Design codes: The Trojan Horse of climate resilience

The impacts of climate change are impossible to avoid. The UN chief, António Guterres, has acknowledged it is now “inevitable” that humanity will overshoot the target set in the Paris climate agreement, and the UK's Climate Change Committee has told the government to prepare for 2°C of global temperature rise by 2050.

So, the question shifts from climate mitigation alone to how we navigate the changes that are here and accelerating. This article does not address the ongoing efforts to limit climate change, although it remains a critical goal for humanity. It focuses on how we can endure climate change and navigate the shocks ahead.

We all know the story of the Trojan Horse; the Greek myth of an innocuous wooden horse used to infiltrate and ultimately destroy the city of Troy. The Trojans welcomed the horse, rolling it within the city walls and underestimating the threat that lay within. Our industry does something similar with design codes. We embrace them, give them pride of place, and assume resilience is factored in. But the horse holds surprises: stationary assumptions, update cycles that lumber, fragmented hazard treatments, and a habit of smoothing over compound risk. If we fail to pull off the panels and look inside, we risk building tomorrow’s failures today.

In this article, we look at the risks, the problems hiding in today’s codes, and what can work instead.

Where risk creeps in – How we use climate futures

Making future predictions about the climate is challenging. Climate modelling is complex, and outcomes depend on socioeconomics, technology, policy, and land-use. To understand the range of futures that could occur, climate practitioners use scenarios developed by the Intergovernmental Panel on Climate Change (IPCC). These Shared Socioeconomic Pathways (SSPs) are commonly summarised as:

• SSP1-1.9 (Sustainability): Very low emissions; closest to a 1.5°C world by 2100.
• SSP1-2.6 (Sustainability): Low emissions; approximately 2°C by 2100.
• SSP2-4.5 (Middle of the road): Intermediate path; emissions plateau then decline but not to net zero by 2100.
• SSP3-7.0 (Regional rivalry): Medium-high emissions with fragmented geopolitical progress.
• SSP5-8.5 (Fossil-fuelled development): High emissions driven by intensive fossil fuel use.

Given this range, codes and regulations can – and should – explicitly reference scenarios, time horizons, and asset design lives. There are many plausible futures, so practitioners test performance across multiple scenarios and periods (e.g., 2030s, 2050s, 2080s), not just a single point estimate. Due to uncertainty and the long life of assets, a range-based approach with adaptive strategies is essential.

The Trojan Horse problems hiding in today’s codes

Design codes are meant to keep people and infrastructure safe. Yet in a rapidly changing climate, some of the very assumptions built into today’s standards can create blind spots. Hidden within them are vulnerabilities that quietly increase exposure to future climate risks. These “Trojan Horse” issues are not obvious at first glance, but they sit inside many of the codes we rely on every day.

1.Most standards use historical data and statistics

Many design actions are derived from historical observations under an assumption of stationarity, that yesterday’s climate is a good guide to tomorrows’. For instance, wind, snow, rainfall, and temperature loads often use historic return periods (e.g., 1-in-50-year events) calculated from past datasets. In a non-stationary climate, those return periods shift and what used to be “rare” can become regular within the life of the asset.

2.There is a lag between standards and accelerating climate risks

Standards bodies do revise codes, but the cycle of research, consensus, drafting, consultation, adoption, and enforcement can take years. Climate hazards are evolving faster than many update cycles. Even where forward-looking guidance exists, adoption by jurisdictions can lag, creating a patchwork of requirements where resilience is often optional rather than prescriptive.

3. Projections are not granular enough

Design needs local, asset-relevant variables, such as rainfall intensities, wet-bulb temperatures, compound flood drivers (pluvial + fluvial + coastal), heat island effects, wildfire spread, and drought persistence. Many projections are coarse in space and time, and this downscaling through codes adds uncertainty and risk.

4. Cherry-picking climate scenarios and horizons

A common pitfall is mixing aggressive scenarios with near-term horizons or vice versa and claiming robustness. For example, using SSP5-8.5 for the 2030s may look conservative, but the spread across SSPs is relatively small up to about mid-century; the divergence explodes beyond 2050. If an asset’s life extends to 2080 or beyond, a single “2030s high scenario” check does little to inform the end-of-life performance. Conversely, using a low scenario for a long-lived critical asset can understate tail risks.

5. Fragmented treatment of compounding and cascading risks

Standards tend to silo hazards: wind here, snow there, heat somewhere else. Real-world failures often involve compounding drivers (e.g., storm surge plus fluvial flow plus heavy rainfall) and cascading impacts (power outage → cooling failure → indoor overheating). Codes rarely require multi-hazard, compound-event stress tests across future climates.

Why the traditional approach no longer works

In a rapidly changing climate, long-held assumptions begin to break down. Historical return periods no longer reflect the frequency or intensity of future hazards. Assets designed to last 50 to 100 years will face late-century conditions where extremes and scenario divergence matters most. Relying on traditional safety factors cannot compensate for shifts in hazard patterns, nor for the socio-technical complexity of modern systems, where small errors can trigger cascading failures across infrastructure networks. In short, the old model of designing for yesterday’s climate is no longer fit for purpose.

What a better approach looks like

A more resilient design culture is possible, but it requires embedding future climate conditions directly into our standards and processes.

1. Make future climate explicit in codes
This requires the use of multiple SSPs, time horizons, and local guidance aligned to design life, while documenting why those scenarios, sources, and assumptions were chosen.

2. Move toward performance and resilience-based design
We should define the levels of service and survivability an asset must maintain under future extremes.

3. Improve granularity and traceability
Regionally downscaled datasets and, where needed, bespoke modelling can translate climate projections into engineering-relevant variables, complete with uncertainty ranges and sensitivity tests.

4. Stop cherry-picking climate scenarios
Long-lived assets should be designed for at least a 2°C-consistent 2050 world, with stress-tests against higher warming pathways (3–4°C) and extreme tails.