The increasing availability of high spatial resolution data of velocit
y and intensity variations on the Sun has stimulated the development o
f helioseismological techniques that probe the solar interior in local
ized regions. The techniques developed so far(1-4) have yielded inform
ation on physical quantities (such as the flow velocity and magnetic f
ield) below the surface, but are still far from providing a detailed p
icture of local subsurface inhomogeneities. Here we report the develop
ment and application of a new method for constructing three-dimensiona
l solar images, utilizing acoustic noise (or stochastic P-mode oscilla
tions) in the Sun. We treat a region of the solar surface as a phased
array of acoustic sensors, which acts as a computational 'lens'; acous
tic waves 'scattered' by local inhomogeneities, such as sunspots, are
collected and summed in phase, based on the knowledge of how (on avera
ge) they travel within the Sun. In this way, we are able to construct
a three-dimensional image of a region of the solar interior.