I've wanted to do this for a long long time. June 21, 2017 and the era of Fake News was the kick in the pants that finally got it started.

The Earth goes around the Sun, but it's not a circular orbit. Around the first of the year we are closest to the Sun, and in early July the farthest. That means that the speed of the Earth going around the Sun changes. Faster at the beginning of the year and slower in the middle. If you have a clock that ticks off 24 hours at the usual rate, and a way to figure out the position of the Sun each day you can show that the Sun does not arrive 'straight overhead' at the same time every day, because of the change in speed. On top of that the same setup can show how the Sun moves north and south during a year, because of the tilt of the Earth. It's all interrelated. Allegedly.

Since you can't trust anyone these days I built an analemma generator to prove or disprove the paragraph above.

The office building where I work has a small patio on the south side. It has a gutter along the roof that is magnetic. There is a cement ramp that runs north and south from the parking lot up to the patio.

Behold...the analemma device! If you guessed that this device was made of an old hard disk drive magnet, a couple 1" standoffs, and section of spoon handle from a plastic Sonic Drive-In spoon then you would be 100% correct. It does what it needs to do, and it was all I could come up with in a 1/2 hour.

At about 12:30pm on the 21st of June it hit me to start this project. The 21st was the Summer Solstice in these parts in 2017. With the Sun all of the way north what better time to start? I flapped around for a bit and finally came up with the apparatus in the picture above.

Below is vandalism of government property with a Sharpie. With the device in place on the gutter I waited until 1:00:00pm MDT and marked the spot on the cement ramp in the center of the not very well focused, yet still eyeballable, square made by the magnet, standoffs, and spoon handle. The dot "A" is that first dot.

I decided to mark the spot every Wednesday at 1pm. Why 1pm? Because the humans on the planet decided to mess with mother nature and set their clock ahead one hour during the summer. When we go back on standard time then dots will be plotted at 12pm. The date and time for the dots are below the picture. The dot between C and D is fake. You just can't get away from that stuff.

A - 2017-06-21 1300 MDT, 1900 GMT

B - 2017-06-23 1300 MDT, 1900 GMT  Friday. Was just checking to see if the Sun was working.

C - 2017-06-29 1300 MDT, 1900 GMT  A Thursday. I forgot on Wednesday.

D - 2017-07-05 1300 MDT, 1900 GMT  Back on schedule. Thanks, iCal!

E - 2017-07-12 1300 MDT, 1900 GMT  I keep forgetting to throw in a ruler. Next time.

F - 2017-07-19 1300 MDT, 1900 GMT  Where's the ruler??

G - 2017-07-26 1300 MDT, 1900 GMT  Just ignore that extra dot. A ruler!

H - 2017-08-02 1300 MDT, 1900 GMT  All ducks in a row! Starting to speed up. I'm gonna need a longer ruler. The G dots last week were just plain wrong, because of stupidity. They've been corrected.

I - 2017-08-09 1300 MDT, 1900 GMT  Another sunny noon. Apparent solar time was about 1146, meaning that the Sun would not be 'straight over head' for another 14 minutes.

J - 2017-08-16 1300 MDT, 1900 GMT  A cloud got out of the way 5 seconds before to about 5 seconds after 1pm. That was close. Apparent solar time was about 1148.

K - 2017-08-23 1300 MDT, 1900 GMT  Still there after the eclipse. Apparent solar time was about 1150.

L - 2017-08-30 1300 MDT, 1900 GMT  Apparent solar time was about 1152.

M - 2017-09-06 1300 MDT, 1900 GMT

N - 2017-09-13 1300 MDT, 1900 GMT  Apparent solar time was about 1156.

O - 2017-09-20 1300 MDT, 1900 GMT  Apparent solar time was pretty close to 1200. I'm going to have to figure out how to put letters next to the dots way up at the top. May have to change to two pictures. That's 4'8" from top to bottom. Equinox was last week. N-O isn't really as much farther than M-N as the picture implies, but it is a little farther.

P - 2017-09-27 Wednesday  Raining and cloudy. Drew all kinds of dots and lines to try and figure out where The Dot should be, but I don't think I was even close. The line with five dashes does look like it was pretty close to being the edge of the roof, but at the time I wasn't sure.

Q - 2017-10-04 1300 MDT, 1900 GMT  It almost ended up like last week, but caught a break in the clouds just before 1PM. It got cloudy again after 1. Forgot to look at the apparent solar time.

R - 2017-10-11 1300 MDT, 1900 GMT  Apparent solar time was almost 1206.

S - 2017-10-18 1300 MDT, 1900 GMT

T - 2017-10-25 1300 MDT, 1900 GMT  Apparent solar time was about 12:08:15. We had a party on the patio, so had to wait until later to take the picture. The dot was under one of the the tables.


U - 2017-11-01 1300 MDT, 1900 GMT  Apparent solar time was about 12:09:40. The line of dots is about 9-1/2' long. Next week daylight saving time goes off.

V - 2017-11-08 1200 MST, 1900 GMT  Apparent solar time was about 12:09:30. The line of dots is about 9-1/2' long. Daylight saving time cancelled! OK. Until next year. Looks like things are starting back toward the middle.

W - 2017-11-15 1200 MST, 1900 GMT  Apparent solar time was about 12:07:45.

X - 2017-11-22 1200 MST, 1900 GMT  Apparent solar time was about 12:06:30? I looked, but don't remember. Distance between the dots are getting shorter and the line is starting to curve back to the right. Seems like a lot will have to happen in a month. We must be going fast. I'll have to figure out how to take a good picture of the whole thing when we hit the solstice next month. Picking up some fall flora...

