The degree-day method is deceptively simple. Heating degree days count how far the average daily temperature falls below the heating base temperature, while cooling degree days count how far it rises above the cooling base temperature. Because they require little more than a thermometer and a threshold, degree-day models have become the workhorse of regional energy assessments, climate impact studies and integrated assessment models. But the threshold itself is anything but universal. In the United States, modellers conventionally use 18.3 degrees Celsius; the United Kingdom uses 15.5 degrees for heating and 22 degrees for cooling; China typically uses 18 and 24 degrees. Applying a single number across a continent ignores the reality that base temperatures shift with local climate, building insulation, occupant habits and wealth. The authors note that a one-degree error in the base temperature can add or remove roughly twenty cooling degree days per month in places like San Luis Obispo, California, and can distort the relationship between degree days and actual energy use in nonlinear ways.

The conventional way to pin down a region’s true base temperature is the energy signature method, which plots energy consumption against outdoor temperature and identifies the inflection point of the resulting curve, or the closely related performance line method, which fits energy use against calculated degree days. Both approaches have produced valuable local results, such as median base temperatures of 18.3 degrees for cooling and 17.4 degrees for heating across sixty Indian cities, or values of 15.42 and 21.18 degrees in Kerman, Iran. The catch is data. These methods demand high-resolution energy consumption records matched with meteorological observations, which exist mainly in wealthy, data-rich countries. For most of Africa, South America and Central Asia, no such records are available, so global assessments fall back on harmonized, one-size-fits-all thresholds that wash out regional differences.

The new study attacks this bottleneck with machine learning. The researchers assembled electricity and natural gas demand time series from 172 regions worldwide, spanning building-level smart meter data at second, minute and hourly resolutions as well as regional daily and monthly statistics. From these they distilled training targets using segmented linear regression, a statistical technique grounded in the energy signature method that fits straight lines to different portions of the energy-temperature relationship and locates the breakpoints where heating or cooling behaviour begins. Rigorous quality control followed: time series without seasonal cycles were discarded, abnormal daily records were removed with density-based clustering, and only regressions achieving a coefficient of determination above 0.5 with physically plausible thresholds between 10 and 30 degrees were retained. The result was a labelled set of cooling and heating base temperatures for 131 geographic units across 24 countries.

Those labels then fed a Bidirectional Long Short-Term Memory neural network, or BiLSTM, equipped with an attention mechanism. LSTM networks are designed to preserve information across long sequences through gated memory cells, and the bidirectional variant processes weather records both forwards and backwards in time, capturing how past and future conditions jointly shape energy behaviour. The model ingested monthly meteorological variables from the ERA5-Land reanalysis, including two-metre air temperature, dew point temperature, wind speed and surface solar radiation, alongside annual socioeconomic data such as population density and economic density, plus geographic coordinates. An attention layer learned to weight the months most relevant to base temperature estimation, and stratified sampling based on geographic clustering ensured that training and test sets represented the full spread of world regions. An adaptive loss function allowed the model to train jointly on both thresholds even where only one was available for a given region.

The performance figures are strong. The model achieved root mean square errors of 1.39 degrees Celsius on the training set and 1.33 degrees on the test set, with Pearson correlation coefficients of 0.84 for cooling base temperature and 0.70 for heating. Benchmark comparisons against ridge regression, random forest and a unidirectional LSTM confirmed that the bidirectional architecture earned its place, particularly for the harder heating threshold, where it reached a cross-validated correlation of 0.63 against 0.52 or less for the alternatives. Applied globally, the trained model produced estimates for 3,385 first-level administrative regions, each roughly the size of a U.S. state or Chinese province. The predicted cooling base temperatures range from 19 to 25 degrees Celsius, with a mean of 22 degrees, while heating base temperatures span 14 to 18 degrees, averaging 16 degrees, values consistent with physical expectations and indoor comfort research.

The global map reveals telling patterns. North America and Europe, dominated by wealthy countries with well-insulated building stocks, show the lowest median thresholds for both heating and cooling, echoing earlier findings that affluent populations switch on air conditioning at lower outdoor temperatures. Asia, stretching across the widest latitudinal band and the largest economic disparities, exhibits the broadest range of values, with southern China and India showing higher thresholds. The predicted median heating base temperature for India of 17.05 degrees closely matches the 17.4 degrees derived independently from empirical energy signature analysis, a reassuring sign of internal consistency. Within Africa, the dataset even captures heterogeneity among countries such as Libya, Gabon, Algeria, Botswana and South Africa, which show relatively low heating thresholds compared with their neighbours.

