After scrutinizing my previous theory about Hyrule's changing geography (see Article 10: Hyrule's Geography), I've decided that some of my conclusions were false, while others simply need more explanation.
First of all, I'd like to apologize for not having a good map drawn yet. My friend Chris Cummins is a good artist, but the way I was having him draw the map was way too confusing. I wanted him to draw one map, dated from after LoZ, but I still wanted him to show all the old structures. Now I've changed my instructions, and am having him draw a map from each of three different time periods: OoT, ALttP, and LoZ/AoL. He will, hopefully, have these done soon, and as soon as I have them I will append them to the end of this article.
Secondly, I would like to refute a theory put forth by . He suggested that the area south of Death Mountain in AoL (the "Death Mountain Area"), is really the entire area from LoZ, and says that the official NES Game Atlas supports him. Maybe the Game Atlas does say this is true, but that doesn't automatically make it fact - I only consider manuals and games to be canon sources. Perhaps the game designers did intend to make the Death Mountain Area look like the overworld from LoZ, but this doesn't change the fact that this area is too small in relationship to the rest of the AoL map, and too different from the LoZ overworld map.
The Death Mountain Area does share similar features to the LoZ overworld - a body of water that that drains into the ocean, a lake next to a forest in the east, and a graveyard and forest in the west. However, there are many things that are different - where is Zora's Domain and the water flowing from in into the moat? why is the southwestern forest suddenly alive and bigger than the eastern Lost Woods? why is the graveyard inclosed within mountains, when in LoZ there were only mountains to the north? With so many unanswered questions, and with the fact that the geography of the Death Mountain Area would totally disharmonize with my theory, I am forced to assume that Mariner's, and the NES Game Atlas', assertions are not correct.
But most important of all is the following change to my theory. This image of the LoZ overworld comes from the game manual (it's been modified, of course, to remove the storyline text). You will immediately notice that the water from the moat runs not into Lake Hylia, as my theory suggests, but into the ocean. The in-game overworld map never shows this, so I simply ignored it at the time.
However, I don't think it can be ignored any longer. Look to the east of the moat and you will see two bodies of water. The northern one is a fairy fountain (I think). It makes perfect sense that the southern one is the dried-up remnants of Lake Hylia! It is in the correct location, after all, and the valley south of the lake even looks like the dried-out Great River, which used to flow from Gerudo Valley into Lake Hylia in OoT.
The moat itself also requires some explanation. Here's a theory: in ALttP, the castle was built in the middle of a moat which took up most of Hyrule Field. After ALttP, royal engineers decided that it would be more advantageous to have the moat flow into the ocean, so that trade ships could more easily reach the interior of the kingdom. Thus, Hyrule Castle was torn down and moved to the northern province of the kingdom (North Castle in AoL), and the moat island was split in half to accomodate more ships. By LoZ, the constant warfare of the Age of Chaos and Ganon's army reduced the kingdom to a shambles, so nothing is left of the engineers' dream but the two islands within the moat. Makes sense, doesn't it?