Velocity and quality factor (Q) values are important model parameters for seismic imaging and interpretation. This dissertation describes two new methods for inverting these parameters through tomographic approaches from seismic reflection data. Chapter two describes reflection attenuation tomography, a method to constrain a depth model of Q distribution from reflection data in a layered media. The objective is to differentiate the seismic amplitude change due to attenuation from the other factors that also influence seismic amplitude and then invert for Q values using amplitude ratios between neighboring reflections. The use of the amplitude ratios between neighboring rays having the same source avoids the amplitude variation between different sources and radiation pattern around the source. A synthetic test and a field example indicate that this method can resolve well the depth models for Q values. Chapter three presents ray-count weighted multi-scale tomography (MST) to improve velocity field in the poor ray coverage area. Ray-count weighted MST parameterizes the study area into two groups of overlapping sub-models of different grid sizes. The inverted velocity perturbation value for a specific location is the ray-count weighted superposition of velocity perturbation values of two groups of sub-models, which emphasize the regional component in the area with small ray count and local component in the area with good ray coverage. In a case study with strong lateral variation in the shallow zone, the results show the artifacts using the single scale tomography (SST) were suppressed significantly by using ray-count weighted MST. Both migrated stack and common image gathers show improvements in the velocity field from ray-count weighted MST.