How can climate adaptation policy enhance urban climate resilience? Empirical evidence from China

PLOS One | https://doi.org/10.1371/journal.pone.0335736 November 6, 2025 6 / 21 H5: The CACCP policy effect can be optimized by reducing the dependence on mineral resources. 3. Methods and materials 3.1. Overview of the research area The Yellow River is China’s second-longest river, spanning the country’s eastern, central, and western regions. Its total basin area is approximately 795,000 square kilometers. The Yellow River flows through nine provinces, including Qinghai, Sichuan, Gansu, Ningxia, Inner Mongolia, Shaanxi, Shanxi, Henan, and Shandong, covering 69 cities [48]. Among them, 36 are mineral resource cities, making the region a significant energy hub in China. However, the coal energy industry dominates the region’s economic development, which has led to the dual challenges of industrial transformation and carbon emission reduction. At the same time, the Yellow River Basin has a fragile ecological environment, characterized by frequent regional droughts and floods in recent years and a significant increase in extreme weather events [49]. The Yellow River Basin is identified as one of the key strategic regions in the National Climate Change Adaptation Strategy 2035. It proposes action measures such as “comprehensively implementing water conservation and control actions in the Yellow River Basin, strengthening ecological restoration and governance, promoting networked, integrated monitoring and assessment of climate change, fine-tuning the forecast, and constructing an integrated transportation network, infrastruc-

ture, and public service system.” This study assesses the potential of the CACCP policy to enhance urban climate resil- ience in the Yellow River Basin. Specifically, it provides insights pertinent to climate adaptation strategies in ecologically vulnerable regions. For the analysis, cities exhibiting substantial missing data from 2010 to 2022 were excluded. Conse- quently, the final sample comprises 57 cities, including nine pilot cities and 48 non-pilot cities. 3.2. Methods 3.2.1. Policy effect model. To examine the impact of CACCP policy on urban climate resilience in the Yellow River Basin, a model is constructed using panel data from 57 cities from 2010 to 2022, with 2017 as the year of policy implementation, as shown in Eq (1). Resilience it = α0 + βtreated it + 3 ∑ j=1 γj X it + u i + λt + εit (1) Where i denotes city and tdenotes year. Resilience is the outcome variable representing urban climate resilience in this study. treated represents a dummy variable for policy treatment. When the city is in the experimental group, it is assigned a value of 1 after the policy is implemented and a value of 0 before it is implemented. When the city is in the control group, it is assigned a value of 0. X represents the control variables. β is the estimated coefficient of the policy effect we are interested in, i.e., the average treatment effect. α0 is a constant term. γj is the coefficient of the control variable. u i and λt denote the city and year fixed effects, respectively. εit is a random error term.

Given that the Synthetic DID method improves upon the traditional synthetic control method (SCM) and the difference- in-difference method (DID), its estimators exhibit better precision and robustness, particularly in scenarios involving small treatment group sizes [50]. In this study, with nine treatment groups and 48 control groups, we selected the Synthetic DID method for policy effect estimation. This method primarily consists of three steps: first, identify the individual weighting variable ˆωsdid i such that the pre-treatment trend of the outcome variable in the control group is as close as possible to that in the treatment group; second, identify the time weighting variable ˆλsdid t , which is used to weight the outcome vari- ables prior to the event to match them with the outcome variables post-event; finally, incorporate ˆωsdid i and ˆλsdid t into the DID two-way fixed effects model to estimate the average treatment effect, yielding the Synthetic DID estimate, as shown in Eq (2).