Green hydrogen (GH2) serves as a vital substitute for fossil fuels and is essential in reaching the goal of net-zero carbon emissions. Despite GH2 potential, the widespread adoption of GH2 production technology faces several significant Challenges. Machine learning techniques can contribute to overcoming these challenges by accurately forecasting solar irradiance, estimating green hydrogen yield, and enhancing the efficiency of renewable energy systems. This research examined weather data to investigate the intricate connections between environmental parameters and solar radiation. Several machine learning algorithms were trained and evaluated using R \(^\textbf{2}\) , Mean Squared Error (MSE), and Mean Absolute Error (MAE) as performance metrics. The Random Forest (RF) model proved to be the most accurate, achieving an R \(^\textbf{2}\) score of \( \mathbf{0.744} \) , an MSE of \( \mathbf{25,471.64} \) , and an MAE of \( \mathbf{90.14} \) . Using the predicted solar radiation values, green hydrogen production was estimated with a \({\boldsymbol{2}0\%} \) energy conversion efficiency. The study estimated an annual hydrogen production potential of approximately \( \mathbf{452.36} \) kg, with variations depending on model accuracy and environmental conditions.