Read novels set in the period. In my case I had already read lots of inspiring nautical adventures such as Patrick O'Brian and CS Forester, but I have read many more since I started the research.
I also had to read all the older and current maritime novels and adventures aimed at children, even if they're bad: firstly to make sure I wasn't replicating anything, and secondly to get the hang of the vocabulary and feel of the reading age (9 - 12).
Read history texts until my eyes fell out. Looked at maps, original (or facsimile) manuscripts, engravings, paintings, newspapers and pamphlets - anything. Read other stuff - tangential but interesting histories - because you never know what you might find. I didn't know about the uprising against the French invasion of Malta when I first started writing book one: I just stumbled across it, and found it so fascinating that it became central to the plot of the trilogy.
Once the narrative and the sense of time and place is clear, there's an awful lot of referencing, fact-checking and map-staring that has to happen. This can be particularly difficult if you're stupid enough to set three books on the other side of the world, and live in a city without a vast collection of references on Malta. The internet helps a great deal, of course, and through it I found brainy people in Malta who could answer dumb questions for me.
But the web can also mislead. Many websites (like my own) are written by enthusiastic amateur historians - even Wikipedia. This is a great and wonderful thing, unless you're relying on them for absolute accuracy. They will sometimes be wrong. So will the professionals, even in books. I read about four different locations for the church in Mdina where the uprising took place, for example, some not even in Mdina at all. I couldn't be sure until I stood outside it.
I keep a spreadsheet of real life action tracked alongside fictional action, which includes things like seasonal changes (which wind will be prevailing, for example) and actual events. Sometimes I needed to track the action and characters hour by hour - other times it's week by week. This is particularly important in books two and three where the characters get more caught up in real life events on Malta, as well as lots of fictional events.
I didn't keep proper records of where I'd found certain items of information (I got bored with keeping card files, which is what I usually do) and as a result drove myself completely mad looking up things all over again.
4. Stand there
I didn't feel that everything was right until I could stand in the limestone dust of Malta and feel the sun on the back of my neck and stare at the sea and just - know.
From now on I am always going to base my books somewhere fabulous so I have to go visit. Often.
5. Check everything again
Redrafting can be as much about checking and refining information as it is about language and character. You end up taking out a lot of those historical details that seemed so critical at the beginning, and I spent a lot of time working on how to convey information without it feeling like a history lesson. Looking back, I think I got better at it by halfway through book two.
Even though the narrator, Lily's, voice is really rather modern, I tried to check the etymology of every phrase and significant word to avoid glaring anachronism. I double and triple-checked maps, dates, language, clothing, food, ship details - everything. I hope. No doubt there's something stupid stuck in there somewhere.
This is how editing works. The manuscript is edited, then I check it, then it is finalised by the in-house editor, and then typeset (beautifully) and I check the pages again, then they are proof-read, then the editor looks through them one last time.
In the early stages, I can still fix things that I've realised aren't quite right, say if I've woken up in the middle of the night in a cold sweat because I'm unable to remember which arm Nelson lost. Editors can ask clever questions like "Why is the candle burning when it's broad daylight?" (Answer: Because the author is an idiot).
I'm a (magazine) editor by trade, so I do this stuff for a living and my work ought to be flawless - and there's still a bloody typo on page 89. No, don't look.
Sandra Gulland, who wrote a successful trilogy on Josephine B, has a great website, and she records some of her less reliable research methods, all of which I also did:
I spent too much money on books;
I collected tacky memorabilia;
I travelled long distances to go to museum shows;
I grew teary-eyed on the cobblestones of Paris...