Science starts with observations, then hypotheses, then tests. If tests start to yield positive results, then we can start to develop a theory (a word meaning "explanation of a phenomenon supported by experiment and observation" not "something that occurred to someone that might make sense"). Proper theories should make testable predictions, and only the theories that can make a lot of successful predictions hold up over time.
The spherical Earth/Sun-centered model can predict eclipses with amazing precision decades in advance. RE scientists can tell you when, where, how long, and what it will look like. For example, a total solar eclipse will just about come over where I live on August 12, 2045. Many more in many places will happen before then.
https://www.timeanddate.com/eclipse/list.htmlYou can even look up when the next eclipse will come over your house and what it will look like. Try it and see if RE scientists (using computer models programmed with RE science) are right.
Here's another little test that you can run: Every time one of these pops up, check the phase of the Moon. Solar eclipses always and only happen right at the hour of the new moon, and lunar eclipses always and only happen at the full moon. Always.
(The flat Earth model predicts that FE supporters will make up random stuff whenever they get cornered by actual science. Let's see if that prediction comes true, too!)