Delhi, the capital of India, is prone to severe seismic hazards, not only from local events but also from Himalayan earthquakes at distances of 250–300 km. Standard techniques are not sufficiently reliable to completely characterize the seismic hazards in this case due to the difficulty of predicting the occur- rence of earthquakes (frequency–magnitude relations) and of properly treating the propagation of their effects (attenuation laws), especially their long-period components. In order to give a sound description of the seismic ground motion due to an earthquake in such a given range of distances (and magnitudes), we use model- ling techniques developed from physics of the seismic source generation and propagation processes. Such models take into account the directivity effect of rupture propagation and the attenuation of (long-period) ground motions. The generated ground motion scenarios permit us to build a very important knowledge base to be fruitfully used by civil engineers, since long period ground motions, especially if amplified by deep sedimentary basins, can represent a severe threat for large scale structures (e.g. lifelines and bridges) and tall buildings, which are widespread in fast-growing megacities. In this study, we simulate the ground motion, at bedrock level, in Delhi city, for an earthquake scenario corresponding to a source of Mw = 8.0 located in the central seismic gap of Himalayas, at an epicentral distance of about 300 km from Delhi city. By means of several parametric studies, we simulate the time histories using Size Scaled Point Source, Space and Time Scaled Point Source and Extended Source models. Together with the complete time histories (displacements, veloci- ties and accelerations, from which the peak amplitudes have been extracted), we have also used the displacement response spectrum to characterize the seismic input at Delhi. Not only is the dis- placement response spectrum of great significance to modern displacement-based design engineering approaches, but it is probably the best parameter by which to characterize the destruc- tiveness potential of earthquakes located at such great distances from the target sites (of the order of 300 km), since the energy of the seismic input is mainly concentrated at long periods (in general, greater than 1 s) and it cannot be determined by straightforward integration of velocity or acceleration response spectra.
