A radar simulator capable of generating time series data for a polarimetric phased array weather radar has been designed and implemented. The received signals are composed from a high-resolution numerical prediction weather model. Thousands of scattering centers (SCs), each with an independent randomly generated Doppler spectrum, populate the field of view of the radar. The moments of the SC spectra are derived from the numerical weather model, and the SC positions are updated based on the 3-D wind field. In order to accurately emulate the effects of the system-induced cross-polar contamination, the array is modeled using a complete set of dual-polarization radiation patterns. The simulator offers reconfigurable element patterns and positions and access to independent time series data for each element, resulting in easy implementation of any beamforming method. It also allows for arbitrary waveform designs and is able to model the effects of quantization on waveform performance. Simultaneous, alternating, quasi-simultaneous, and pulse-to-pulse phase-coded modes of polarimetric signal transmission have been implemented. This framework allows for realistic emulation of the effects of cross-polar fields on weather observations, as well as the evaluation of possible techniques for the mitigation of those effects.