In this project, we explore the impact of waterquality data and buoy data collected from the Chesapeake bay on chlorophyll concentration over the recent decades. Chesapeake bay is an important and sensitive region on the east coast of the USA, in the atlantic ocean. The data itself is sparse, but records a large number of variables. We explored the data with Boosting methods: XGBoost and LightGBM with different sets of hyperparameters, followed by SHAP analysis to find the parameters that had highest correlations with chlorophyll content, such as Salinity, as well as those that were not, such as seconds and minutes. We also explored simple MLP regressor and Recurrent Neural Networks to observe how well they perform. Recurrent neural networks had a significantly lower loss when compared to dense networks.