Abstract
Energy spectra of high–altitude atmospheric turbulence data were analyzed. Kolmogorov's similarity theory for the inertial subrange was applied to infer energy dissipation rates from the spectra. The dissipation rates increase as the third power of the truncated root-mean-square values of turbulence: thus, the former can be inferred directly from the later with relative case. In the stratosphere the dissipation rates can occasionally he of the same order as those in the surface layer; a single curve can hardly represent the dissipation state in the atmosphere.