Currently, earth observation satellites have the characteristics of speed, dynamism, and accuracy, making them essential monitoring tools for high-accuracy monitoring of natural disasters. However, there are numerous earth observation satellites with significant differences in their main tasks and resolutions. Rational allocation of satellite tasks is crucial for emergency monitoring of natural disasters. Based on the analysis of the requirements of natural disasters, this study proposes a satellite observation planning method for natural disasters based on temporal and spatial constraints and task priorities. By establishing a satellite planning model and optimizing it through a combination of greedy algorithms and genetic evolution algorithms, this study conducts simulation experiments to verify the reliability of the model. The results indicate that greedy algorithms and genetic evolution algorithms are suitable for the planning of Earth observation tasks for natural disasters in medium-sized and large areas respectively.