Initial description of SUNCALC and some follow-up information ============================================================================== Date: Thu, 28 Apr 94 20:10:05 CDT From: rusin (Dave Rusin) To: rec-gardens@cs.utexas.edu Subject: How to answer, "Is there enough light?" I'm making a short computer program available for your planning pleasure. A recent post concerning shady gardens suggested the comment that gardens in the far north, or on the north side of a house, don't get enough sunlight to grow certain plants. Of course, the most convincing way to decide this is: try it and see. If instead you want to try and decide this in advance, you may want to try and figure out how much light falls in your area at certain times. As it turns out, this very question came up this winter in misc.consumers.house in regard to how much heat you can expect to gain from sunlight through windows and skylight. There too, the answer is really "it depends", since things like the position of your neighbor's house, or the fraction of the time your sky is overcast, can significantly affect the answer. But the main ingredient is the determination of the sun's apparent position in the sky given your location, the time of day, and the time of year. Factual evidence to the contrary notwithstanding, I couldn't believe that we in the northern hemisphere would ever have direct sunlight thru the north windows; being a mathematician I thought I ought to figure it out. Now that spring is here I am glad to see the calculations agree with experience: yes, the sun does rise north of east. Thinking that others may need to investigate this too, I wrote up a little program to compute things like sun position, length of [sunlit] day, and relative energy gain from sunlight. It's written in a variant of BASIC so you can adapt it at will (specifically, it runs under the delightful and publicly-available interpreter UBASIC, available in the SimTel PC archives). If you want to play with this, I have left it out for anonymous FTP or gopher access at: math.niu.edu in the directory /pub/papers/Rusin. It's called "suncalc.ub" (ASCII file). As is often the case I suspect no one else will be as interested as me but I found it amusing to try to figure out how, for example, a 12hr day here changes in a continuous way to a 6 month day inside the polar circles. If you have comments or questions, use the address below to reach me. I can also run calculations for you given latitude (degrees), day of year (in days since last equinox) time of day (in hours since local midnight), and (if you're thinking of solar collectors or greatly-sloped hills for landscaping) the direction your collector faces. dave rusin, rusin@math.niu.edu ============================================================================== Date: Sun, 1 May 94 08:39:17 CDT From: rusin (Dave Rusin) To: rjpn@cbnewst.att.com Subject: Re: here comes the sun I should add that even though you get many hours of sunlight on the north side of the house by late June, this doesn't help your garden as much as you might like. Remember that the sunlight, when shining on the north face, comes in at a low angle, so that a unit amount of sunshine gets distributed across a wider surface. You can think of it this way: near dawn and dusk, the plants tend to cast longer shadows on each other, so there is less light per plant than there is at noon on the south face. This "light collecting efficiency" is what I measure in the SUNCALC program under the heading "energy rate". ... ============================================================================== Date: Fri, 1 Jul 1994 13:04:34 +1200 From: John Harper To: rusin@mp.cs.niu.edu Subject: Re: Sunrise,sunset alogorithm Glad you replied to the one about sunrise/set. It would also help your users to know how big the effects you didn't consider can be (e.g. non-spherical Earth, atmospheric refraction) This can matter for legal purposes (a driver kills someone claiming to have been dazzled by setting sun: was it in fact visible from where he was at the time he said he was there? Another person claims he didnt see the other car as it didnt have its lights on. Was it more or less than 30 min after sunset (which is our legal lighting-up time?)) John Harper Mathematics Dept. Victoria University Wellington New Zealand ============================================================================== Date: Thu, 30 Jun 94 22:48:10 CDT From: rusin (Dave Rusin) To: John.Harper@vuw.ac.nz, rusin@mp.cs.niu.edu Subject: Re: Sunrise,sunset alogorithm Yes, thanks for the input. As far as I know the biggest error comes from neglecting the earth's revolution in the course of a day; a 1/365 change in the length of a day is 4 mins (or, obviously, 1/4 of 1%). On the face of it, the biggest change (of which I am aware) ought to be from the non-circularity of the earth's orbit; it is really quite close to circular, but when you work out the equations of the ellipse, there is a 3% change in the earth-sun distance from January to June, and so (by Kepler's laws) a 6% change in the earth's speed of revolution. But unless I am missing something, this is 6% change in a quantity which already has only a 2nd order effect. I hadn't even thought about the refraction of light. I suppose this makes the apparent position of the sun dependent on the frequency of light you think you see the sun with! I am glad others are responding with pointers to more detailed calculations to cover these other effects. Really, I only worked through the model in order to convince myself the the sun really could rise north of east here in the northern hemisphere (in the summer) -- this past winter I nearly convinced myself I had been hallucinating, and that the northern gardens could never get direct sunlight. I was wrong, of course. dave rusin@math.niu.edu ==============================================================================