Untangling the role of aerosols in cloud development requires continuous and concurrent measurements of aerosols, clouds, and meteorological conditions. The intensive operational period (IOP) during the U.S. Department of Energy (DOE) Tracking Aerosol Convection Interactions Experiment (TRACER) field campaign provides such datasets. Based on K-means clustering analysis of aerosol data during TRACER-IOP, we found four dominant aerosol clusters present in La Porte, Texas: <em>Continental</em> (24 %), <em>Extreme Pollution</em> (8 %), <em>Sea Breeze and Rural </em>(45 %), and <em>Sea Breeze and Pollution</em> (23 %). The two <em>Sea Breeze</em> clusters are characterized by the inclusion of relatively large particles sized around 180 nm, along with the peak frequency in the afternoon and evening, which are both consistent with the characteristics of sea breeze. These two clusters are a mixture of different aerosol types, which confirms the complexity of aerosol characterization in this region. Utilizing the representative number concentration and size distribution of these aerosol clusters, large eddy simulations of shallow cumulus clouds were performed with the Cloud Model 1. Our comparison of simulated cloud characteristics using a bulk microphysics scheme with those using a Lagrangian scheme (the super-droplet method) reveals that the bulk scheme tends to produce too much precipitation too early, despite the fact that we do not observe surface precipitation from these shallow cumulus clouds of interest, also in contrast to the results with the Lagrangian scheme. This discrepancy is due to the underestimation of droplet number concentrations and/or the overestimation of collision-coalescence rates in the bulk microphysics scheme.