Trying to nail down proper GCP density for a municipal infrastructure project. Client specs say "survey grade accuracy" but I'm seeing wildly different recommendations:

• Japanese regs: 24 GCPs/km? for 20cm accuracy
• Korean standards: 9 GCPs/km? for 2.9cm RMSE
• Some papers suggest 12 GCPs/km? at 500m perimeter spacing

What density are you actually using in practice? Using Pix4D with RTK-enabled drone, processing desktop version. Need horizontal RMSE <3cm, vertical <5cm.

Those standards assume different conditions. With good RTK and strong signal, I regularly hit <2cm horizontal with 4-6 GCPs per square km. The key is GCP placement - perimeter + center works better than uniform distribution.

New to this - why such huge differences between countries? Is it just different risk tolerance or are there technical reasons? I've been following the "more is better" approach but sounds like I might be over-doing it.

Recent research shows sub-centimeter GCP accuracy matters more than quantity. I use 12/km? for critical infrastructure but could probably get away with 8-9 if my GCP survey was perfect. Ground conditions and RTK strength affect this way more than published guidelines suggest.