Photoacoustic (PA) image reconstruction of data acquired with conventional linear arrays suffers from incompleteness due to limited bandwidth and limited view. This problem is compounded by the dominance of boundary signals suppressing other weaker PA sources. In this study, we propose a method that uses a naturally coregistered ultrasound image to enhance the PA image reconstruction, with the dual aim to reconstruct the sources that suffer from incompleteness and reveal weak sources near a strong boundary. In this method, an ultrasound image provides the input for a PA wave field simulation. The simulated PA field can be combined with the measured fields to complete the partial reconstruction or to suppress the dominant boundary signals. Experimental validation of the method was performed with two different phantoms and in vivo data from the lower arm of a healthy volunteer.