This question is quite complex, and here's a somewhat imprecise way to develop an intuitive understanding:

Let's first discuss the patterns of changes in frequencies received by ground stations.
As shown in this figure, in the intermediate case, the source – i.e., the satellite – is closer to the observer, while the observer appears to be in a region with higher frequency. Conversely, when the satellite is further away from the observer (the rightmost case), the observer appears to be in a region with lower frequency. From a qualitative perspective, as the satellite rises and falls relative to the ground station, the derivative of its slant distance (u) with respect to time goes through the following process: u continuously increases, from negative to zero to positive. The received frequency is approximately f0(1-u/c), where f0 is the frequency of the source (satellite) and c is the speed of light. Thus, our received frequency, which is storedDownward frequencyIt must be from large to small.
It is worth noting that the Doppler effect of mechanical waves only relates toGaze speedThey are related. The recent infographic will give readers a false sense of Doppler effect and distance relationship.
Now, let's return to the analysis of the uplink frequency. It seems we can reverse their roles: the transmitter becomes a ground station, and the observer becomes the satellite. In the case of fixed-frequency satellite repeaters that we are discussing, the receiving frequency set by the satellite itself remains constant. If the frequency transmitted from the ground remains unchanged, as in the previous analysis, then the frequency received by the satellite should also beFrom largest to smallestHowever, as we just mentioned, the satellite's receiving frequency remains constant, so ground-based compensation is needed, meaning that when transmitting from the ground, the frequency needs to be changed to.From childhood to adulthoodThis is why the uplink frequency must be ordered from lowest to highest.
However, it should be noted that the figure cited in the answer isMechanical waveThe Doppler effect. Strictly speaking, the calculation of the frequency of mechanical wave's Doppler effect cannot be based on arbitrarily exchanging the relationship between the source and the receiver due to relative motion – we easily fall into the trap of thinking, "Since the source is moving towards us, the stationary observer, then when calculating, we can naturally switch to a scenario where the source is stationary and we are moving towards the observer," which is incorrect. This is because the propagation of mechanical waves depends on the medium; this relative motion is essentially the relative motion of the medium, and the exchange of the source-observer relationship does not take into account the existence of the transmission medium (treating the reference frame of the medium as a primary reference frame).
Furthermore, the Doppler effect of electromagnetic waves also has properties that differ from those of mechanical waves. First, electromagnetic waves themselves do not rely on a medium and therefore do not have a preferred reference frame based on the medium; it is meaningless to distinguish between the source and the observer as being at rest or in motion. Second, when calculating the Doppler effect of electromagnetic waves strictly, one must also consider the effects of special relativity, i.e., the relativistic effects make the Doppler effect need to take into account lateral movement (which would not have an impact under non-relativistic conditions). However, in the case of satellite motion, this low-order approximation is sufficient and does not require the introduction of higher-order relativistic effects.