Celestial Navigation Plugin Redux

[rgleason]

NAV your help is appreciated.

[seandepagnier]

Originally Posted by NAV

Three more observations:

1. In the plugin, eye height (h) can only be entered in whole meters (m). For more accuracy and easier checking against example data in The Nautical Almanac and The Admiralty Manual of Navigation, I suggest to offer the possibility to enter eye height in meters with one decimal (m.m).

Ok, but I must use a text like the index error, not a spinbox

2. Dip in the plugin is calculated by the formula 1'.753*SQRT(h). Both the Admiralty Manual of Navigation and the Nautical Almanac use 1'.758*SQRT (h). I suggest to use this last formula.

ok

3. The Nautical Almanac uses a standard atmosphere for the calculation of refraction (Temperature=10 degrees Celsius and Pressure=1010 hPa). I suggest to set the default values in the plugin accordingly.

1010 is probably closer, but 10C seems a little cold for an average temperature. I think maybe because they are english or something, in any case I will use 10C because the user can change it.

Originally Posted by NAV

Well, things look fine, I guess. But checking goes slow when example data cannot (yet?) be entered with same accuracy and when - partly - slightly different formulas are used. A difference of 1' in altitude is relatively large in astronomical positioning, so while checking I try to establish the cause of differences > 0.1' I see (that's 0.00167 degrees).

This brings me to a fourth suggestion:
Although calculations are probably done to full precision, for the presentation of data 3, 4 or 5 decimal figures are used. I suggest to make that 4 decimal figures for all values in degrees. For presentation of minutes (angular) 1 decimal should be sufficient.

Yes, full precision is used, but not displayed.
ok.

Question for Sean: will the POT-file be updated too?

yes

all done, in git now

[NAV]

Originally Posted by boat_alexandra

I think maybe because they are english or something, ...

The Nautical Almanac is prepared jointly by UK en US ;-)

[NAV]

From the code I understand that the Bennett formula is not used anymore for the calculation of refraction:

/* compensate for refraction */
double RefractionCorrection;
#if 0
/* old correction not used */ THIS IS BENNETT's FORMULA
double Ha = m_Measurement - m_EyeHeightCorrection;
double Ref = 1/tan(d_to_r(Ha + (7.31/(Ha + 4.4))));
double RefImp = Ref - .06 * sin(d_to_r(14.7*Ref + 13));

RefractionCorrection = RefImp * .00467 * m_Pressure / (273.15 + m_Temperature);



Refraction seems to be calculated only by the following code:

#else
double x = tan(M_PI/180 * ApparentAltitude + 4.848e-2*(M_PI/180) / (tan(M_PI/180 * ApparentAltitude) + .028)); m_CalcStr+=wxString::Format(_("\nRefraction Correction\n\
x = tan(Pi/180*ApparentAltitude + 4.848e-2*(Pi/180) / (tan(Pi/180*ApparentAltitude) + .028))\n\
x = tan(Pi/180*%.3f + 4.848e-2*(Pi/180) / (tan(Pi/180*%.3f) + .028))\n\
x = %.3f\n"), ApparentAltitude, ApparentAltitude, x);
RefractionCorrection = .267 * m_Pressure / (x*(m_Temperature + 273.15)) / 60.0;
m_CalcStr+=wxString::Format(_("\


I can't relate this to any formula I've found, although a small part of it seems to be similar. In the Admiralty Manual the following procedure is described:

"Calculate refraction (R) at a standard temperature of 10 degrees Celsius and pressure of 1010 mb:
R0 = 0.0167 degrees/TAN(Ha+7.31/(Ha=4.4))
If temperature T in Celsius and pressure P in mb are known, calculate the refraction from:
R = f*R0 where f = 0.28P/(T+273)
otherwise set R=R0 "

Sean, can you please advice me what formula you are using and why?

[NAV]

[QUOTE=NAV;1469356]

"Calculate refraction (R) at a standard temperature of 10 degrees Celsius and pressure of 1010 mb:
R0 = 0.0167 degrees/TAN(Ha+7.31/(Ha=4.4))
If temperature T in Celsius and pressure P in mb are known, calculate the refraction from:
R = f*R0 where f = 0.28P/(T+273)
otherwise set R=R0 "

QUOTE]

In the Nautical Almanac of 2005 the bold formula above (read '+' for the equal sign after the second Ha) is slightly different:
R0 = 0.0167 degrees/TAN(Ha+7.32/(Ha+4.32))

[rgleason]

I had a question about setting time...Should we be using Greenwich time? If so, should it say so in the instructions? --Or does it use the computer''s clock to start with and then calculate Greenwich mean time?
I am a little unclear how you would use a quartz watch to adjust...to get the actual time.

Probably a dumb question, but it came up because I had My ship Location set and tried entering sights and was getting circles halfway around the world. Does Celestial use My Ship at all for setting the likely position?

[rgleason]

I forgot, I've got to upload Sean's changes to improve the interface, which compiled and works.

Hakan and I are working on batch file installers, so we can get the data directories and files installed properly. See OpenCPN Internationalization ( i18n ) - Page 48 - Cruisers & Sailing Forums where Hakan made a batch file for plugins language files. I've asked if he could add the data files too.

