【Objective】This article primarily analyzes the effect of climate change on the apples initial flowering date. The most important meteorological factors will be identified. Two methods will be used to forecast the apples initial flowering date in this article. Some scientific results will be provided to better allow for identifying meteorological disasters before they occur.【Methods】Used partial least square regression method to analyze the correlations between the apples initial flowering date and meteorological factors. Use the improved accumulated temperature method, linear correlation method and partial least square regression model to forecast the apples initial flowering date. Meteorological factors about light, heat, water were obtained from the Xifeng Agricultural Meteorological Stations observations. Data about apple florescence were obtained from the Xifeng Agricultural Meteorological Station which is located at Dongzhi Loess Plateau, Qingyang city, Gansu province. The observed variety of apples for testing was the Red Fujiapple. Meteorological data was obtained from the Xifeng National Based Station in Gansu province. The years used for testing were all 22 years from 1994 to 2015.【Results】(1) There was a significant negative correlation between the date of apple florescence and the averange temperature from February to early April. There was a low correlation between the date of apple florescence and the average temperature from November last year to January this year. When the average temperature was higher, the apple blossomed earlier. (2) There was a significant correlation between the accumulated temperature (accumulated temperature≥0 ℃ and ≥5 ℃) and the apples initial flowering date during November 1st last year to April 10th this year. The larger the accumulated temperature, the earlier the apples initial flowering date. It indicated, with the heat condition increasing year by year, that the apples initial flowering date also showed an advancing trend. The warmer winter temperatures make apple trees safer to go through the winter. (3) When the average temperature from March to early April was higher than the minimum temperature for the growth of apple trees, the apples initial flowering date was more than 20 days behind the first date of 5 ℃, and was closer to the first date of 10 ℃. (4) The relationship between ten days or more of monthly precipitation and the apples initial flowering date had very little correlation. Therefore, precipitation was not the factor that restricted the blossoming date of the apples. (5) There was a strong negative correlation between the apples initial flowering date and the sunshine duration in early January and March. Especially in March, the sunshine duration had passed the alpha0.05 significance test, which means in March, when the sunshine duration has increased, the apples initial flowering date will also advance. (6) In using the improved accumulated temperature model to forecast the apples initial flowering date from 1994 to 2015, there was a 81.8% year forecast date which was closer to the actual date. The maximum time difference was 16 days. The years with errors of more than 7 days reached 5 years. Then forecasted date of the apples initial flowering date in 2013 was April 23th, while the actual date was April 7th.