NavList:

>>One of the problems with statements on Nav List is that people don’t give sufficient information for the statements to be analysed properly by others.  Was this day or night?  Was it a fix or a position line? I’m assuming it was a noon fix.  What was GPS being used for?  Was it to provide a source of time or to provide a position to compare the astro to? <<

Good questions, David.  And before you, by Frank, wondering how we could TELL if the timepieces were wandering.

First, what was the GPS being used for?  It was the gold standard for location data. 

As I practiced with my sextant, and in particular seeking to develop my skill in taking a sight just as we hit the crest of a 14 foot wave, I evaluated my skill in terms of how close I could get an LOP to come to our GPS fix. 

As for time, generally speaking the GPS display and my Casio wrist watch agreed with each other within 1/2 second or so.  I used my watch (which was NOT syncing with WWVB at the time) freely in taking sights, correcting my time by its typical rate of change, and then glancing at the GPS time display - which typically agreed with my watch.

Second, did the time discrepancy I observed (or would this discrepancy) make any practical difference navigationally?  No.

     My celestial LOPs agreed with reported-GPS
     locations within a range of from one to six 
     miles.  I was thrilled even to get within six miles.   
     
     As with many of the things that motivate 
     me around celestial nav, my question was 
     driven by curiosity about the way things work
     more than any practical considerations.

Third, how did I recognize a discrepancy?

     The captain and I both had timepieces we 
     had rated.  On Day 1, after correcting for 
     known rates of gain/loss, his watch, my 
     watch, and the display on the GPS receiver 
     all agreed with each other.

     On Day 2, the watches agreed with each 
     other, but the display on the GPS was fast.

     On Day 3, the watches and the GPS all 
     agreed with each other.

     This sequence happened twice over the 
     course of 19 days.

The discrepancy was, as I recall, two seconds in one case and five seconds in the other.

Am I suggesting that the atomic clocks on the GPS satellites are drifting, or that they are transmitting flawed signals?  No.

What I am wondering is if GPS receivers are found by most people to be accurate in calculating/displaying the correct time.  All receivers get time-related information in the signals they receive from the satellites.  It would be up to the software to slice and dice that information to come up with a time to display.  Alas, after all these months have passed, I have a zero memory of who the GPS unit in question was manufactured by.  I was mostly focused on the trip itself as it happened.  It was only after I was well and truly back on shore that I started to reflect on my experiences...and among them, a small puzzle related to time.

This discussion has helped me refine my question.  So let me restate it:  Has anybody else seen behavior in a GPS's time display that made them raise their eyebrows?  Has anybody else seen their GPS time display showing odd behavior?

On this trip, we had three relevant time pieces.  Two of them were rated watches that always (after correctcions were applied, to within a fraction of a second) agreed with each other.  One timepiece, the GPS, had two occasions where, over the course of a full day, it disagreed with the watches...but for the balance of the trip DID agree with the watches.  That is, the GPS time display would look odd for a day, but the next day seemed to straighten itself out.

If people's experience of GPS receivers is that they always see an accurate time, then I will start to ponder if there were environmental effects that might have caused the wristwatches to drift, more or less in sync, with each other.  Perhaps a quartz watch can vary its rate in response to temperature changes, or in response to vibration.

Alternatively, perhaps GPS receivers (this one was wired into the navigation station; it was not a hand held unit) can be affected by environmental or electrical variations.

The electrical status of the boat was that we ran our diesels for 3 or 4 hours/day to keep the batteries charged, and the rest of the time we sailed and ran our electrics off batteries.  There was once that we were becalmed as the North Pacific High shifted its location, and motored for 20 hours trying to get back to some wind.  As I say, I failed to correlate these events with the time discrepancies...mostly because I was puzzled at the level of raising my eyebrows at the time displays available to me.  I had not yet gone so far as to formulate a concrete question...still less, to formulate a hypothesis that I could test.

Our sailing conditions ranged from 16 foot waves and 25 knot winds to flat calm. 

Our temperatures ranged from tropical too-hot-to-sleep heat to North Pacific wear-two-pairs-of-long-underwear chill.  I did not think to correlate any of these environmental changes with the wristwatch sync puzzlements.  Towards the end, I was focused mostly on staying warm in the steady drizzle of rain rather than on timepiece issues. 

I had always assumed the display on the GPS unit was the gold standard.  My questions only began to arise after the skipper came by one day and asked, "Are you noticing anything odd about the time on the GPS unit?"

Thank you all for the questions and thoughts you have shared.

Bob