Urban areas are responsible for 80% of GHG emissions, leading scientists and policymakers to act towards city greening. The contribution of these urban green spaces to storing carbon and climate change mitigation is unknown, particularly in the tropics. Here, we present a methodology that combines high-resolution satellite images (3 x 3 m) with biomass sampling in Brasilia, Brazil's capital. The phytosociological parameters of the woody vegetation and the total aerial biomass of the grass layer were calculated. Our model used a CO2 flux index calibrated with field data. The CO2 flux index was produced by the Planet satellite constellation. The woody plant community had 59 individuals (21 species), and the average biomass of the grass layer varied between 8,915.68 and 11,545.12 kg.ha− 1. The CO2 flux reveals areas with higher carbon flow, indicating greater carbon sequestration due to more expressive photosynthetic activity. Linear regression (R2 = 1) was used to estimate total carbon storage in eight urban parks, and cerrado areas exhibit higher values, while degraded or constructed areas show the lowest values. Our simple and low-cost method using field data with remote sense imagery produces precise biomass estimations at regional levels, which greatly apply to tropical landscapes. By enhancing knowledge about the urban positive and negative influence on the carbon cycle, our method could support scientists and policymakers in targeting climate and biodiversity-friendly agendas.