Two-point lapse rate

[Peter9192]

EWED profiles are not smooth enough to map directly to a single fixed free troposphere lapse rate.
Instead, we want an option to set the free troposphere lapse rate as a piecewise linear function. For the timing being with two segments only.

Profiles will look like:

h: 0, h, h, z_ft, z_ft+400
theta: [theta, theta, theta+dtheta, theta+dtheta+gamma_theta * (z_ft-h), theta_dtheta+gamma_theta*(z_ft - h) + gamma_theta2 * 400
q: similar

Note: if h > z_ft, we skip that segment.

So we need to add z_ft (initially as number, could later be extended to number array), gamma_theta2, and gamma_q_2.

[Peter9192]

Alternatively, we could specify z_ft to include the random 400 number that I used now, so we always have a defined top. Then, we could make gamma, q, and z_ft number arrays of equal length. I think this is cleaner and easier to extend later on.

[Peter9192]

Looks like example specified different profiles for different variables, that makes it much less elegant.. We will get z_ft_theta, z_ft_q, z_ft_u, z_ft_v.

[Peter9192]

Closed in #144