Can integrated ecological restoration mitigate urban heat? Evidence from China

China’s most ambitious ecological restoration program, the Shan-Shui Initiative, appears to be gently cooling the country’s cities, according to a new analysis that tracks land-surface temperatures across 291 prefecture-level cities from 2006 to 2023. The study, published in Regional Environmental Change, is among the first to test at national scale whether landscape-scale restoration—rehabilitating mountains, rivers, forests, farmland, lakes and grasslands as interconnected systems—can meaningfully alter the thermal climate of urban areas. The answer, the researchers find, is a cautious yes: the effect is real and statistically robust, but modest, amounting to a reduction in the urban heat island of roughly three hundredths of a degree Celsius.

The urban heat island, the phenomenon in which cities run hotter than their rural surroundings, has long been recognized as one of the most tangible consequences of urbanization. Concrete, asphalt and other impervious surfaces absorb and re-emit solar radiation, while sparse vegetation reduces the evaporative cooling that shading and transpiration normally provide. Decades of research, from Oke’s classic energetic account of the heat island to recent global satellite surveys, have documented how surface temperature differences between cities and their hinterlands amplify heat stress, raise cooling energy demand and exacerbate health risks during heat waves. What has been far less clear is whether large-scale ecological restoration programs—policies aimed primarily at biodiversity and ecosystem services—produce measurable thermal dividends for the cities embedded within restored landscapes.

To answer that question, the research team, led by Jiaxing Ren of Northeast Forestry University together with Hong Chen and Jiyue Zhang, exploited the staggered rollout of the Shan-Shui Initiative’s first three pilot waves. Because different cities entered the program in different years, the researchers could apply difference-in-differences methods in the modern, heterogeneous-treatment-effect framework, combining event-study designs, cohort-specific estimates, and double machine learning techniques to isolate the policy’s effect from background trends in urbanization and climate. Rather than relying on a single temperature metric, the team constructed a 1-kilometer dynamic equal-area urban–rural land-surface temperature difference as their primary outcome, and benchmarked it against conventional daytime and nighttime surface urban heat island measures derived from MODIS satellite observations.

The headline result is a preferred estimate of a 0.0325 °C reduction in the urban heat island intensity attributable to the initiative. When the team decomposed this effect by time of day, daytime surface urban heat islands declined by 0.0846 °C and nighttime ones by 0.0374 °C. Event-study and cohort-specific estimates both point to a post-policy decline in urban heat, and the effect survives a battery of robustness checks. The coefficient remains negative and statistically significant after 5 percent winsorization of the data and after adding city-specific linear trends, although its magnitude is attenuated under those stricter specifications, suggesting some of the raw estimate may reflect pre-existing local trajectories.

A central concern in any policy evaluation is whether the estimated effect is genuinely caused by the intervention or simply by pre-trends—cities that joined the program were perhaps already cooling for other reasons. To address this, the researchers conducted date-shift placebo tests, artificially moving the policy’s adoption date two and three years earlier in their statistical model. The placebo estimates were negative but statistically insignificant, providing no evidence of a discrete cooling response before formal adoption. In other words, the data show no sign that treated cities were on a distinct cooling path prior to enrollment, strengthening the case that the observed temperature declines coincide with the restoration program itself.

Perhaps the most practically useful finding concerns who benefits most. The cooling effect was stronger in cities that started with high pre-policy urban heat island intensity, higher levels of urbanization, greater vegetation cover, and lower concentrations of PM2.5 particulate pollution. This heterogeneity makes physical sense: cities with more existing green infrastructure have more vegetation available to expand evapotranspiration and shading, while the most overheated cities have the greatest thermal margin for improvement. The finding that lower air pollution amplifies the benefit also hints at interactions between aerosols and surface energy budgets that could reward cities pursuing air quality and greening goals in tandem. For urban planners, the message is that restoration investments are likely to pay the largest thermal dividends when they build on, rather than substitute for, existing green cover.

