The article contains an error regarding the electron spectra displayed in Figs. 4 and 5 and the data extracted from these spectra. The measurements were made with the SESAME magnetic spectrometer, the working principle of which is recalled in Fig. 1. Specifically, a magnetic dipole is used to separate charged particles (electrons in the case of this experiment) depending on their energy, charge and mass. The deflected particles then hit an imaging plate (IP) and deposit energy in its sensitive layer. The kinetic energy of the particles can be evaluated from their impact position on the IP and their number can be inferred from the local energy deposition.

There are three available IP positions on the SESAME spectrometer, located at respectively 0.5, 1.5, and 2.5 cm away from the magnetic dipole. The electron spectra were extracted assuming that the corresponding IP was located at the closest position to the dipole, when it was actually at the furthest position. As can be seen in Fig. 1, this entails a significant error in the evaluation of the kinetic energy from the impact position. The corrected spectra for the three shots are shown in Figs. 2 and 3 together with Maxwellian fits.

The corrected number of electrons above 2.5 MeV and total ejected electronic charge infered are displayed in Tables I and II. The new electron number above 2.5 MeV is close to that previously given, since the amplitude of the spectra is determined by the PSL value in each pixel, and the PSL response function of the IPs to electron irradiation is roughly constant above 2 MeV. Although the revised hot-electron temperatures are lower than the original estimates, the main conclusion of the article remains unchanged: the hot-electron temperatures are higher than those expected from the standard ponderomotive scaling.

We would like to sincerely thank Lucas Ribotte (CEA/DAM/DIF) for reanalyzing the data and contributing to the correction of the results presented in this erratum.