For the Atlantic storms in ERICA IOP 1–5, NMC operational surface analyses, both manual and automated, were compared with two sets of research analyses prepared later. The positions of cyclone centers agreed within 100 km on average only between the two research sets. Root-mean-square deviations of the automated analysis positions from the research positions were 180 km. Central pressures were not deep enough, especially in the automated analyses. Comparison of reported pressure with the research analyses shows that those from the moored buoys and C-MAN stations were most accurate and reliable. The drifting buoys were nearly as good, as were the best ships. Analyses are shown in detail for the IOP 2 storm, during its evolution from a complex multi-centered system to a single center of great intensity. Careful consideration of low-level aircraft data and of observations from ships (with detection and correction of their errors), was necessary for reconciliation of analyses differences. There were not enough observations to resolve all problems. The final great intensity of the center would not have been known without a low-level aircraft traverse. A small sample of delayed ships' observations received after completion of the research suggested that the root-mean-square error of the pressure analysis was about 1.5 mb.