This research focuses on the detailed modeling of solar PV (Photovoltaic) in urban areas by integrating a detailed reconstructed digital surface model (DSM) and high-temporal-resolution satellite data. This study introduces enhancing method for open digital surface model from medium to high resolution to capture building height details, and integrating the new DSM with high-temporal satellite data to get detail spatial and temporal resolution of solar PV. The DSM reconstruction, building heights are determined using a combination of multi-data and multi-machine learning regression approaches to achieve building height results closely aligned with actual height. In Jakarta, building heights ranged from 5 to 223 m, while in Bandung, they ranged from around 3.58 to 254 m. The combination scenarios had a mean absolute error of 1.584 m, and the coefficient of determination (R2) was 0.754. The integration of new DSM detail data and high-temporal data employs an approach to integrate surface solar irradiance by shadow, sky view factor, and reflectance irradiance from surrounding object. The annual average energy potential in Jakarta and Bandung ranges from 260 to 420 W/Wp. Areas near tall buildings exhibit lower potential because of the shadow effect and sky view factor. This research hopefully can contribute to urban planning for carbon neutrality by providing a more detailed understanding of solar PV potential in urban areas in the absence of detailed data limitations.