An understanding of radiation e†ects on the evolution of shock waves is of great importance to many problems in astrophysics. Shock waves driven by a laser-heated plasma are attractive for laboratory investigation of these phenomena. Recent studies of intense short-pulse laser interactions with gases of atomic clusters indicate a potential avenue to access this regime of radiative hydrodynamics. We have measured the energy absorption efficiency of high-intensity, picosecond laser pulses in low-density gases composed of large atomic clusters and Ðnd that the energy absorption can be very high ( [ 95%), producing a high-temperature plasma Ðlament which consequently produces a strong blast wave. Interferometric characterization of these shock waves indicates that in high-Z gases such as Xe, radiation transport plays an important role in the evolution of the shock wave. Subject headings : atomic processes È hydrodynamics È plasmas È radiative transfer È shock waves Understanding the dynamics and evolution of shock waves is crucial to understanding the structure of the interstellar medium (McKee & Draine 1991 ; Blandford & McKee 1976 ; Draine & McKee 1993 ; Shull & McKee 1979 ; Strickland & Blondin 1995). For example, shock wave evolution and the growth of hydrodynamic instabilities such as the Rayleigh-Taylor and the RichtmyerMeshkov instabilities play a fundamental role in the evolution of expanding supernovae (Arnett 1996). In such interstellar shocks, the e†ects of radiation transport are quite important and can dramatically a†ect the shock dynamics (McKee & Draine 1991 ; Draine & McKee 1993 ; Shull & McKee 1979). Radiation transport a†ects the shock dynamics in the interstellar medium in a number of ways, such as by radiative cooling and via the formation of a radiative precursor ahead of the shock front (Shull & McKee 1979). Because of the great importance of these dynamics on astrophysical structure, there is a strong motivation to study these e†ects in laboratory-produced plasmas (Rose 1991). There has been some recent success in accessing shock physics of relevance to astrophysics using shocks driven by large-scale, long-pulse lasers (Remington et al. 1997). For example, the presence of a radiative wave preceding a blast wave in xenon was observed by Bozier et al. (Bozier et al. 1986), and Grun et al. have seen evidence of the growth of instabilities in blast waves propagating through xenon (Grun et al. 1991). These experiments have inferred the presence of radiation transport ahead of the shock front through streak camera diagnosis of the shock wave front. However, no measurements have been reported to date on the complete shape and evolution of a radiative shock wave. To access the regimes of interest for radiative hydrodynamics studies, it is desirable to produce plasmas with initial temperatures above a few hundred eV in a lowdensity material. A unique method by which hightemperature plasmas may be produced in gas is possible utilizing ultrafast lasers and gases containing atomic clus1 Imperial College of Science, Technology, and Medicine, Blackett Laboratory, Prince Consort Road, London SW7 2BZ, UK. ters (Ditmire 1997). Recent studies have shown that a gas of large atomic clusters ([ 1000 atoms cluster~1) presents a radically di†erent environment for laser-plasma interaction dynamics (Ditmire 1997 ; Ditmire et al. 1995 ; Ditmire et al. 1997a) ; Ditmire et al. 1997c ; Ditmire et al. 1996 ; Ditmire 1997b ; Ditmire et al. 1998 ; Lezius et al. 1997 ; McPherson et al. 1994 ; Purnell et al. 1994 ; Shao et al. 1996). In general, low-density gases are expected to exhibit very low absorption efficiency (\ 1%), and the plasmas produced by intense irradiation will generally be quite cold (10È100 eV). The presence of clusters in a gas changes this situation dramatically (Ditmire et al. 1997b). Though the average density of a gas containing clusters is low, the local density within the cluster is near solid and, consequently, will be subject to the rapid heating experienced by a solid target due to collisional inverse bremsstrahlung. Very bright X-rays have been observed from gas target plasmas produced by intense femtosecond illumination of clusters (Ditmire et al. 1995 ; McPherson et al. 1994), indicating that electron temperatures in these plasmas were quite high, far in excess of those expected from a gas composed only of single atoms. In fact, absorption measurements indicate that cluster gases are at least as efficient as solid targets in absorbing short pulse laser energy and are, therefore, ideal media for the production of strong blast waves. Large atomic clusters can be produced in the expansion of a gas jet into vacuum under certain conditions. The cooling associated with the adiabatic expansion of the gas will cause certain atomic species to coalesce into van der Waals bonded clusters. An ultrafast laser pulse can then interact with clusters on a timescale faster than the cluster disassembly. Recent studies of intense, subpicosecond laser interacations with individual clusters with greater than a few hundred atoms per cluster have conÐrmed that hot electrons (up to 3 keV) are produced during the laser-cluster interaction (Shao et al. 1996). Furthermore, under certain conditions, an even greater energy can be deposited in the ions when these hot, highly ionized clusters explode (Ditmire et al. 1997c). These studies indicated that the clusters are rapidly heated by the laser, to a nonequilibrium, superheated state, in large part because of the passage of the free electrons in the cluster through a Mie resonance with