We all know that we put LESS air per second into the instrument the higher we go. And because the frequency is higher, the amount of air contained in each wave cycle is EVEN LESS. So as we raise a note the momentum of air leaving the mouthpiece is less than the momentum of the air in the return waves from the old note. Where they collide at the antinodes, there is therefore a nett momentum left over - a force on the air column acting back towards the mouthpiece that pushes the antinodes, and hence the nodes and hence the entire air column back towards the mouthpiece shortening the wavelength and increasing the frequency.
Another research paper in the offing. I expect a co-authorship credit