Does environmental carbon pressure lead to low-carbon technology innovation? Empirical evidence from Chinese cities based on satellite remote sensing and machine learning

Climate change has become the most significant environmental issue facing human society in the 21st century, and a low-carbon transition is imminent. In order to comprehensively evaluate the balanced relationship between carbon activities and ecological environment under different regional scenarios. Opening up the black box of low-carbon technological innovation in Chinese cities under the pressure of climate change. This work combines the characteristic forecasting of random forest with the ordinary least square model to assess the contribution value and effects of various characteristic variables. The findings demonstrate that (1) China's carbon pressure level has changed from equilibrium to overload. The number of carbon high-pressure cities is increasing year by year. Carbon high pressure cities have risen from 12 to 68 in 20 years. (2) The carbon pressure on the city is directly influenced by the secondary industrial production value and the land urbanization ratio. Nevertheless, the level of environmental regulations imposed by municipal governments remains rather consistent, irrespective of the degree of urban carbon emissions. (3) Urban carbon pressure has a direct impact on the development of low-carbon technologies, with characteristic contributions of 0.91, 0.52, 0.43, and 0.3 at various pressure levels, respectively. (4) Population size and the level of economic development have a significant positive impact on the promotion of low-carbon technological innovation. The government should give full play to the talent and technology advantages of the core city clusters, and create a technological innovation alliance between the city clusters through the scale effect of economic factors. This study helps determine the correlation between carbon pressure and low-carbon technological innovation, and helps provincial and local governments identify their own carbon pressure status and choose differentiated emission reduction models.