Crucially, the team validated the dataset against 16 independent energy demand records from every major continent, none of which had been used in training. Direct comparison showed average prediction errors of 0.68 degrees for cooling and 1.08 degrees for heating. Indirect validation, in which degree days computed from the predicted thresholds were regressed against actual energy consumption, delivered the headline result: the average correlation between degree days and energy demand rose from 0.67 to 0.81, and the root mean square error of both cooling and heating demand models fell by approximately 10 percent compared with official or empirical base temperatures. In a South Australian home, using the predicted thresholds cut the cooling model error by 21.2 percent; in Tacoma, Washington, the regional heating model error dropped 12.85 percent against the official 18-degree standard. Sensitivity analysis confirmed that the predicted thresholds sit close to the maximum achievable correlation in nearly every dataset tested.

The authors are candid about limitations. The dataset rests on electricity and natural gas statistics, so heating from coal, diesel or biomass in energy-poor regions remains invisible, and building characteristics such as age, insulation and floor area are not differentiated. Uncertainty, quantified through 30 Monte Carlo dropout runs, ranges from 0.56 to 2.26 degrees Celsius, with the highest values concentrated in sparsely sampled regions of Africa, South America and western Asia, where training coverage is thinnest. Even so, relative uncertainty hovers around 10 percent for both thresholds, and the team shows the labels are robust to the choice of fitting threshold and to an independent UK Met Office-based breakpoint method, which agreed with their estimates to within 0.83 degrees on average across twenty British regions.

The implications reach well beyond energy statistics. Because degree-day models underpin projections of how climate warming will shift demand between heating and cooling, more accurate thresholds translate directly into better climate impact assessments, heat exposure and heatwave warning studies, and protection strategies for vulnerable populations. The dataset is also designed for coupling into Earth system models such as CESM and integrated assessment models such as GCAM and IMAGE, where it could reduce long-standing uncertainties in climate-energy interactions and sharpen the decarbonization pathways that policy makers depend on. Freely available on figshare, the first global map of the temperatures at which humanity reaches for the thermostat may prove one of those quiet datasets that quietly rewrites how the energy transition is planned.

Subject of Research: A global machine-learning dataset of heating and cooling base temperatures for residential building energy demand modelling

Article Title: A global base temperature dataset for residential building energy demand modelling

Article References: He, X., Eom, J., Yu, S., Liu, S., Xu, W., & Zhou, Y. (2026). A global base temperature dataset for residential building energy demand modelling. Earth System Science Data, 18(10), 7417-7454. https://doi.org/10.5194/essd-18-7417-2026

Image Credits: AI Generated

DOI: 10.5194/essd-18-7417-2026

Keywords: base temperature, degree days, building energy demand, BiLSTM, machine learning, heating and cooling, energy modelling, climate change, ERA5-Land, decarbonization, Earth System Science Data, thermal comfort

The methodological heart of the study is the China Integrated Sustainable Power System Optimization model, known as CISPO, which integrates land-use constraints and renewable resource assessments directly into a capacity-expansion and hourly-dispatch framework. The researchers simulated five-year intervals from 2030 to 2060 across three contrasting land-supply scenarios, ranging from an open regime in which developers can access most technically suitable terrain to a conservative regime reflecting strict protections for cropland, forests, grasslands and water shorelines. These scenarios draw on China’s actual regulatory landscape, including national rules governing the use of forest land for photovoltaic stations and wind farms, the Grassland Law, the Land Administration Law, and guidelines restricting construction along rivers and lakes. By embedding these real policy boundaries into the optimization, the team captured how legal and environmental restrictions, not just physical geography, determine where wind and solar can actually be built.

One of the study’s most striking conceptual contributions is the treatment of land as a spatially heterogeneous production factor with a quantifiable marginal value. The researchers extracted the shadow prices of the variable renewable energy capacity constraints in each grid cell, effectively asking how much the whole power system would save if one more unit of renewable capacity could be squeezed into a particular location. The results reveal a dramatic geographic asymmetry. High-value land is concentrated in eastern and central China, close to the demand centers where electricity is consumed. Siting additional renewables there reduces the need for expensive long-distance transmission lines and for overbuilding remote generation in the resource-rich but demand-poor west. In effect, a square kilometer of suitable land near Shanghai or Wuhan is worth far more to the system than a comparable parcel in the sparsely populated northwest, even if the northwest enjoys better sunshine and stronger winds.