The study also ventures into institutional territory, asking whether the initiative changed local economic and policy environments. The analysis finds that participation in the program was associated with increases in green finance, human capital, and exposure to climate-policy uncertainty. Product-of-coefficients estimates linking these three institutional responses to the temperature outcome were negative and statistically significant, but the authors are careful to characterize them as exploratory indirect associations rather than causal mediation. In plain terms, the data are consistent with a story in which restoration programs attract green capital, skilled workers and heightened climate-policy activity, and these channels plausibly reinforce cooling—but the evidence cannot yet prove that these mechanisms carry the effect.

The authors frame their conclusions with deliberate restraint. The results, they write, support a modest city-scale cooling effect rather than a comprehensive climate-adaptation effect. That distinction matters. A reduction of a few hundredths of a degree in a city-scale temperature differential, while encouraging, is small compared with the multiple degrees of urban–rural contrast documented in many Chinese cities, and far smaller than the temperature increments projected under global warming. The Shan-Shui Initiative was never designed as a cooling program; its primary goals concern ecosystem integrity and biodiversity. Thermal benefits, on this evidence, are a genuine but secondary co-benefit—real enough to register in satellite data across hundreds of cities, but not a substitute for dedicated heat-mitigation strategies such as reflective materials, urban ventilation corridors or targeted tree canopy expansion.

The methodological contribution may prove as influential as the substantive one. By combining a dynamic equal-area urban–rural temperature metric with staggered-adoption causal inference and benchmarking against standard MODIS measures, the study offers a template that other countries with large restoration programs—among them China’s Grain for Green afforestation efforts, Africa’s Great Green Wall, and large-scale reforestation initiatives elsewhere—could adopt to audit whether ecological investment translates into urban climate benefits. The funding for the work came from the National Social Science Foundation of China and from Xinjiang University of Finance and Economics, and the authors report no competing interests. As cities worldwide confront intensifying heat, the study’s central lesson is measured but encouraging: restoring landscapes at scale does seem to nudge urban thermometers downward, and the effect is largest where restoration builds on the green foundations cities already possess.

The thermal dividend the study documents, though small, aligns with a body of experimental and observational work on how vegetation cools cities. Research on tree canopy in the United States has shown that the cooling benefit of green cover is scale-dependent: scattered trees can actually worsen daytime heat in some neighborhoods by shading the ground while blocking airflow, whereas large contiguous canopy patches paired with reduced impervious surface deliver measurable relief. That finding helps explain why the Shan-Shui effect concentrates in cities with greater existing vegetation cover, where restoration likely expands connected green infrastructure rather than isolated plantings.

The choice of measurement also matters for interpreting the results. Satellite-derived surface urban heat islands are known to be sensitive to observation conditions; recent global assessments have shown that clear-sky satellite observations can overestimate surface heat island intensity in humid cities, and that daytime and nighttime patterns diverge because surface and air temperatures respond differently to solar forcing. The study’s use of a dynamic equal-area urban–rural difference, benchmarked against conventional MODIS daytime and nighttime measures, reflects growing awareness in the remote-sensing literature that single-metric estimates can mislead cross-city comparisons.

The findings also connect to a broader evidence base on restoration outcomes. A widely cited meta-analysis in Science concluded that ecological restoration enhances biodiversity and the ecosystem services it underpins, but thermal regulation at city scale had rarely been quantified as a policy outcome. By treating temperature as a measurable dividend of a national restoration program, the study extends that literature from plot-level ecosystem function to city-scale climate conditions.

Finally, the modest magnitude of the effect carries practical weight for energy and health planning. Urban overheating raises cooling energy demand and amplifies heat stress during heat waves, and even small reductions in the urban–rural temperature differential compound across large populations. The evidence suggests restoration programs can contribute to thermal comfort portfolios, but as a complement to, not a replacement for, engineered heat-mitigation measures.

Subject of Research: Can integrated ecological restoration mitigate urban heat? Evidence from China

Article Title: Can integrated ecological restoration mitigate urban heat? Evidence from China

Article References: Ren, J., Chen, H., & Zhang, J. (2026). Can integrated ecological restoration mitigate urban heat? Evidence from China. Regional Environmental Change, 26(4), Article 187. https://doi.org/10.1007/s10113-026-02677-w

Image Credits: AI Generated

DOI: 10.1007/s10113-026-02677-w

Keywords: integrated, ecological, restoration, mitigate, urban, heat, Evidence, China, scientific research