The spatial identification and intensity differentiation mechanism of carbon effect of urban residents 'energy use mode is an important content of the research on the low-carbon transformation of residents 'life, which is of great significance to promote the modernization construction of harmonious coexistence between man and nature. However, the fine-scale assessment of residential carbon effect is relatively weak, and the nonlinear mechanism of carbon emissions from residential energy use is not clear. Therefore, this paper used multi-source data such as night lights, Internet map services, and mobile phone signaling, used the random forest model, took Zhengzhou City as an example, revealed the spatial heterogeneity characteristics of residential electricity carbon emissions, and explored the nonlinear action law of residential electricity carbon emissions under the action of built environment, climate elements and socioeconomic factors. The results show that: (1) the average electric carbon emission of residents in the main urban area of Zhengzhou is 2.07t, and the spatial distribution is low in the central urban area and high in the peripheral urban area. The spatial distribution pattern of total carbon emissions is opposite, showing the characteristics of high in the central urban area and low in the pe-ripheral urban area. (2) The contribution of the built environment, climate factors and social economy to the change of carbon emissions of residential electricity decreased in turn, and the contribution degree was 50.07%, 26.17% and 23.75%, respectively. Among them, when the density of retail and commercial facilities and basic education facilities are in the range of 0-2000 /km2 and 0-100 /km2, respectively, the improvement of built environment diversity significantly increases residential electricity carbon emissions. And when the green land rate is 0-0.25, the impact of urban green space on residential electricity carbon emissions is first promoted and then suppressed. (3) There is significant spatial heterogeneity in the influencing factors of residential electricity carbon emissions. Land surface temperature is the most important contributor to the change of residential electricity carbon emissions in and within the second ring road, and the two have a "U" -shaped relationship. Building height was the most important contributor to the variation of residential electricity carbon emissions at and beyond the third ring road, and the two showed a positive nonlinear relationship. The research results improve the spatial cognition of urban residential carbon emissions at fine scales, reveal the differentiation characteristics of residential carbon emissions under the synergy of built environment, climate factors and socio-economic factors, and provide a scientific basis for the construction of low-carbon cities and climate resilient cities.