Elevated CO2 concentrations have been observed to accelerate flowering time in Arabidopsis through the action of a highly conserved regulatory network controlled by miR156 and miR172. However, the network's robustness to the impact of increasing CO2 concentrations on flowering time remains poorly understood. In this study, we investigate this question by conducting a comprehensive analysis of the global landscape of network dynamics, including quantifying the probabilities associated with juvenile and flowering states and assessing the speed of the transition between them. Our findings reveal that a CO2 concentration range of 400-800ppm only mildly advances flowering time, contrasting with the dramatic changes from 200 to 300ppm. Notably, the feedback regulation of miR156 by squamosal promoter binding protein-like proteins (SPLs) plays a substantial role in mitigating the effects of increasing CO2 on flowering time. Intriguingly, we consistently observe a correlation between delayed flowering time and increased variance in flowering time, and vice versa, suggesting that this might be an intrinsic adaptation mechanism embedded within the network. To gain a deeper understanding of this network's dynamics, we identified the sensitive features within the feedback loops of miR156 SPLs and miR172-APETALA2 family proteins (AP2s), with the latter proving to be the most sensitive. Strikingly, our study underscores the indispensability of all feedback regulations in maintaining both juvenile and adult states as well as the transition time between them. Together, our research provides the first physical basis in plant species, aiding in the elucidation of novel regulatory mechanisms and the robustness of the miRNAs-regulated network in response to increasing CO2, therefore influencing the control of flowering time. Moreover, this study provides a promising strategy for engineering plant flowering time to enhance their adaptation and resilience.