The invention provides an area division method based on improved orthogonal decomposition. The method comprises the following steps of firstly, acquiring the coordinate time sequence observation valueof a GNSS station, and obtaining a residual coordinate time sequence according to the coordinate time sequence observation value; then, calculating the public epoch of all the stations and constructing the observation sample matrix of all the stations at a preset direction based on the public epoch, constructing the correlation coefficient matrix Bn*n of all the stations, and then carrying out orthogonal decomposition on the correlation coefficient matrix Bn*n to obtain an orthogonal matrix P; and then, acquiring a load matrix A according to the observation sample matrix and the orthogonal matrix, and acquiring a characteristic vector from the orthogonal matrix P, and selecting K cumulative characteristic vectors according to the characteristic values of the characteristic vector; processing the K cumulative characteristic vectors again, and carrying out linear transformation on the orthogonal matrix P to obtain a load matrix B' after linear variation and a principal component G; andfinally, analyzing the load matrix B' after the linear variation to obtain an area division result. In the invention, the accuracy of area division can be greatly increased.