GNSS Jump and a Poorly Conditioned Star Fix
At 0430 UTC on 14 May, in position by DR 34°18′N 023°42′E, your GPS receiver shows 34°22.6′N 023°36.1′E - a 5.2 NM jump within 12 minutes - and the receiver's status page shows only 4 satellites with two flagged unhealthy. There is no other GNSS. You observe three stars: Vega altitude 34°12′.4, azimuth 052°; Arcturus altitude 28°05′.1, azimuth 145°; and Spica altitude 11°22′.8, azimuth 198°. The intercepts, plotted from the DR, are 0.6 NM toward for Vega, 0.8 NM toward for Arcturus, and 2.9 NM away for Spica; the cocked hat is 2.6 NM across. The horizon is clear, but there is a 2′ dip uncertainty from a residual swell. Dawn is at 0505, and the sky will be too bright for star sights by 0515. You have a radar range and bearing to a lighthouse charted at 34°25′.0N 023°30′.0E, but the lighthouse is not yet visually identified.
1. Assess the reliability of the three star sights and state the most probable fix. Explain the geometry and likely error sources.
Read the worked answer
The weak sight is Spica: altitude 11° is low, so refraction, dip and horizon error are amplified; the stated 2′ dip uncertainty alone is about 2 NM at that altitude. Its azimuth (198°) is also only about 53° from Arcturus, so it gives a poor crossing angle. Vega (34°) and Arcturus (28°) are higher and their azimuths are about 93° apart, which is a reasonable cut. Their intercepts, 0.6 NM toward and 0.8 NM toward, are consistent with a fix about 0.7 NM from the DR in an easterly direction, not with the GPS position 5.2 NM away. The 2.6 NM cocked hat is therefore best explained by Spica's error, not by a real 2.6 NM uncertainty in all directions.
I would reject the GPS jump as the primary fix: a 5.2 NM change in 12 minutes is not physically credible for this vessel, and the receiver has only four satellites with two flagged unhealthy. I would plot the Vega/Arcturus fix, mark it as a lower-confidence celestial fix, and use the radar lighthouse as an independent check. If the radar confirms the celestial fix within about 0.5 NM, I would plot the terrestrial fix as the primary position and retain the celestial fix for comparison.
2. How would you use the radar lighthouse to confirm the fix, and what checks would you make before plotting it?
Read the worked answer
Identify the lighthouse echo by its charted position, expected range and bearing, and any radar-conspicuous feature or racon. Use the VRM and EBL on a short range scale; check that the heading marker is aligned and apply the gyro error. A one-point fix is not enough for confidence, so I would also use a second radar echo (headland, breakwater, or another charted object) and compare the plotted position with the celestial fix and the DR. Before plotting, check the chart datum/geodetic shift, radar range accuracy, and whether the echo could be a similar structure or a vessel. If the lighthouse is not visually identified, I would not treat the echo as confirmed until its range/bearing agrees with the chart and with the celestial fix.
3. State the actions and log entries after rejecting the GPS position.
Read the worked answer
Inform the master and switch the primary position source to the celestial/terrestrial fix; plot the fix on the paper chart and ECDIS, label it as an estimated position if there is any doubt, and keep the GPS displayed for comparison but not for navigation. Increase the frequency of fixes, use the echo sounder, and check the vessel's position relative to the TSS/lane boundaries and any hazards. Log the time, GPS coordinates, satellite status, star sights, intercepts, radar ranges/bearings, the decision to reject the GPS, and who was informed. Report the GNSS anomaly to the company and, if appropriate, to the coastal authority/VTS; do not send a distress alert unless the position uncertainty creates an immediate danger.