After a very difficult last 10 miles, Jacob has safely arrived at Trinity Beach, Yorkey's Knob marina according to local newspapers. We do not have any word from him to confirm any details. Readers must buy a subscription to read the articles... no teasers; no samples! Once the newspaper gets the story onto their facebook page there might be a link to read it. A historic voyage completed, with many records under his belt. I thought he might have anchored last night, but it seems he did not... just powered on through.
Update, the next morning: He did not anchor, as speculated here, but indeed went straight through to land in Trinity Beach (Yorkeys Knob Marina), north part of Cairns.
We can see on the Google Earth presentation of Jacob's track that the reported GPS positions are more scattered over a larger range than when he is anchored than when he is underway. You can see this in the Google Earth presentation online. Zoom into the track before he anchored at Flynn Reef. Below is a sketch of the situation.
I apologize for all of these rough sketches in this note. This is a passing thought (and timely) that does not justify, and i hope not require, finer artwork.
The yellow range shows the position reports from an anchored position. They spread over about 4.5 miles range. GPS has an inherent accuracy of about 100m or better. The pink positions are his reports from the same system once underway. In both cases, our presentation of each hourly position report is a sliding average over the past 3 hr.
Here is a plausibility argument of why the position scatter might be larger when anchored.... and indeed an argument that he might be anchored once again (03z Jun 8).
The position accuracies are poor because the GPS antenna is below decks with a limited view of the horizon. With limited view, it sees only a few satellites, and the intersection angles are narrow, which introduces large fix uncertainties.
Here schematically we see LOP1 intersecting with LOP2, and then consider a small error in LOP2 called 2'. We see the resulting error in intersection is large, whereas with a steeper intersection (ie nearer 90ยบ intersection) the same small error in one of them has less effect on the fix. And on top of that, a 2-d GPS fix (ie elevation at sea level is known) actually requires 3 LOPS. The third one is key to how GPS works; it is used to remove the time uncertainty in the signals—the "LOPs" are actually range rings measured by how long it takes the signal to get to the GPS antenna from the satellite.
Here is a schematic of the restricted view of the sky from within the vessel...
The outer circle is the sky; the areas on the sky are what is seen through two port light windows on either side of the vessel—I do not know the real geometry of this on Jacob's boat. When the boat rolls, one side sees higher, the other looks into the water. When the boat yawls, one window looks more forward; the other more aft, and so on.
When anchored the boat does move around at the end of the anchor line, but if the wind is strong, it is likely to be held more or less in line, and if it does sail back and forth some in the wind, it will be a relatively slow change.
It seems the view of the sky will change with time when anchored, but that it will be much slower and less dramatic than when sailing (rowing).
When underway, this theory is that the two portlight views sample much more of the sky during a fix than they do at anchor, which gives the GPS access to more satellites and hence better fixes than when anchored with limited view.
This may or may not be true. There may be other factors at play. I am standing by to hear your thoughts on this. Post a comment if you have another interpretation of any support for this idea.
For now, and for the moment, I will use this argument in a video below that he has anchored again, near Cairns anchorage CA1 or CA4. I say that because his track progressed normally from the last anchorage to this location and then just stopped and scattered.
The alternative is he is fighting against the wind and not making progress, which is effectively like anchoring, when it comes to sampling the sky through limited access.
I have not heard from his since he left Flynn Reef, so i do not know the answer.
Here is a video discussion of the points above and others related to his current position:
Jacob got started later start than anticipated, but he did leave and indeed got by the two reefs he needed to pass. Now on the entrance range to Cairns, about 9 nmi off. We shall see how he proceeds in this last short leg. It will be light in about 2 hours.
Just got word from Jacob that he is getting underway shortly. We can watch him on Google earth; download the KML file above and drag to Google Earth. This is likely the most exciting leg of the entire year-long voyage.... this last 20 nmi getting across the Great Barrier Reef!
* * * NOTE: our GE positions are 3-hr averages, so we will not get a good track until 3 hr after he leaves. At 00z he said "in a few hours," maybe 9z or so we might see the track.
Safely anchored off of Flynn Reef, Jacob is still waiting for lighter wind.
We discuss in the video below the situation, including charting anomalies, and glimmers of faint hope in the forecast for tomorrow... which is just about 10 hr from now.
One on the things we think we are learning is the Oceanic version of the National Blend of Models might provide best extended forecast. We have one more test of this underway that is Jacob oriented, but then we can do the same thing across the globe by watching conditions at several buoys that have live data to compare with. We are homing in on the best ways to present this. The best source of data is clear, namely LuckGrib, because it is the only source and viewer for the probabilistic data from NBM and the statistical data from the GFS Ensemble runs, GEFS.
We got word yesterday that the wind was relentless at 170 to 180T at 30 kts gusting to 35. This was simply unmanageable so he looked for first opportunity to anchor and wait it out. He found a place just on the west side of Flynn Reef, and we made a new chart insert to show this on the google earth. Please update the KML to see it.
As luck would have it, after his anchor caught (a 16# Danforth, with 200 ft of line) and he played out the line, he noticed a mooring buoy right there so he is tied to it and his anchor...which is good news as the winds are still strong.
It's hard to see in this pic, but he says he can see the green water ahead where the reef is located. ("Green water" and green reef on the chart is a coincidence of course. The latter simply means that at zero tide height the reef is exposed... depending on the waves.) We also see here his two radar reflectors; the antenna on the top is maybe his AIS transceiver to monitor traffic and be monitored. The white instrument looks like a sonic anemometer. The boat is moored at the bow.
