Satellites using sun-synchronous orbits can circle the Earth multiple times in a day.Compared with a stationary observer on the ground,they are moving at a relatively fast speed.Therefore,as long as they delay the release of observational data by a random time period,there won't be this issue.Only geostationary satellites would have this problem.
But a bigger problem comes before the above issue:most of the current human meteorological satellites do not have stealth capabilities.You can see them directly.Perhaps your idea will become a practical problem when satellite stealth technology matures.
I did not see your comment before I left mine! But yes the second half makes sense.
I do think you might able to deduce the orbit even if the data release is delayed by a random time period. If you’re a foreign adversary that has its own satellites, you can measure the same information from a known orbit. Then one could compare the published data with one’s known dataset to deduce things like the angle from which the data was measured.
But a bigger problem comes before the above issue:most of the current human meteorological satellites do not have stealth capabilities.You can see them directly.Perhaps your idea will become a practical problem when satellite stealth technology matures.
This is a translation.