One of the key reasons Ernest Hemingway is a compelling author is because he writes with the authority of someone who was actually there.
The retreat and collapse of the Italian army in A Farewell To Arms (Hemingway's best novel in my view) was vivid because it contained a level of detail that could only come from an eyewitness. Likewise, The Old Man and The Sea and To Have and Have Not gained immeasurably thanks to his intimate knowledge of the people and environs of Cuba and Key West.
In short, Hemingway did his research. It may not have been "research" in the academic sense, but it was serious and deep research nonetheless.
Barbara Kingsolver also does immense research for her novels. In any other author's hands, The Poisonwood Bible would have been an overreach. The setting and historical backdrop--the Congo in the 1950s--would have been too exotic for almost all authors describe believably.
Kingsolver's research wasn't limited to place names and descriptions. She tied in contemporaneous political events, included vivid details on the flora and fauna, and even incorporated linguistic details into her story. Even her lesser-known works (Prodigal Summer comes to mind) contain agonizingly well-researched details on the insect and mammal life of Appalachia, all of which which add to the realism of her writing. I've read only four of her novels, but Kingsolver strikes me as the kind of author who has never taken a research shortcut in her life.
It goes without saying that you should never write about a subject unless you know it well. What's less obvious is even when you do know a subject well, you probably don't know it well enough.
Do your research, and then do more research. Just one mistake will reveal to your readers that you don't really know what you're talking about.