The invention discloses a full-time-scale spatial correlation analysis method of different spatial points of a near surface wind field. The analysis method comprises the following steps: synchronously collecting the wind field time sequence information of different spatial points, carrying out pairwise coupling on the wind field time sequence information of different spatial points, and independently carrying out deviation accumulation on each pair of non-stable wind field single-point time sequences; independently carrying out halving from a first data point and a tail data point to obtain intervals which are not overlapped so as to obtain a plurality of isometric intervals; fitting data by a least square method to obtain a local non-stable tendency in each interval; calculating the removal tendency mean square errors of the plurality of isometric intervals and the removal tendency mean covariance of the plurality of isometric intervals; obtaining the mean value of the removal tendency mean square errors, and carrying out extraction of a root to obtain a mean square root deviation under a time scale; obtaining the mean value of the removal tendency covariance in the plurality of isometric intervals to obtain an average covariance under the time scale; and calculating full-time-scale cross-correlation measure between every pair of non-stable wind field single-point time sequences measured at two different spatial points of the near surface wind field. The analysis method can integrally display a change situation of cross-correlation intensity among different time sequences along with the time scale.