As I got closer to breaking ground on this observatory a few choices about design had to be made.
Background Research
In the last few years, I've looked at a number of sources for information on observatory builds for inspiration. This includes the books "Building a Roll-Off Roof or Dome Observatory" by John Stephen Hicks, "Setting up a Small Observatory" by David Arditti (both part of Patrick Moore's Practical Astronomy Series), the "How to Build an Observatory" pamphlet from Outsideology, and the build plans from both SkyShed and Backyard Observatories.
Dome or Roll Off Roof?
As I mentioned in my earlier discussion of the weather, I expect wind will be a major factor for the telescope unless the observatory can provide significant protection. For this reason, I'd thought about using a dome style observatory early on and I'd looked at some of the commercial options available.
In the end, I decided against using a dome. I'm not really a big fan of domes for small observatories. It seems like many people build them for the look. I also worry about the chimney effect of having any heat in the structure funnel out through the dome slit in front of your telescope and impact the seeing. Finally, I prefer the simplicity of a roll off roof. The control system and electronics are far simpler because there's no need to bridge electricity across to the rotating part of the dome in order to operate the shutter. While this is not too difficult these days (during my time as a postdoc I built a pair of robotic domed observatories), I wanted to keep my own build as simple as possible. As a result, I decided that I was willing to live with the lesser wind protection of a roll off roof.
As a result, I'll keep the walls relatively high to shield the telescope from wind. This actually provides another advantage: the telescope will be able to slew around inside the closed observatory. This is important because if the roof can not hit the telescope at any position, then it will be able to close without waiting for the telescope to park. This is really nice because if there is a problem with the telescope mount or computer control of the mount, you may not be able to close the roof in case of bad weather. That entire class of problem goes away if you can open and close the roof without worrying about the telescope's position.
In addition, one can connect the weather station to the roof controller directly and have the weather station trigger a close without going through the control software on the computer. This eliminates another entire class of failure: if the computer has crashed, the system will still close in the event of bad weather. It will be much easier for me to sleep while the system is in operation if this is the case.
Concrete Slab or Deck Flooring?
This question was always a relatively easy one for me. Concrete is a thermal mass which would radiate heat away during the night and cause seeing issues. The big research telescopes are built on concrete foundations, but they are also air conditioned during the day so that the interior (including the concrete flooring) is kept at nighttime temperatures.
To be fair, I doubt this would have a big negative effect on a small observatory like the one I have in mind. In addition, while being on the "dry side" of Waimea means I have some clear skies, it also means that I'm on the lee side of a mountain range, so I don't expect great seeing on a typical night. Just going out and watching the twinkling of stars by eye, I can tell there are a number of nights with quite poor seeing that I'll be dealing with.
Finally, I didn't want to deal with arranging for a cement truck and pouring a large cement pad. It will likely be more work to build footings and a deck, but I prefer that solution.
Dimensions
How big should this be? To get a starting idea, I set up a couple of my telescopes and measured the rough swing area to get an idea of how much clearance I'd need. My Stellarvue SVX152 refractor with an eyepiece and diagonal on a G-11 mount needs about a 3 foot radius to avoid collisions. It'll need additional space to accomodate an observer. Call that an extra 2 feet radially at minimum and you get about a 5 foot radius. My 16 inch f/4.5 TeleKit dobsonian also has a swing radius of about 5 feet. While I don't expect to set up the dob in the observatory, it is a nice reference to have in case I change my mind. Fortunately, with a dob, the observer stands next to the eyepiece and not behind it, so I don't need much extra space to accomodate the observer.
Given those numbers, I'm looking at something at least 10 foot square. Given that I'm now building a bit bigger (both in footprint and height) than the small, imaging only enclosure I'd originally envisioned. I think I'm going to base my build on one of the pre-existing plans I've purchased rather than design my own from scratch. I've heard good things about both SkyShed and Backyard Observatories (BYO), but the CJE1 plans I bought from BYO are a bit roomier. The largest SkyShed in the plans is 10x12 feet and the BYO is 11.5 x 11.5 feet.
I actually considered going even larger in order to accomodate more storage and even space for a small work area and was looking at the BYO CJE2 model which is 11.5 x 15.5 feet, but I decided it would be better if this plan stopped growing at some point, so I decided to stick with the CJE1.