Frozen weather baselines distort cooling energy demand and emissions

Air conditioners whirr in millions of homes, and we tend to accept cooling energy demand as a non-negotiable price of comfort. The common assumption is that tomorrow will resemble today, only with a slightly higher bill.

The same mindset shows up in the models governments rely on when preparing for a hotter world. Many of these tools hold the weather constant at temperatures from decades ago. A new analysis argues that this convenience quietly skews the results.

Frozen weather baseline

A number of major climate-planning models take an unusual approach when estimating how much energy buildings will consume. Even while global temperatures rise, they keep heating and cooling needs tied to climate conditions from decades past.

This decision was not made casually. It started as a practical method for separating human behaviour from the variability of a changing climate, and because integrating detailed weather into large-scale economic models is difficult.

Because buildings already account for nearly one third of global energy use, that simplification influences a large share of the projections. Yang Ou, a researcher at Peking University (PKU), and colleagues set out to unfreeze the baseline.

To do so, they inserted updated temperature projections into one of the few global models that explicitly links building energy use to outdoor heat and cold.

Cooling energy demand gap

The difference is substantial. When the model sticks to historical temperatures, it underestimates cooling demand by 23% under modest warming. Under a high-emissions pathway, the shortfall grows to 79% by 2100.

Heating moves in the opposite direction. If winters are assumed to be as cold as the historical record, the older method overstates heating demand by up to 14% in the gentler scenario and 40% in the harsher one.

Most previous research in this area focused mainly on cooling. By modelling heating and cooling together, the team captured effects that earlier studies overlooked.

These two pressures work against each other. Leaving either out distorts the overall projection.

A surprise in emissions

The counterintuitive finding is this: once actual warming is incorporated, projected carbon dioxide emissions from buildings fall rather than rise.

The decline in demand for furnaces and gas heating more than offsets the additional electricity required for cooling. By 2100, the estimated reductions range from 83 million to 1.6 billion metric tons of carbon dioxide per year.

At the upper end, that cut is close to 4% of today’s total emissions. Earlier work had already suggested how strongly warming can reshape energy demand.

That may sound like good news, and in a narrow sense it is. However, carbon dioxide is not the only relevant gas. Cooling depends on refrigerants, which themselves trap heat.

Hidden refrigerant cost

As air conditioning becomes more widespread, leaks of F-gases increase as well. These refrigerant chemicals inside cooling systems warm the planet far more than carbon dioxide, though in milder futures their overall effect remains limited.

In the hottest future, that warming impact expands into a serious issue that older models did not highlight. They largely missed it because they underestimated how much cooling capacity the world would add.

This blind spot is most damaging in fast-growing, hot countries across the Global South, where separate research on developing economies has mapped the gap in who can afford to keep cool.

So the improved carbon dioxide picture comes with important caveats: more accurate heating estimates on one side, and a growing refrigerant problem on the other. The second risk is easy to overlook if carbon dioxide is the only metric counted.

Why geography decides

When you examine individual countries, the global headline breaks apart. Colder, wealthier nations such as the United States, China, and Russia benefit because warmer winters mean less fuel is burned for space heating.

Hotter regions face the reverse. Across Southeast Asia, the Middle East, and South Africa, rising temperatures push cooling demand upwards. Where electricity supply still depends largely on fossil fuels, emissions increase rather than decrease.

The Middle East shows some of the largest cooling-demand gaps of all. That divergence sits at the core of the warning.

A model built around a single global average can capture gains in a cold country while missing a growing burden in a hot one. That is exactly the kind of error that can send investment and policy to the wrong places.

How planners can adapt

The remedy is straightforward. Planning tools should use current and projected temperatures, region by region, rather than relying on a historical average that no longer matches real-world conditions.

Countries confronting rapidly rising heat can respond with stronger building standards, cleaner power grids, and faster deployment of efficient cooling - from heat pumps to materials designed to shed heat naturally.

Places experiencing milder winters can reduce their reliance on fossil-fuel heating instead. What has changed is the picture itself: the evidence now shows that historical temperature baselines misjudge both the scale and the location of future demand.

The next generation of strategies can be built around the climate we are moving into, not the one we have left behind.