<p>Wetlands play a crucial role in mitigating climate change via the process of carbon sequestration (CS). The spatiotemporal evaluation of methane (CH₄) emission, and CS throughout the pre-monsoon, monsoon, and post-monsoon seasons was carried out in the Purbasthali wetland, an oxbow lake, situated in the Purba Burdwan region of West Bengal, India. Standard process-based mathematical equations were applied to compute the net CS. Various soil and water variables were also observed spatiotemporally and a machine learning-based artificial neural network (ANN) model was employed to find out the major factor influencing CS. Moreover, the net CS along with water quality index (WQI) was used to assess the ecological restoration in this lake. The study showed that emission was higher in the monsoon months (1083.59 ± 56.54&#xa0;gCO₂e&#xa0;m⁻²&#xa0;yr⁻¹), whereas CS was higher (2411.88 ± 72.26&#xa0;gCO₂e&#xa0;m⁻²&#xa0;yr⁻¹) in the post-monsoon season. The average total net CS in the Purbasthali throughout the study was 13.96 ± 8.35&#xa0;tCO₂e&#xa0;ha⁻¹&#xa0;yr⁻¹. The ANN model revealed that dissolved oxygen (DO) and temperature are most important features that influence net CS. These findings can be utilized to develop an effective monitoring system aimed at optimizing CS rates, reducing CH₄ emissions, and identifying ecological imbalances. The observed variations in net CS and WQI for the Purbasthali wetland are linked to shifts in wetland usage patterns and the seasonal enhancement of ecosystem services. Notably, the positive net CS indicates improved environmental conditions within the lake and its surroundings, reflecting the success of restoration efforts undertaken by the local community and government. On a global scale, restored and constructed wetlands have proven to be highly effective in carbon storage, pollutant removal, and habitat improvement, contributing significantly to climate goals and urban sustainability objectives. The work was inclined to meet the Sustainable Development Goal 14.</p>