We construct an axisymmetric model for the Crab Nebula. The flow dyamics is based on the model by Kennel and Coroniti (1984), but we assume that the kinetic-energy-dominant wind is confined in the equatorial region. We reproduce spacially resolved spectra which agree well with the Chandra results. However, the reproduced image is not a ring but lip-shaped. In addition, brightness contrast between fore and back sides of the ring cannot be reproduced if we assumes that the sigma-parameter is as small as 10**-3. We suggest that the nebula field is highly disordered due to for instance magnetic reconnection. The estimate of sigma can be larger than previously expected.