This research creates a predictive model for geomagnetic storm forecasting using information from the National Oceanic and Atmospheric Administration's (NOAA) Deep Space Climate Observatory (DSCOVR) satellite. The satellite DSCOVR is currently used to monitor space weather as it measures the main factors of the solar wind and the local magnetic field. These measurements are critical for understanding space weather processes and their effects on Earth. The predictive model was built using data on solar wind speed, density, temperature, and magnetic field intensity. When it comes to sequential data, Long Short-Term Memory (LSTM) models typically perform better than different neural network patterns. This is because LSTM models have memory cells, which improves their ability to identify long-term dependencies in the data and makes them ideal for use in weather prediction models. The model tries to predict significant shifts in the magnetic field and solar wind to provide early warnings of geomagnetic storms. Forecasting these storms is critical for minimizing any interruptions because they threaten communication networks, power grids, and satellite activities. This paper advances space weather forecasting and shows a direct use of machine learning techniques in space science. The LSTM model's predictive capabilities allow us to better prepare for geomagnetic storm events, protecting technology such as satellites and advancing our knowledge on space weather.