Aeroacoustic imaging is a standard technique for mapping positions and powers of acoustic sources with microphone arrays. In this paper, we firstly propose an improved forward model of acoustic power propagation by considering background noise and model uncertainty caused by wind tunnel effect. To obtain super spatial resolution in strong noise interference, we then propose a robust super-resolution approach via sparsity constraint by jointly reconstructing source powers and positions, and adaptively estimating the background noise and sparsity parameter. We investigate the sparse distribution of source powers, and apply the incoherence of signal propagation matrix to determine the sparsity parameter. Our approach is compared with classical beamforming, deconvolution and sparse regularization methods by simulated, real data and hybrid data. It is feasible to apply proposed approach for monopoles and extended source imaging with the 2D non-uniform microphone array in wind tunnel tests.