Abstract: Abstract: The surface of the coal mining subsidence area will continue to sink at a relatively slow speed until it reaches limit, and the continuous settlement will inevitably affect the choice of land reclamation time, the growing crops which need nutrients from the land, and the normal use of agricultural facilities. It is very necessary to study the dynamic settlement of land over time to guarantee the time of land reclamation and the effectiveness of agricultural production. Time effect parameter in the Knothe function was analyzed in this study, and found that the parameter value had a great influence on the dynamic calculation of the surface subsidence and stability time. In the short period after mining, the bigger the parameter was, the greater the surface subsidence, and vice versa. And the bigger the parameter, the shorter the time needed for the surface from starting sinking to reaching the stability, and vice versa. In fact, the parameter was changing over time; the parameter in different periods were obtained based on the surface dynamic subsidence and time firstly, then the values were fitted, and the parameter in any period could be obtained according to the fitted formula. The dynamic settlement of coal mining subsidence area was closely related to the time; the Knothe function was introduced into the probability integral model, and the new idea of the ultimate subsidence and stage subsidence in the coal mining subsidence area was proposed. The prediction model of land dynamic settlement in coal mining subsidence area was constructed based on the Knothe function, which could be used to calculate the stage subsidence value at any time according to the engineering requirements. The subsidence factor, which changed with the time and was obtained from the time effect parameter, was the most important parameter in the prediction model. In order to verify the reliability and practicability of the prediction model, the paper took the measured surface subsidence data from the 2407 working face as an example, and carried out the application of the model as well as got the parameter. The time effect parameter at the 6th, 7th, 8th, 9th and 10th observation were calculated, respectively, and the empirical formula of the time effect factor with the time was obtained by fitting. By using the fitting formula, the time effect factor at the 11th observation was calculated. The calculated and the actual observation value of the time effect factor were compared, and the relative error was only 2.6% and the fitting effect was good. In addition, the measured and predicted values of the surface subsidence at the 9th observation were compared and analyzed. The results showed that the maximum relative error of single point was 8.8%, the average relative error of single point was 1.2%, the overall root mean square error was 187 mm, and the overall relative root mean square error was 5.6%. The difference between the predicted and measured value was small, which further proved the reliability of the model. According to the calculation model, the ultimate and the stage subsidence in multiple periods for the surface of coal mining subsidence area were calculated. The analysis showed that the prediction model conformed to the actual condition.