1: \begin{abstract}

2: We report the first maximum-light far-Ultraviolet to near-infrared spectra (1000\AA\ $-$ 1.62$\mu$m, rest) of a hydrogen-poor superluminous supernova, Gaia16apd. At $z=0.1018$, it is the second closest and the UV brightest SLSN-I, with 17.4 magnitude in {\it Swift} UVW2 band at $-11$\,days pre-maximum. The coordinated observations with {\it HST}, Palomar and Keck were taken at $-2$ to $+25$\,days.  Assuming an exponential (or $t^2$) form, we derived the rise time of 33\,days and the peak bolometric luminosity of  $3\times10^{44}$\,erg\,s$^{-1}$.  At maximum, the photospheric temperature and velocity are 17,000\,K and 14,000\,km\,s$^{-1}$ respectively. The inferred radiative and kinetic energy are roughly $1\times10^{51}$ and $2\times10^{52}$\,erg. 

3: Gaia16apd is extremely UV luminous, emitting 50\%\ of its total luminosity at $1000 - 2500$\AA.  Compared to the UV spectra (normalized at 3100\AA) of well studied SN1992A (Ia), SN2011fe(Ia), SN1999em (IIP) and SN1993J (IIb), it has orders of magnitude more far-UV emission.  This excess is interpreted primarily as a result of weaker metal line blanketing due to much lower abundance of iron-group elements in the outer ejecta. Because these elements originate either from the natal metallicity of the star, or have been newly produced, our observation provides direct evidence that little of these freshly synthesized material, including $^{56}$Ni, was mixed into the outer ejecta, and the progenitor metallicity is likely sub-solar. This disfavors Pair-Instability Supernova (PISN) models with Helium core masses $\geq90M_\odot$, where substantial $^{56}$Ni material is produced. Higher photospheric temperature definitely contributes to the far-UV excess from Gaia16apd.  Comparing with Gaia16apd, we find PS1-11bam is also UV luminous.  

4: 

5: \end{abstract}

6: