I'm demo-ing Carlson Survey. One area of interest is Solar Observations. I note some discrepancies between Carlson and MICA. Here's an example:
From MICA (Topocentric, Local Horizon):
12mar16, 1200 hrs UT1; 72 w. Long; 43 n. Lat; Angle to mark=0; Angle to Sun=10; SD=16'-05.6". That would give me 102-59-11.4 to the center, and 102-43-05.8 to the trailing edge of the sun. Subtract 10 degrees: 92-43-05.8 to the mark
From Carlson:
Same data; (True bearing; Trailing edge); Azimuth to the mark: 92-42-45, a difference of 20.8"
Is this attributable to the possibly less granular ephemeris in Carlson? Something else? Something I'm doing wrong in Carlson?
Thoughts?
FYI, Lee Green's old DOS program put it right between the two at 92-42-56.5...
I'd sure hate to have to start putting multiple answers from different ephemerides into Star*net for the same shot.:excruciating:
I wonder if they are all using the same kind of time, although it wouldn't make that much variation. The earth rotates 15 arc seconds per time second so UTC vs UT1 could make several seconds.
Topocentric versus geo-centric would change it by a few seconds.
Bill93, post: 363550, member: 87 wrote: I wonder if they are all using the same kind of time, although it wouldn't make that much variation. The earth rotates 15 arc seconds per time second so UTC vs UT1 could make several seconds.
Topocentric versus geo-centric would change it by a few seconds.
Interesting. I just ran MICA at one second (in time) intervals for that exact period of time and found that the change in azimuth was 10"+ per second of time. I didn't realize it was that much. I guess in the days that solar was "important", anything less than a minute of arc in azimuth was considered good.
That would mean that the Carlson and MICA data are only two seconds in time off or so, which I suppose is not that great, especially if one is going to do multiple observations and let Star*net do it's thing. Hard to believe the MICA database wouldn't be spot on; can't speak for Carlson. I did pose the same question of Carlson Technical Support but have not heard anything back yet. Curious.
Well here's some kudos to Carlson for good technical support:
Dean Goodman of Carlson responded with the following. He actually looked at the code and discovered that there seems to be a discrepancy in the Semi Diameter calculation between Carlson and the MICA database (or how I'm reading it). Here's what he says:
The ephemeris algorithm we use was written back in the mid 80's. I periodically check it to make sure it is still valid. It generally is less than 10 seconds off.
One thing I noticed is that you used a Sun Semi-Diameter of 16'-05.6". I debugged into our code and we call a function that calculates the Sun's semi-diameter. Below are the calculations:
int semi_calc(SOLAR_STRUCT *ss)
{
double h;
if(ss->shot_type == 0){ /* center of sum */
ss->semi_diameter = 0.0;
return(0);
}
h = (sin(ss->latitude) * sin(ss->dc)) + (cos(ss->latitude) * cos(ss->dc) * cos(ss->zt));
if( fabs(h) >= 1.0)
return(-2);
h = atan(h / sqrt((-h * h) + 1));
ss->semi_diameter = fabs(ss->semi_diameter / cos(h));
if(ss->shot_type == -1) /* trailing edge */
ss->semi_diameter = -ss->semi_diameter;
return(0);
}
The zt and dc variables are calculated elsewhere but this gives you the general idea.
The value this function gets for the Sun's semi-diameter is 16'-20" which would change your ephemeris azimuth to 92-42-51 which comes closer to matching Carlson's azimuth of 92-42-45 (within 6").
The formula for "h" seems identical to that shown in the Sokkia manual:
I thought it might be because I chose topocentric rather than geocentric in MICA, but that makes less than a tenth of a second difference in the SD.
Now it's bugging me. Why is there any discrepancy at all? This math has been done since Ptolemy did it.
Where's MathTeacher when you need him?:-S
Just did a check of Jerry Wahl's Ephemeris...For ALL of 2015, the sun's semidiameter never exceeded 16'-16", so I'm still trying to figure out how Carlson's formula has it at 16'-20".
If you want a really good view of all of this, purchase an astronomical almanac for the current year.
I used to buy one every year. They have a lo of good explanations in there. For even more in-depth information, get the
John Hamilton, post: 364041, member: 640 wrote: If you want a really good view of all of this, purchase an astronomical almanac for the current year.
I used to buy one every year. They have a lo of good explanations in there. For even more in-depth information, get the
John:
I did. MICA is the Multi Year Interactive Computer Almanac from 1800 to 2050. (ought to last me for a while):-D
http://aa.usno.navy.mil/software/mica/micainfo.php
So do I believe it, or Carlson?
Mica. It is the gold standard
Solar irradiance, I believe is the term, that causes the apparent size of the sun to be larger than it actually is. This is discussed in the Sokkia handbook. It's an estimate. I have never done this, but I believe the most accurate approach is to average trailing edge and leading edge observations, or use a scope with a solar reticle or use Roelof prism. These would effectively rule out semi diameter.
Shawn Billings, post: 364058, member: 6521 wrote: Solar irradiance, I believe is the term, that causes the apparent size of the sun to be larger than it actually is. This is discussed in the Sokkia handbook. It's an estimate. I have never done this, but I believe the most accurate approach is to average trailing edge and leading edge observations, or use a scope with a solar reticle or use Roelof prism. These would effectively rule out semi diameter.
Don't have a Roelof or a scope with a solar reticle, so no can do. It's not called irradiance (that has to do with the Watts/sf radiation the sun emits or something), but I think what you're referring to is this additional formula in the Sokkia manual:
I can't see where Carlson uses this in their code (I'm not very good at reading C++ or whatever it's written in), but could it be that MICA is NOT using this additional correction? Or that Carlson is not? Can anyone tell me what "h" is in the Sokkia formula? This is all getting way too far down in the weeds for me. I just want to go out and USE the sun for Azimuth. Perhaps they didn't worry about 10" here or there in the day, because their instruments were only good to a minute or so.
Don't know. Would just like to know what data to rely on to reduce my observations.
See page 92 of the pdf from rollanet.org (These are the guys that "wrote the book" so to speak). It's been a few years since I played with the very things you are, so I wasn't sure of the terminology. They refer to it as irradiation of the sun.
http://www.rollanet.org/~eksi/2008EPHEM.pdf
I believe that refraction can also make the sun appear larger, much like the moon appears larger close to the horizon compared to high in the sky. There are enough random errors in the process that the variation in SD is probably not that critical, but you asked. 🙂
You don't necessarily need a Roelof's or a Solar reticle (which is just a circle that is part of your cross hairs that is slightly smaller than the diameter or the Sun, maybe about 15 arc minutes). You can just take a series of leading edge shots and trailing edge shots, keeping track of which is which. The average should be the azimuth as if you used the center. No worries then about how someone estimated the atmospheric effects and optical effects on the apparent size of the Sun.
By the way, with a good ephemeris such as Elgin Knowles and Senne's book data I could easily get within 15 arc seconds using a filter, trailing edge, and a Garmin eTrex for time and position (+/- 30 feet).
Using the internal ephemeris of my old TDS data collector I could get 30 arc seconds. The nice thing about the TDS collector was that I could perform a Sun shot while the rodman moved from the backsight to the first foresight in many instances, just as a quick check on my azimuth. It was pretty slick. No reason with today's data collectors with internal GPS that this couldn't be done in literally seconds with no manual entry by the user, with an accuracy of 15 arc seconds or better. Good luck convincing any of the total station data collection software writers to see the value in that.
h is altitude of the sun.
Shawn Billings, post: 364171, member: 6521 wrote: By the way, with a good ephemeris such as Elgin Knowles and Senne's book data I could easily get within 15 arc seconds using a filter, trailing edge, and a Garmin eTrex for time and position (+/- 30 feet).
Good luck convincing any of the total station data collection software writers to see the value in that.
That's for sure. It's impossible with SurvCE to even just record an azimuth only...no distance. I've tried everything and asked here. No dice. No way to just use the DC as a "note taker" for the shots.
