Linear prediction methods, based on a Hankel data matrix, suffer fromsubspace leakage and degraded resolution when applied to data models that do notresult in a mode matrix with Vandermonde structure, such as the constant-Q model.In the absence of noise, the Vandermonde structure ensures the equivalence betweenthe number of backscattered signals and the rank of the data matrix. This paper firstidentifies the origin of subspace leakage residing in subspace-based and linear predictionmethods when applied to data of the constant-Q model. Second, it proposesa frequency-distortion technique, based on the extension theorems, for suppressingthe leakage and preserving the time resolution performance of these methods. Theeffectiveness of the distortion technique is then demonstrated on GPR simulated databy extending the damped MUSIC algorithm to the joint parameter estimation of theconstant-Q model.