Client studies cloud formation from aerosols and needs a model to predict particle number concentrations. The model uses reanalysis data on temperature, carbon monoxide and date to predict particle concentrations over 100nm in size (N100) which are a proxy for cloud formation. A linear regression model was created relating N100 to temperature, carbon monoxide and season. It explained 29.2% of variation but underestimated high values. Improving the model by addressing non-linear relationships and outliers could increase accuracy. A web app was being developed to demonstrate the concept.