored readiness (lower deaths) means codes improved. A later early-warning coverage series would change the meter. It would not mint a second adaptation driver.