Existing methods for evaluating and predicting the health status of ecosystems suffer from low efficiency and low prediction accuracy. Therefore, an urban ecosystem health evaluation system and simulation prediction method based on geographic information systems and improved Markov models is proposed. Firstly, the basic geographic data of the research area is obtained through the spatial analysis function of GIS and the interpretation technology of urban remote sensing images. Based on the pressure-state-response model framework, 15 quantitative indicators were selected from three dimensions: environmental pressure, ecological status quo, and human coping to construct a comprehensive evaluation index system. In the stage of health status assessment, the particle swarm optimization algorithm is adopted to intelligently optimize the penalty factor and kernel function parameters of the support vector machine. By adaptively adjusting the position and velocity of the particles to find the optimal parameter combination, an ecosystem health classification model with the best classification performance is finally established. The prediction module integrates the dual advantages of Markov chains and cellular automata, uses the historical state transition probability matrix to represent the evolution law of the time dimension, combines the neighborhood transformation rule of CA to describe the spatial diffusion effect, and generates future multi-scenario prediction results through explicit iterative calculations in space. The prediction accuracy of the proposed method reached 95.2%, with a root mean square error of only 0.08. After applying the proposed method to practical ecosystem health management, the probability of the health status of the corresponding regional ecosystem transitioning to a healthier state reached 0.45% and 0.30%, respectively. In addition, the health status recognition accuracy of the proposed method also reached over 98%. The urban ecological health assessment and simulation prediction method proposed in the research can effectively assist in the effective implementation of ecosystem maintenance work and provide a reliable basis for urban ecosystem management.