The growth in energy demand is a becoming critical issue worldwide, which leads to the exhaustion of conventional energy (e.g., fossil fuels) and causes adverse environmental pollution. Thus, this promotes wider adoption of renewable energies (e.g., solar, bio, hydro, wind, geothermal and ocean energy). According to the report of the World Energy Outlook 2021, solar photovoltaic (PV) represents one of the most economic sources of new electricity generation. Hence, solar energy is increasingly becoming an appealing source of electricity worldwide. However, for a densely built city-state country, it faces challenges to find large spaces for utility-scale deployment of solar panels. Integrating solar PV systems into the urban environment via building rooftops is an effective solution for this problem. This project develops a comprehensive model for accurate estimation of solar irradiation on the building rooftop for a dense and high-rise city. Firstly, this paper uses the machine learning model to establish a robust relation between cloud optical thickness, s aerosol optical thickness, clear-sky irradiation, and land surface solar irradiation. Then, this project performs parametric study on the impact of urban morphology on the solar energy potential on the building rooftop in Hong Kong context. Meanwhile, deep learning method is used for extract the available building rooftop area from the satellite images. Finally, available rooftop area is combined with the estimated urban solar irradiation to generate the solar irradiation map on the building rooftop in Hong Kong.