I am going to try to get celestial working with a batch file.

Here's the dll. Take off the DOC.

[NAV]

Originally Posted by rgleason

I had a question about setting time...Should we be using Greenwich time? If so, should it say so in the instructions? --Or does it use the computer''s clock to start with and then calculate Greenwich mean time?

Yes, GMT (or more precisely UT) should be used for astronomical calculations. An indication at the field where time can be entered should be useful.

On board you maintain ship time (a local time or a zone time). The time you read from your clock (or watch, or computer) for the time of altitude measurement, should be corrected for clock or watch error and for the difference between ship time and UT.

[NAV]

While checking the calculation of altitude corrections (for Sun LL, Sun UL, Moon LL, Moon UL, Venus and Polaris) I do find differences >0.1' for Semi Diameter, Refraction, Horizontal Parallax and Parallax in Altitude. This must be caused by the formulas used, which differ from those used in The Admiralty Manual (and in the Nautical Almanac) for the example data.

E.g. for the Moon LL example I find HP=0.9475 and PA=-0.7890, while the example data give HP=0.9100 and PA=-0.7598 (differences of 2'.25 and 1'.75).

[rgleason]

From Information in Plugin Taking a Sight:

Date and Time for the sight must be the same as the exact moment the measurement is taken. If the sight was taken a while ago, and we have moved since then, do not modify the date and time values. Instead, enter course dead reckoning for distance traveled and bearing to shift it into position to use along with current sights. Once a sight is shifted, it is no longer be used for a position fix (it is very complicated to implement this math) but a fix can still be visually determined.

PItfalls:

Time The time is syncronized with the system clock when a new sight is selected. Be sure to manually adjust the time to make sure it corresponds to the exact moment the sight was taken.

=====

Rgleason

I had a question about setting time...Should we be using Greenwich time? If so, should it say so in the instructions? --Or does it use the computer''s clock to start with and then calculate Greenwich mean time?

BCN

Yes, GMT (or more precisely UT) should be used for astronomical calculations. An indication at the field where time can be entered should be useful.

On board you maintain ship time (a local time or a zone time). The time you read from your clock (or watch, or computer) for the time of altitude measurement, should be corrected for clock or watch error and for the difference between ship time and UT

You use the "Ship's Time" (local time or a zone time) to:
1. Correct for clock or watch error.
2. Get the Difference between ship time and UT for Greenwich Mean Time GMT.

What do you do with these?
Are you trying to enter/correct for the Actual UT /GMT of the sighting?
I am still guessing that you do not want your Local time entered for the sighting.

[rgleason]

Sean would you mind checking how something works?
Since I am not taking sights, for a new sight after putting in my approx position at the top, I pick NEW, then select a star and hit FIND. The FIND popup shows a lat and long that never change, but the last one Longitude is never correct, so I change it to my approximate location and enter the altitude shown in the previous popup. Of course it is just approximate so the fix is wrong, but it is closer than what pops up initially.

Where do the LAT LONG values come from for the FIND popup? I don't think they are right!

Thanks

[rgleason]

Will users want to save sights?
Or is it best to force them to record them in a book?

[NAV]

Originally Posted by rgleason

You use the "Ship's Time" (local time or a zone time) to:
1. Correct for clock or watch error.
2. Get the Difference between ship time and UT for Greenwich Mean Time GMT.

What do you do with these?
Are you trying to enter/correct for the Actual UT /GMT of the sighting?
I am still guessing that you do not want your Local time entered for the sighting.

The 'problem' is this:
1. For the calculation UT of altitude measurement must be used for the calculation (this UT must be entered in the time entry fields).
2. You read the time of altitude measurement from a clock (or from watch or from a computer). In almost all situations this time is not equal to UT, so it must be corrected before you can input the UT in de time fields.

The quote you give from the Information page, refers to a change of position (distance, course) between your first observation and a second one. Usually this applies when taking a sun sight in the morning and a sun sight at noon. For both sights the correct UT of both measurement should be used for the calculation. The resulting LOP of the first sight can be moved to the second sight LOP by moving it over a distance equal to the time difference between the sights times the speed, to obtain a fix of two LOPs at the time of the second measurement. The plugin doesn't use LOP's though, but COPS, for which this cannot be done as I understand from the explanation. Of course the accuracy of a fix based on one shifted LOP and another one, is less than a fix of multiple star LOP's taken at about the same time (say within a period of about 10 minutes) in morning or evening twilight.

[NAV]

Originally Posted by rgleason

Will users want to save sights?
Or is it best to force them to record them in a book?

Yes, all users want to save sights, I expect. And if they don't, they should in my opinion. So, preferably this functionality should be build into the plugin. As long as this functionality is not available, use a notebook for sights only.

[NAV]

Originally Posted by rgleason

Does Celestial use My Ship at all for setting the likely position?

It doesn't, as far as I can see.

Another suggestion: for the Find page it would be easier to offer the possibility to enter your estimated dead reckoning position in degrees and minutes. You don't have a My Ship position when you must use astronomical positioning, because other systems are not available.