A new method called DIDW was developed to improve geospatial interpolation. It combines traditional IDW with data-to-data correlation and adjusts the exponent locally. DIDW was tested against IDW and ordinary kriging using the Walker Lake dataset. Results show that DIDW produces more accurate estimates than IDW and kriging, especially when dealing with clustered data and varying variogram parameters. This makes DIDW a reliable option for spatial interpolation in various applications.