A new method is proposed for the characterization of the atmospheric turbulence strength parameter (
C
n
2
) based on the principal component analysis (PCA) of the texture of the speckle intensity pattern obtained on long-range propagation of a focused charge +1 vortex beam. Employing the split-step propagation method, datasets containing instantaneous intensity images of the focused vortex beam are generated for three distinct
C
n
2
values, specifically
1
×
10
−
14
m
−
2
/
3
,
5
×
10
−
14
m
−
2
/
3
, and
1
×
10
−
13
m
−
2
/
3
, representing medium to high turbulence levels for a 2 km propagation distance. The gray level co-occurrence matrix (GLCM) methodology is employed to extract key texture attributes, like contrast, correlation, homogeneity, and energy from the intensity images. These extracted texture parameters serve as inputs for training a PCA model, enabling the identification of associated
C
n
2
values. The PCA analysis exhibits distinct clustering of the first three principal components for each of the three
C
n
2
values, forming individual clusters on the PCA plot. Standard deviational ellipses are drawn to clearly demarcate these clusters on the PCA plot. The texture-based PCA classification of atmospheric turbulence was also performed for a focused Gaussian beam. The comparison of PCA plots between vortex and Gaussian beams showed that a pronounced clear separation of
C
n
2
values is obtained for the vortex beam. This indicates that the non-zero orbital angular momentum of the vortex beam also plays an important role in achieving the distinct separation of
C
n
2
values on the PCA plot. The proposed method can provide efficient real-time turbulence estimation solely on the basis of texture of the instantaneous intensity speckle with prior training and therefore may simplify the estimation of turbulence strength parameter.