I thought his discovery of a mooring buoy was really unusual, but these are what the dive boats hang onto while diving the reef. He is just off the reef to the west, both anchored and moored.
You will also notice on the Google Earth display a huge spread in the GPS fixes. This is not normal! This comes about because his GPS receiver antenna figuring the tracking positions is on Iridium Go box and that is below decks, in fact even worse, under a solar panel. The GPS antenna can barely see the horizon, which limits the view to only a few intermittent satellites, maybe even in the same direction, and hence to poor fixes. In this case really poor fixes. If we had this GPS in hand, it would be warning us of this; the tracker code does not do that.
Users of the popular Iridium Go should be aware that the GPS antenna used for the tracking signals is in the stubby antenna connected to the box. Even if you have a very nice external antenna that enhances the satellite connections this does not improve the GPS. There is no GPS receiver in that external antenna.
Below is our system, which is a sliding average over the past 3 hrs of PredictWind positions.
True position is the red and yellow dot. The red line is 4.5 nmi long, which shows a huge range in GPS positions, all artificial. A GPS with good antenna would be well within a 100 meters or so. Dotted circles are the averages of past 3h reported positions from PredictWind tracker, shown below
Blue dot is the true position. In this static collection of data points, it is not clear if the Starpath averaging is improving the display, but underway it is definitely better which can be seen by comparing our GE display with the PW display.
We will watch the weather to see when he might get underway again. ASCAT passes show strong will still, and models do not show relief this week. See how we evaluate the forecasts at this note on crucial weather routing. Our conclusion is that the FV3 GFS model did the best for the longer term (~96h) forecasts. The Oceanic National Blend of Models was the best for short term (day or two). We can see the tides in that region from within OpenCPN, where we have the AU charts installed. The tide data come with OpenCPN. We see below that the tide in this region has a range at the moment of about 2m or about 6.5 ft. That seems to imply that he might see the reef dry briefly if the waves don't interfere.
A high pressure system [1039 hPa], moving slowly eastwards over the Great Australian Bight, will extend a firm ridge over northern Queensland waters for the next few days with fresh to strong southeasterly winds.
Forecast for Tuesday until midnight
Strong Wind Warning for Tuesday for Cairns Coast
Winds
South to southeasterly 20 to 25 knots, reaching up to 30 knots in the late morning and afternoon.
Seas
1 to 1.5 metres, increasing to 2 to 3 metres outside the reef.
Swell
Easterly 1 to 1.5 metres outside the reef, decreasing to around 1 metre later in the evening.
Weather
Partly cloudy.
Wednesday 5 June
Strong Wind Warning for Wednesday for Cairns Coast
Winds
Southeasterly 20 to 25 knots, reaching up to 30 knots inshore north of Cape Grafton in the afternoon.
Seas
1 to 1.5 metres, increasing to 1.5 to 2.5 metres outside the reef.
Swell
Easterly 1 to 1.5 metres outside the reef, tending southeasterly 1 to 1.5 metres during the morning.
Weather
Partly cloudy.
Thursday 6 June
Winds
Southeasterly 20 to 25 knots.
Seas
1 to 1.5 metres, increasing to 1.5 to 2.5 metres outside the reef.
Swell
East to southeasterly 1 to 1.5 metres outside the reef.
Jacob is now anchored off of Flynn Reef. Southerly winds of 30 kts gusting to 35 were too much to cover come. He missed first attempt to grab an anchorage, but caught the next one and is anchored now waiting for new wind... looks like he may not see that till Thursday.
New KML on line now with new charts and routes June 2 1524 PDT
For those who might be following this blog post, now is the time to turn on Google earth for hourly updates. Jacob has turned to enter the region of the reefs so this is the last crucial step of the voyage. Positions online are updated hourly, but they are the average of the reported positions over the past 3 hours... we have to do that because the raw GPS data from the vessel is uncertain for reasons we explained earlier. Also keep in mind it is pitch dark (no moon) where he is at the moment, so he is "flying by instruments," which he is well accustomed to as a former combat pilot. PS... he did not choose to enter in the dark! In these strong winds without a sea anchor or drogue, he cannot slow down. He is there, when he is there.
This is 12:11 PM Seattle time, which is 0511 in AU, which we see here is still dark. It is winter there. The meridian lines shown on this map are 2 hr a[art, so he has about 2 hr more of darkness at this time. This screen cap is from the free PC program called orbitron, which i am running on a Mac using WINE. I will post a couple articles about this program in our Starpath nav blog. We use it to predict passing times of Metop-A and Coriolis, the scatterometer satellites.... the daylight terminator is a bonus to us, but crucial to those who want to visually see the satellites cross the sky. A video note on this announcement:
We are updating frequently now, so best to download the KML file and update your Google Earth.
Jacob is making good about 2.3 kts, on course to Flora Pass, which at 18z on Sat Jun 1 is just 62 nmi off. He is getting some help from the currents, and wind is strong and right on the stern. We see this in the ASCAT images, but strangely the WindSAT data do not agree. This we will look into elsewhere. (We also have to consider that Jacob has indeed seen periods of wind at 10 kts from 070, so maybe there is some chance the WindSat captured this unusual transient wind.)
We note that the local coastal forecast from Cairns is still 20 to 25 kts, but we have at least one example where this was an overestimate. For Jacob and his boat, 20 kts is difficult to steer in, but 25 kts is notably more difficult, with much limited range of headings.
Below is a video discussion of these points and others...
