2024 04/25
Jaymes Kenyon
A Primer on Numerical Weather Prediction

Jaymes discussed the historical developments in numerical weather prediction of V. and J. Bjerknes in 1922 in Norway and their development of theories. Jaymes goes on to describe The Primitive Equations and the fundamental equations to figure out six variables to stem further from. These equations also show tendencies and gradients which are also super important, “prognostic” and “diagnostic”. 

Next, he describes Model Discretization, the next step after the equations, and how time ties into it. The next major experiment in developing numerical weather prediction was a British man named Lewis Fry Richardsowhoch wasn't accurate in weather prediction but his math was on the right track.

Next is a Hungarian man named Max Margules he discovers the flaws in the previous theories to add new theories to them.

Then in 1943-1945 one of the first digital programmable computers was called the Electronic Numerical Intergrator and Computer (ENIAC) which was a breakthrough.

Another study was found in 1939 named Carl Rossby of more theories.

In 1950 the first successful operational NWP using the ENIAC.

All of this has led to NOAAs supercomputing productions today at 29 petaflops (29 quadrillion a minute).

Jaymes talks about the very complex and unresolved processes they have today and the things they can do to best estimate.

Jaymes concludes with how technologically advanced everything has advanced. As well as how the HRRR is being replaced by the RUFES (which Jaymes is currently a part of helping with the process of this new model). Jaymes also talks about how AI is being introduced but that it will be post-processing instead of driving it, that everything goes back to human calculations of the primitive equations.

You can view all of his presentation slides below.