To address the rapid attenuation of signal amplitude and the difficulty in recovering effective responses under low signal-to-noise ratio (SNR) conditions in long-range detection of moving magnetic targets using high-temperature SQUID planar gradiometers, this study investigates the influence of the relative distance between the target and the sensor on the extractability of magnetic gradient signals. For moving magnetic targets, the Closest Point of Approach (CPA) serves as a key geometric parameter characterizing the minimum distance between the target and the sensor.First, the spatial distribution characteristics of magnetic gradient fields generated by large-scale magnetic targets are analyzed using a COMSOL based finite element model. The results show that as the CPA increases, the lateral coverage of the magnetic gradient disturbance on the observation plane expands, while its peak amplitude decreases significantly. Subsequently, under the condition that the observation distance satisfies the magnetic dipole approximation, a moving magnetic dipole model is established to theoretically analyze the time-domain responses at different CPA. The analysis indicates that, at large CPA, the magnetic gradient signals persist in the form of weak amplitudes with broadened temporal structures. To extract such weak structured signals from noisy measurements, an Orthogonal Basis Functions (OBF) based signal extraction method under a representation normalized with respect to the CPA is introduced. By constructing a structural subspace that matches the response characteristics of the moving magnetic dipole, the proposed method enables structured recovery of magnetic gradient signals under low SNR conditions. Experimental results demonstrate that the proposed method can effectively recover the target response waveform for a magnetic target with a moment of 2.5 A·m² at a maximum distance of approximately 2 m, where the signal approaches the sensitivity limit of the gradiometer, even when the initial SNR is as low as −7 dB. significantly improving signal detectability under weak-signal conditions.The results provide methodological support for the application of high-temperature SQUID planar gradiometers in long-range magnetic anomaly detection.
