Friday, March 22, 2013

A Comparision of Ocean Current Models

We have been struggling to learn what might be the best ocean current model for forecasting currents for the OARNW trans Atlantic row, now underway. Our original goal was to measure the currents continually from the vessel and thus get a good check on the forecasts. Unfortunately, although all the other sensors are working fine, we do not have a knotmeter speed, so we cannot directly measure current.

Thus we are left with only very general comparisons during those incidents when the boat position coincides with an unusual current forecast, such as very favorable or very adverse.  At times we have had some indication that this might work,  but for the most part this has been very discouraging exercise.  But the problem is not just the measurement. What we have learned well is the various forecasts available do not agree with each other, so this has become a really hit and miss exericise.

To illustrate this point, we show here the predictions from five sources, which are more or less independent of each other. We also show what the same data looks like when obtained from different sources.

In each of these five cases we have marked two regions to compare the predictions. The top one is about where the boat is at this moment. The bottom one is just marking another area for comparison. But it is easy to focus on other areas for comparison as the maps are all on about the same scale.



Global HYCOM model from via WeatherNet
US Navy Global NCOM from

And we see immediately the problem. These are the latest data from each of these sources. Not only do the speeds vary by huge amounts, even the directions are reversed in some cases. If you see two that agree in the top box on some level, they will much disagree on the bottom one, 180 nmi south.

Another big difference not shown here is the RTOFS data do vary notably from day to day, even hour to hour, within a general pattern, and the OSCAR data are essentially sliding averages over the past 5 days.

In any event, our great excitement about predicting ocean currents underway has to be put on hold for awhile till we learn more.

One thing that is encouraging is that two of the popular models have data available to mariners underway in GRIB format at no charge, that is the RTOFS from ViewFax and the OSCAR data from These are shown below as they arrive from these services, which can be compared to what they look like in the displays above created in Panoply from the raw data.

OSCAR data delivered by email from and displayed in ViewFax

RTOFS data downloaded and displayed from ViewFax

These data do not agree with each other, but they do agree with the primary sources that we show in the top pictures.  We can hope that one of these turns out to be useful.

For the moment, we are leaning toward the OSCAR data for our application, primarily because we have several indications that the RTOFS speed predictions are quite a bit high.... however, I must state immediately, we do not have proper scientific measurements of this.  It is clear that the two models do not agree, that we can see easily, but which one is right or at least best, we do not know.

The indirect data we get is whenever the boat is forced to go onto a sea anchor because of bad seas. Then we get an indirect measure of the current from the COG and SOG as they drift.

Here is another bit of data which favors the OSCAR results, which is the track of the boat over the past couple days viewed in Google Earth.  During the time they were in a steady favorable 0.3 kts according to OSCAR, the boat did indeed make progress in the direction of the current at roughly this much faster than their best rowing speed, ie they had their best 24 hour run of the trip, averaging about 3.0 kts.

Tuesday, March 12, 2013

Course Boxes - Still Valuable in Modern Sailing

On any long voyage, ocean or coastal, or even a long inland sail, it is generally crucial that the helmsman has in mind the proper or desired course at all times. This is especially true on overnight sails and any time in bad conditions. Generally in big waves and strong winds it is important to know the right course that keeps you in tune with the sail trim, wind, and seas.

On a long trip, when the course is changing frequently, it is easy to lose track of the course. For example, if you steered 220 all day yesterday, but now this morning the right course is 245, then if you are thrown off course by a wave, you might instinctively turn back and try to settle in at 220, which would be wrong. The delay in doing this right could be crucial in very tough conditions.

Besides safety, there is efficiency. With changing crew, you might forget to tell the new watch that the course changed, etc... and off you go for some period of time on the wrong course.

The solution to this is to have some prominent place in clear view of the helmsman that shows the present proper course. Whenever the navigator  comes up with a new course, it is marked in that sign, and all know it.  You could rig any device for this.  At times we have used a vertical piece of duct tape, and then just crossed off the old course and wrote in the new one below it.  That has the virtue of keeping a record of your courses if doubt comes up.  You could also just write it on the bulkhead with some type of erasable marker after testing that it does not stain your fiberglass (most do!)  Or better, use a small dedicated board for this purpose.

In the old days of sailing, they used a course box. I found this antique one for sale on ebay in Sept, 2009. This one is about 10 inches across. You lift the pins to rotate to a new digit.  Nice idea, and modern equivalents are easy to imagine.  The existence of these show clearly that this was an issue faced by sailors since earliest times.

Sample of a course box seen on ebay. This one was from a Russian ship. The numbers on both of these are about 2 inches tall.

This one from a friend's collection is the same as I have seen in museums. This one is older, dating back to maybe late 1800s or early 1900s.

Note: in earlier writings we have referred to this as a "route box," but it seems course box is a better term. I do not know what the official name was or is. As we run across this in our work we are changing all to course box.

Thursday, March 7, 2013

More mainstream media on Viking sunstones

Believe it or not, there is another rash of mainstream media articles on sunstones that do not mention the work of Leif Karlsen and his book  

Therefore I am adding another link to the front of our blog that goes back to the post we made last year when the same thing occurred.

This was originally posted here in August last year.

Leif's book came out in early 2003, and he reported then the crystal find on an  Elizabethan wreck that took place while his book was in final production phase in 2002.  That was 11 years ago.  Yet now all the media refer to this as a recent find, with an exciting new interpretation.  All the stories are the same, so someone fed them the story and they repeated it, unchecked.

They even speculate now that it actually could be used for navigation! They really should read Leif's book, especially the sections on his own use of the sunstone navigating Viking replica vessels along known Viking routes.

... and they should get a lesson on the use of Internet search engines, or failing that, maybe call a library and ask for help from a librarian.

Maybe title of the book was not clear enough?

Saturday, March 2, 2013

Finding the North Star

We have had a request to discuss finding Polaris from Facebook friends in India, and we are happy to respond.

The subject is covered in much detail in our  book Emergency Navigation, which covers not just the North Star, but how to find directions from any stars in the sky.  The North Star is just a bit easier.

First and foremost, it will be due north, at a height above the horizon equal to your latitude. Another general way to think of it is it will be a faint star to the north, about halfway between Cassiopeia and the Big Dipper, but often this entire space is obscured by haze, which leads to the value of having many ways to find out where the North Star is, even when we cannot see it.

Here is one method based on Cassiopeia. These are all from Chapter 5 of the that text.

Figure 5-6. Finding Polaris from Cassiopeia. Any view of Cassiopeia alone will tell where north is. Find the nearby region in the sky from which the constellation would look like an M. Imagine a line along the base of the M. Twice that distance from the base at the trailing star of the M is the location of Polaris, even if the star happens to be obscured by clouds. Th e squares in the corners are reminders that this distance should be projected at right angles to the base. Directly below that point on the horizon is due north.

The next picture covers all those cases where the pointer distance is about 5 times the pointer length.

 Figure 5-9. Ways to find Polaris. Any one of these widely spaced constellations will locate north. The + or – signs indicate that the polar distances are either somewhat more than or less than exactly five times the pointer spacings. However, it is simpler and adequate to just remember the factor of five—you point with your fi nger, each hand has five fingers. Stars at the bows of the constellations are the leading stars; those at the sterns are the trailing stars.

I hope that helps.  Please see the book for a lot more on steering by the stars in the North and South Hemispheres.

The North Star is fairly faint (magnitude 2) but any of these pointers is just as good as the north star for finding the place where the North Star is, even when covered by clouds, as illustrated in the picture below: