The design of solar energy systems poses significant challenges, encompassing issues related to infrastructure, installation costs, and land availability and selection. In the practical implementation of such projects, location selection is critically dependent on numerous vital criteria, one of which is the estimation of received solar radiation. For instance, solar radiation estimation is pivotal for designing proper energy storage systems, balancing load, and managing photovoltaic and thermal systems. Machine and deep learning approaches have been commonly utilized in solar radiation forecasting and estimation. This paper deploys multiple deep learning techniques, including feed-forward neural network, Long Short-Term Memory (LSTM), and Convolutional Neural Network (CNN) to estimate the amount of Global Horizontal Irradiance (GHI) received in the city of Tustin, California, using multiple metrological factors. California was selected for this study because it is characterized by receiving high amounts of solar radiation and large amounts of solar energy installations and projects. The paper compares the three examined deep learning methods using the commonly used evaluation metric, the Root Mean Squared Error (RMSE). The findings indicate that the LSTM approach is the most effective deep learning technique for estimating the hourly GHI.