In this paper, we will present new compression algorithms to determine optimal layer heights and turbulenceweights for the tomographic reconstruction in wide field AO systems. Among other approaches, a new compressionmethod based on discrete optimization of collecting atmospheric layers to subgroups is discussed. Furthermore,studies of the influence of layer heights and c2n-profiles on the reconstruction quality for differentreconstruction algorithms and atmospheric profiles will be shown. Our comparison suggests that reconstructionson fewer atmospheric layers yield comparable quality with lower computational effort, if an appropriatecompression algorithm is used. The numerical results were obtained on the ESO end-to-end simulation toolOCTOPUS.