The long-term storage of CO2 must be verified to ensure the success of geologic carbon sequestration projects. To detect subtle CO2 leakage signals, we present a strategy that integrates near-surface measurements of CO2 fluxes or concentrations with an algorithm that enhances temporally- and spatially-correlated leakage signals while suppressing random background noise. We assess the performance of this strategy using synthetic CO2 flux data sets and modeled surface CO2 leakage. These simulations provide a means of estimating the number of measurements required to detect a potential CO2 leakage signal of given magnitude and area. Results show that given a rigorous and well-planned field sampling program, subtle surface CO2 leakage may be detected using the algorithm; however, leakage of very limited spatial extent or exceedingly small magnitude may be difficult to detect with a reasonable set of monitoring resources.