Y - 2017-11-29 1200 MST, 1900 GMT  Apparent solar time was about 12:03:45. Kind of a big change.

Z - 2017-12-04 1200 MST, 1900 GMT  Marked this point on Monday, because I wasn't sure what the weather was going to be on Wednesday. Tuesday was bad.

a - 2017-12-06 1200 MST, 1900 GMT  Very sunny today. Apparent solar time was 12:01:05. Directions are changing fast. Hang on!

b - 2017-12-12 1200 MST, 1900 GMT  Tuesday. This dot really should be somewhere in the crack.


d - 2017-12-13 1200 MST, 1900 GMT  Nice day. Will skip letters like "c", because it looks like "C". The shadow is getting kinda diffuse, because it is so far away from the device on the roof. Yup. Dot b was not in the right place.

e - 2017-12-16 1200 MST, 1900 GMT  Saturday, since I was here.

f - 2017-12-19 1200 MST, 1900 GMT  Apparent solar time was 11:55:05. Tuesday. It was cloudy yesterday. My plotting skills are getting worse. I don't think "e" was where I thought it was.

g - 2017-12-20 1200 MST, 1900 GMT  One more day and back we go.

h - 2017-12-21 1200 MST, 1900 GMT  The solstice was at 9-something, so just a couple of hours before this. I'm already starting to lean the other way. I corrected point "b" in this picture. Virtual dots can be placed in expansion joints. The distance from A to h is 166". There is a bunch of little dots below "h". It was partly cloudy and, of course, the clouds were right where the Sun was. I marked the shadow position every minute for the nine minutes before noon in case I had to make a guess. The clouds let the light through right at noon, and then went away four minutes after that. Figures.

i - 2017-12-27 1200 MST, 1900 GMT, apparent solar time 11:51:55. Headin' on back. The "b" dot is those two dots.

j - 2018-01-03 1200 MST, 1900 GMT, apparent solar time about 11:47:50. No wonder my neck hurts. We are on the move.

k - 2018-01-10 1200 MST, 1900 GMT

m - 2018-01-18 1200 MST, 1900 GMT Thursday, apparent solar time about 11:41:55. Was too cloudy yesterday, so had to wait until today. Yes, that picture is upside down. It was taken from the sunny side. It's making me a little seasick.

n - 2018-01-24 1200 MST, 1900 GMT Thursday.

p - 2018-01-31 1200 MST, 1900 GMT, apparent solar time about 11:41:55.

q - 2018-02-07 1200 MST, 1900 GMT, apparent solar time about 11:37:55. Maybe.

r - 2018-02-07 1200 MST, 1900 GMT. Did the dot on Tuesday. Weather didn't look, and wasn't, good the rest of the week.

t - 2018-02-21 1200 MST, 1900 GMT, apparent solar time 11:39:03. Clear, but cold and breezy.

y - 2018-03-03 1200 MST, 1900 GMT. Saturday!? It was cloudy on Wednesday, and I forgot and just missed noon on Thursday and Friday.

AA - 2018-03-07 1200 MST, 1900 GMT, apparent solar time 11:41:40. Nice day.

BB - 2018-03-14 1300 MDT, 1900 GMT. Back to messing with Mother Nature and going on daylight saving time. The perspective here makes this look a bit funny. From the starting point to the crack in the sidewalk is just a couple of feet, but from the crack to the far end of the curve is about 11 feet.

CC - 2018-03-21 1300 MDT, 1900 GMT, apparent solar time 11:45:15. Nice day. The usual Wednesday dot.
VE - 2018-03-20. Yesterday the vernal equinox was at about 1015, so I made a dot at 1300. One more quarter to go.

DD - 2018-03-28 1300 MDT, 1900 GMT

EE - 2018-04-06 1300 MDT, 1900 GMT
FF - 2018-04-07 1300 MDT, 1900 GMT

GG - 2018-04-11 1300 MDT, 1900 GMT

HH - 2018-04-17 1300 MDT, 1900 GMT

II - 2018-04-18 1300 MDT, 1900 GMT

JJ - 2018-04-24 1300 MDT, 1900 GMT
KK - 2018-04-25 1300 MDT, 1900 GMT

LL - 2018-04-30 1300 MDT, 1900 GMT

MM - 2018-05-10 1300 MDT, 1900 GMT
NN - 2018-05-16 1300 MDT, 1900 GMT Back to Wednesday. Have been all over the week lately.

OO - 2018-05-23 1300 MDT, 1900 GMT A month to go!

PP - 2018-05-30 1300 MDT, 1900 GMT

QQ - 2018-06-06 1300 MDT, 1900 GMT I'm no expert, but it sure looks like the dots are going to meet.

RR - 2018-06-11 1300 MDT, 1900 GMT

SS - 2018-06-13 1300 MDT, 1900 GMT
TT - 2018-06-16 1300 MDT, 1900 GMT
UU - 2018-06-19 1300 MDT, 1900 GMT Tried to get in a few extra dots near the end. The summer solstice was at about 0400 MDT on the 20th, so UU is the day before solstice. Now, why doesn't UU and A line up? Well, you can see above the leaf on the left what I was using for a shadow to mark the dots throughout the year. It's a bit sketchy. When I started the dots I was using the middle of the square, but the edge of the roof. At some point I started eyeballing the center of the square, so UU is a little more south than AA. Operator error.

VV - 2018-06-20 1300 MDT, 1900 GMT This is the red 'shadow' to the right of dot UU. This dot is about 9 hours after solstice. Doesn't appear to be going back yet. Guess that's why they call it a solstice.
WW - 2018-06-26 1300 MDT, 1900 GMT WW is a little misaligned with respect to C for the same reason as UU.