Reducing wind speed is the main role of shelterbelt. To reveal wind speed change characteristics in the protected area, establish a simulation model of the average wind speed, height and average wind speed after the forest belt. Established three wind speed variation models respectively by the least squares method, ridge regression and elastic network method. Verified the wind speeds of the windward and leeward sides of eight different structural forests under wind speeds of 8,11 and 14 m·s -1

more »

... y wind tunnel experiments. Model evaluation indicators are the coefficient of determination (R 2 ) and root mean square error(RMSE).The results show that the R 2 of models all are higher than 0.9, and RMSE are 1.05492, 1.05490 and 1.05092 in turn. Comparing the modeling performance of three methods, the generalization ability of elastic network is better. The average of 0-5H wind speed before the forest belt and height can estimate the average wind speed after 0-10H the forest belt.