<p>Potential field downward continuation is a key technology in high-precision airborne gravity and magnetic surveys, which is of great significance for geological structure analysis. However, traditional frequency-domain methods suffer from problems such as low computational efficiency, insufficient accuracy, and poor stability in processing data gap areas. To solve these issues, this paper proposes a potential field downward continuation method that combines the Fastest Fourier Transform in the West (FFTW) algorithm with the minimum curvature method. By constructing an FFTW algorithm environment and designing a frequency-domain iterative solution model under minimum curvature constraints, this method realizes high-precision and large-span downward continuation processing for aeromagnetic survey data with arbitrary grid spacing and data gaps, without the need for pre-edge extension or grid gap filling. Through theoretical model calculations and verification with actual aeromagnetic data, the results show that this method significantly improves processing efficiency while ensuring computational stability, achieves high processing accuracy, and effectively enhances local anomaly information. In addition, the algorithm is integrated into the Geoprobe airborne gravity and magnetic data processing platform, providing strong technical support for refined interpretation of airborne geophysical data and exhibiting broad application prospects in related fields.</p>