We discuss the prospects for improved upper limits on neutrino masses that may be provided by a core-collapse supernova explosion in our galaxy, if it exhibits time variations in the neutrino emissions on the scale of a few milliseconds as suggested by recent two-dimensional simulations. Analyzing simulations of such neutrino emissions using the wavelet technique adopted in [1], we find that an upper limit m_nu ~ 0.14 eV could be established at the 95% confidence level if the time variations in emissions were to be preserved during neutrino propagation to the Earth.

Supernova explosions provide the most sensitive probes of neutrino propagation, such as the possibility that neutrino velocities might be affected by the foamy structure of space-time thought to be generated by quantum-gravitational (QG) effects. Recent two-dimensional simulations of the neutrino emissions from core-collapse supernovae suggest that they might exhibit variations in time on the scale of a few milliseconds. We analyze simulations of such neutrino emissions using a wavelet technique, and consider the limits that might be set on a linear or quadratic violation of Lorentz invariance in the group velocities of neutrinos of different energies, v/c = [1 \pm (E/M_{nuLV1})] or [1 \pm (E/M_{\nuLV2})^2], if variations on such short time scales were to be observed, where the mass scales M_{nuLVi} might appear in models of quantum gravity. We find prospective sensitivities to M_{nuLV1} ~ 2 X 10^{13} GeV and M_{nuLV2} ~ 10^6 GeV at the 95% confidence level, up to two orders of magnitude beyond estimates made using previous one-dimensional simulations of core-collapse supernovae. We also analyze the prospective sensitivities to scenarios in which the propagation times of neutrinos of fixed energies are subject to stochastic fluctuations.

We analyze the timing of photons observed by the MAGIC telescope during a flare of the active galactic nucleus Mkn 501 for a possible correlation with energy, as suggested by some models of quantum gravity (QG), which predict a vacuum refractive index \simeq 1 + (E/M_{QGn})^n, n = 1,2. Parametrizing the delay between gamma-rays of different energies as \Delta t =\pm\tau_l E or \Delta t =\pm\tau_q E^2, we find \tau_l=(0.030\pm0.012) s/GeV at the 2.5-sigma level, and \tau_q=(3.71\pm2.57)x10^{-6} s/GeV^2, respectively. We use these results to establish lower limits M_{QG1} > 0.21x10^{18} GeV and M_{QG2} > 0.26x10^{11} GeV at the 95% C.L. Monte Carlo studies confirm the MAGIC sensitivity to propagation effects at these levels. Thermal plasma effects in the source are negligible, but we cannot exclude the importance of some other source effect.

We correct the fitting formula used in refs. [1,2] to obtain a robust limit on a violation of Lorentz invariance that depends linearly on the photon energy. The correction leads to a slight increase of the limit on the scale of the violation, to M > 1.4 x 10^{16} GeV.

We argue that in fat brane-world scenarios the light propagating in vacuum will, because of massive ``Kaluza-Klein'' (KK) excitations, experience a refraction. The motion of a photon inside a fat brane can be decomposed in the longitudinal and transverse directions with respect to the surface of the brane. Since the light observable propagation is related only with the longitudinal motion, the observed speed of light depends on the value of the momentum transverse fraction contributing as the massive KK excitations. This is directly connected with the energy of the particles emitting the light, and hence with the frequency of the light itself. Using recent results on the arrival times of radiation of different energies from the measurements of gamma-ray bursters with known redshifts, we establish the limit M>620 TeV on the inverse thickness of the brane, and thus on the masses of the KK excitations. This limit exceeds by at least one order of magnitude the typical energy scale currently in use to characterize brane phenomena in the realm of future colliders.

We constrain the possibility of a non-trivial refractive index in free space corresponding to an energy-dependent velocity of light: c(E) \simeq c_0 (1 - E/M), where M is a mass scale that might represent effect of quantum-gravitational space-time foam, using the arrival times of sharp features observed in the intensities of radiation with different energies from a large sample of gamma-ray bursters (GRBs) with known redshifts. We use wavelet techniques to identify genuine features, which we confirm in simulations with artificial added noise. Using the weighted averages of the time-lags calculated using correlated features in all the GRB light curves, we find a systematic tendency for more energetic photons to arrive earlier. However, there is a very strong correlation between the parameters characterizing an intrinsic time-lag at the source and a distance-dependent propagation effect. Moreover, the significance of the earlier arrival times is less evident for a subsample of more robust spectral structures. Allowing for intrinsic stochastic time-lags in these features, we establish a statistically robust lower limit: M > 0.9x10^{16} GeV on the scale of violation of Lorentz invariance.