Incorrect L2norms for 3DH1D simulations
The L2-norms of flow variables (u,v,w,theta) are incorrect in 3DH1D expansions (x-y spectral/hp + z Fourier).
Given a box of fluid in (x,y,z) \in (PI x 2 x 1) with values of u,v,w,theta defined as 1 everywhere in the domain, L2(u,v,w,theta) should be
\sqrt(Lx * Lz * Lz) = 2.50663.
However, a value of 1.77245 was reported instead.
PS: The L2-norms for 2D simulations are correct.