Abstract: 【Objective】We established a landscape evaluation model to determine which landscape factors had the greatest influence on the beauty degree. Based on the evaluation model, we determined the optimal color composition pattern. 【Method】Based on the landscape of the autumn forest in southwest Zhejiang Province, the scenic beauty estimation method(SBE)was used to evaluate the beauty of mountain forest facies, and we decomposed the elements, calculated the color contrast, standardized the data, and established a landscape evaluation model. On the basis of the above model, the optimal color composition pattern could be obtained mathematically according to the relationship among the color components. 【Result】The calculation of the independent variables showed that the partial correlation coefficients of the five landscape factors, hue contrast ratio, green proportion, yellow proportion, red proportion, and canopy density, relatively affected the beauty degree. We decided to use the five landscape factors of 25 categories to establish a landscape evaluation model. The landscape evaluation model was:Y_SBE=-0.548-0.062x_7b+0.302x_7c+1.176x_7d-0.149x_8b+0.640x_8d+0.885x_8e+0.060x_9b+0.274x_9c+0.311x_9d+0.837x_9e+0.170x_11b+0.237x_11c+0.182x_11d+1.554x_11e-0.637x_17b+0.066x_17c-0.578x_17d-0.492x_17e, and the results showed that the evaluation model had a high degree of reliability. The red proportion had the highest percentage of contribution to the largest score, reaching 30.0%. Thus, the red proportion had the largest contribution rate to the model, and hue contrast ratio, green proportion, yellow proportion, and canopy density had a contribution rate that decreased in turn. When the hue contrast ratio was between 72 and 144, the hue contrast ratio was proportional to the SBE. The yellow proportion and red proportion were basically proportional to the SBE. Canopy density was proportional to the SBE value only when canopy density was between 70%-80%. The integrated green proportion, yellow proportion, red proportion, and hue contrast ratio, from the perspective of the color of the visual experience, built the relationship; while we used the enumeration method to determine that green accounted for 50%-60%, yellow accounted for 0-10%, and red accounted for 30%-40% for the mountain landscape in the optimal color composition model.【Conclusion】The color has a significant influence on the beauty degree, and there were five landscape factors, including the hue contrast ratio, green proportion, yellow proportion, red proportion, and canopy density those had a high correlation with the mountain scenery in the autumn. We obtained a linear evaluation model with high degree of credibility. The green proportion was basically proportional to the SBE value when green accounted for more than 60%.The proportion of yellow and red were basically proportional to the SBE value. The best hue contrast ratio was between 36 to 72. The best canopy density was between 70% to 80%. When integrating the green proportion, yellow proportion, red proportion, and hue contrast ratio, the optimal color composition model occurred when the green components accounted for 50%-60%, yellow component accounted for 0-10%, and red component accounted for 30%-40% in the southwest of the Zhejiang mountain forest from the point of view of the color of the visual experience. Finally, according to the research results, the paper puts forward a management strategy and method of forest facies transformation.