When stringent land-use restrictions are imposed, the optimization model responds in a predictable but consequential way: renewable deployment migrates. With the most accessible land near demand centers placed off limits, the model shifts wind and solar construction toward the resource-rich frontier regions. Northwestern solar deployment increases by approximately 48 percent relative to the base scenario by 2060. This relocation is not free. Moving generation farther from load centers raises inter-regional transmission requirements by roughly 20 percent in 2060, meaning China would need to build substantially more high-voltage corridors to carry electricity from the remote interior to the coastal megacities. Transmission lines themselves consume land, face their own permitting hurdles and add capital cost, creating a feedback loop in which land scarcity in one domain generates infrastructure pressure in another.

Despite these frictions, the study delivers a reassuring headline: carbon neutrality remains technically achievable even under the most restrictive land assumptions. The model finds feasible pathways to a zero-carbon power system in all scenarios, meaning land constraints do not derail decarbonization. What changes is the price tag and the configuration of the system. The 4.1 percent cost penalty, while modest relative to the enormous overall investment required for full decarbonization, represents tens of billions of dollars per year by 2060, resources that could otherwise fund storage, grid modernization or electrification of other sectors. The finding reframes land policy as an economic lever: decisions about which parcels are open to renewable development directly shape the national cost of climate mitigation.

The study also connects land scarcity to electricity prices, an issue of growing political salience as China deepens its power market reform. Using marginal-cost-based price formation, the researchers show that the cost differences across land-use scenarios could become visible in market-based electricity prices, depending on market design and cost-recovery mechanisms. Recent reforms, including the 2025 policy notice on deepening market-oriented reform of on-grid tariffs for new energy, are pushing renewables toward full market participation. If land restrictions systematically raise the marginal cost of supply, consumers and industries in land-constrained regions could face persistently higher prices, while the benefits of cheap remote renewables are eroded by transmission charges. This linkage between land policy and retail electricity costs is likely to become a central tension in China’s energy governance.

The research arrives amid a global reckoning with the land intensity of renewable energy. Solar and wind are far less land-efficient per unit of energy than fossil or nuclear plants, and studies from the western United States to Brazil have documented mounting conflicts between renewable buildout and agriculture, biodiversity and community interests. In China, a surprising share of existing photovoltaic capacity has been installed on cropland, prompting regulatory tightening. The new study quantifies, for the first time at national scale, what such tightening costs the power system, transforming an abstract land-use debate into concrete economic terms. It also highlights the value of land-saving alternatives that appear in the literature, including rooftop photovoltaics, floating solar on reservoirs, agrivoltaics that combine farming with electricity generation, and photovoltaic deployment on degraded or industrial land such as open-pit mines.

For policymakers, the implications cut in several directions. First, the shadow-price analysis offers a spatial prioritization tool: protecting the highest-value land near demand centers imposes the largest system costs, so land-use decisions in eastern and central China deserve the closest scrutiny. Second, the 20 percent increase in inter-regional transmission needs under restrictive scenarios argues for proactive grid planning, since transmission projects take years to permit and build. Third, the modest overall cost penalty suggests that China has genuine flexibility; even a conservative land regime does not make decarbonization prohibitively expensive, provided the system can re-optimize around the constraints. The authors’ framework, with data and code publicly available on GitHub, allows other researchers and planners to test alternative land policies, resource assumptions and technology costs.

The study’s scope is limited to the power sector, and the authors note that cost differences will surface in market prices only as reform deepens and depending on how costs are recovered. Extending the analysis to heating, hydrogen production and electrified transport, all of which will compete for the same renewable generation, is an obvious next step. Yet the core message is already clear and widely applicable: as the world pivots from whether to decarbonize to how, the humble question of where to put the panels and turbines has become a first-order economic problem. China’s experience suggests that every country plotting a net-zero grid should start counting not just its terawatt-hours but its square kilometers, because in the land-energy nexus, geography is destiny and scarcity has a price.

Subject of Research: Land availability constraints on China's transition to a carbon-neutral power system

Article Title: Navigating the land–energy nexus for carbon-neutral power systems in China

Article References: Zhu, Z., Yu, R., & Zhang, D. (2026). Navigating the land–energy nexus for carbon-neutral power systems in China. Nature Sustainability. https://doi.org/10.1038/s41893-026-01941-5

Image Credits: AI Generated

DOI: 10.1038/s41893-026-01941-5

Keywords: China, carbon neutrality, power system optimization, land use, renewable energy, solar power, wind power, transmission, shadow prices, electricity markets, Nature Sustainability, energy modelling