Simulations of decaying two-dimensional turbulence suggest that the on
e-point vorticity density has the self-similar form P omega similar to
t f(omega t) implied by Batchelor's (1969) similarity hypothesis, exc
ept in the tails. Specifically, similarity holds for \omega\ < omega(m
), while p(omega) falls off rapidly above. The upper bound of the simi
larity range, omega(m), is also nearly conserved in high-Reynolds-numb
er hyperviscosity simulations and appears to be related to the average
amplitude of the most intense vortices (McWilliams 1990), which was a
n important ingredient in the vortex scaling theory of Carnevale et al
. (1991). The universal function f also appears to be hyperbolic, i.e.
f(x) similar to c/2\X\(1+qc), for \x\ > x, where q(c) = 0.4 and x* =
70, which along with the truncated similarity form implies a phase tr
ansition in the vorticity moments [GRAPHICS] between the self-similar
'background sea' and the coherent vortices. Here c(q) and c are univer
sal. Low-order moments are therefore consistent with Batchelor's simil
arity hypothesis whereas high-order moments are similar to those predi
cted by Carnevale et al. (1991). A self-similar but less well-founded
expression for the energy spectrum is also proposed. It is also argued
that omega(s) = x/t represents 'mean sea-level', i.e. the (average)
threshold separating the vortices and the sea, and that there is a spe
ctrum of vortices with amplitudes in the range (omega(s),omega(m)). Th
e total area occupied by vortices is also found to remain constant in
time, with losses due to mergers of large-amplitude vortices being bal
anced by gains due to production of weak vortices. By contrast, the ar
ea occupied by vortices above a fixed threshold decays in time as obse
rved by McWilliams (1990).