Evaluation of the ecological/environmental damage is critical for the effective implementation of the ecological/environmental damage compensation system. The baseline is a critical gauge and standard for damage validation and successful restoration, which runs through the entire process of the ecological/environmental damage assessment. The method of historical data, which uses the average value of 3 years before the damage occurred, is preferentially used to ascertain the static baseline. However, reflecting the dynamics and uncertainty of the baseline is challenging. Thus, it is imperative to explore the dynamic baseline of the ecological/environmental damage. In this study, we selected the net primary productivity (NPP) as an indicator of the grassland ecosystem damage. Based on the historical data of NPP (2000-2008) before the development in the Bayanbaolige mining area, we estimated the dynamic baselines of NPP between 2009 and 2011 using the grey topological model and assessed the availability of the method by comparing with the static baseline values. The grey topological prediction model exhibited high modeling accuracy and small relative error. Compared with the actual values from 2000 to 2008, the average relative error was 2.88%. Meanwhile, the fitting curve of 2009-2011 was consistent with the actual curve. The findings of the grey topological model as dynamic baselines were better than that of the static baseline in presenting the real situation. In conclusion, the prediction method of the grey topological model offers a possible solution to the baseline determination, which has application value and scientific rationality.