Uplink vs downlink jamming

In plain terms

You can jam a satellite link at the satellite — then it affects everyone in its coverage area — or close to the user, and then only that area.

In practice

In Ukraine, Russian systems such as the Peresvet-M disrupt Starlink terminals around a site: a protected bubble of reportedly some 20 km² (Case study: Russian Starlink jammers (Peresvet-M)), not interference with the whole network. Military satellite links use directed beams and frequency hopping to resist such jamming.

The distinction determines the area of effect — compare the geometry of stand-off vs stand-in jamming (SEAD & DEAD):

Uplink jamming: you jam the satellite's receiver from the ground. If it works, it affects all users within the footprint of that satellite — a wide effect. The drawback: you need a lot of power to reach a satellite in a distant orbit.
Downlink jamming: you jam the user's receiver locally. Much easier (the user is close by), but the effect is limited to the area around the jammer.

Uplink jamming disrupts the satellite's receiver and affects everyone in the footprint; downlink jamming disrupts one user locally. Compare the stand-off/stand-in trade-off in SEAD (SEAD & DEAD).

Satellite dishes of a US 'space aggressor' squadron, which jams SATCOM links in exercises the way an adversary would. Photo: U.S. Air Force (1st Lt Charles Rivezzo) / Wikimedia Commons — public domain.

There is also spoofing/meaconing: falsified or rebroadcast signals — in the spirit of GNSS spoofing (GNSS and navigation warfare), but applied to SATCOM. And dazzling: temporarily blinding an optical reconnaissance satellite with a laser (Laser and military applications, Directed Energy Weapons), so that for a while it records nothing usable.

§21.3 Uplink vs downlink jamming