Sunday, December 2, 2012

Welcome and Voyage Overview

We will use this forum to discuss progress and related matters of navigation and weather for the upcoming OAR Northwest ocean rowing expedition from Dakar to Miami, across the tropical  Atlantic. It will be the longest non-stop, self-supporting ocean rowing venture on record. It is scheduled to depart around the second week of January, 2013 and be underway for about 60 days. The last of the team departs Seattle for Dakar on Dec. 4. The boat is on a ship on the way there, expected to arrive about Jan 4, after several stops.

When we get more specifics on the boat transport, we will post them here. (The first post of this shipment was not right. Storm Sandy had changed scheduling out of NY that not all agents knew about.)

We will address wind, waves, and ocean currents, using the state of the art information.  The boat will be transmitting back the actual data observed, along with much other scientific data, so we can learn how well these modern resources work in practice.

You are invited to post questions and comments here, and we will address them as best we can. It promises to be an exciting event, with much to be learned about ocean rowing navigation, much of which will apply to any vessel transiting the tropics.


Nav Notes, Dakar to Miami

Departure: 14º 38.6'N, 17º 26.0'W

Destination: 25º 45.5'N  080 07.4'W

Distance rhumbline (RL) = 3606.03 nmi on a constant heading of 280.7 T
Distance great circle (GC) = 3561.55, with an initial heading of 291.5 T

At an average speed made good (SMG) of 2.5 kts, this is 60 days. This assumes an average distance per day of 60 nmi.

Currents are variable along the route, but usually less than 0.8. If we could find currents that average a favorable 0.5 kts, then we save about 10 days!  We have notes in this blog on current eddies that will have a dramatic effect on progress, and it will be part of the study to learn about them.

Thus is is difficult to imagine the trip taking less than 50 days under very best conditions.  Over the last several ocean rows, the average speed was 2.5 kts. These trips included times of much more favorable currents, but also times holed up with no speed at all. It is felt that 2.5 kts is a good planning speed for such a trip.  That is, just a bit slower than an average walking speed, so this shows how long it would take to walk to Africa from Miami… which in one sense is not that long. It is a small world after all.

With that said, there are records kept of all ocean rows of all 4-man crews, and those that are mostly in the trades like this tend to average a bit higher, maybe up to 3.0 kts.  For now we are still using 2.5 for our DR (dead reckoning). If we learn different, we will update this.

At present there does not seem to be any route that would favor a direct one (GC or RL) whenever possible. Things that might change this are notably stronger trades to the south of the RL, or the development of anomalous currents.

The Cape Verde (CV)  island group is about 300 nmi from Dakar, right on the route to Miami. The Great Circle route just skims the top of the Cape Verdes. The Rhumbline goes right though 70-mile wide channel over the top of the southern most island. The main first tactical issue (Dakar to CV), however, will be to deal with expected winds that are rather well north of NE. Thus there will likely be a struggle at first just to stay up as close to the GC route as possible.  The wind frees up (goes more NE to E) as they get past CV.

The chart below shows the GC route (blue line) along with the major GC shipping routes (red lines) where they might expect to see some traffic.

The green line is about halfway across, where they will be about 1,000 nmi from land in all directions, and on the way home past that.

Screen cap  from Furuno MaxSea echart program. Furuno is a supporter of the expedition and we are grateful for their valuable contributions in charting and electronics.

We will post a gpx file of these routes as soon as they are tested.... better yet, we will actually set better boundaries on likelyhood of ship encounter based on satellite records of actual ship paths integrated over many years, as shown below.

Friday, November 30, 2012

Irridium Sat Phone on JHR

We post this note here so the crew has a place to check for details that can be easily accessed during the testing and further installations on the boat.  

As of May, 2013, we have redone the main body of this article which can now be found at A few nuts and bolts of sat phone usage.  Please refer to that for the details that were once in part here.

The row boat JRH has several means of high seas communications, one of which is a hand held Iridium 9555 satellite phone (User's Manual is online). This unit can be used for email, downloading weather data, or phone calls, and that is in fact the easiest use of the device.  Turn it on, pull out the antenna, and rotate it so it will be in the vertical position when talking, get a clear view of the sky so you see bars showing (just as with a cell phone) and dial the call: 001 area code number.  001 is for calls from ship to US. Other calls take different prefix.

Once connected the communications are excellent. Better than many cell phones, and almost always better than using the HF radio.

This model 9555 phone the boat has is not the latest model from Iridium, but the one just before latest, which is still popular and is the one often used in rental programs. Sat phones rend for about $200 per month, but we are fortunate to have this rental fee donated to the expedition by our friends at www.ocens.com here is Seattle.


Ocens has also donated valuable software including Ocens Mail (for efficient sat phone email and data transfer), WeatherNet (for downloading weather files underway), and Grib Explorer (state of the art GRIB viewer for weather and ocean data). They have provided a valuable contribution to the project and we are grateful for it. I have used these products for many years and can attest to their dependability and great service for navigation and communications underway.

A very nice feature of the Iridium program is the ability to send short (160 characters) messages to the phone at no charge to sender or receiver.  There will be numerous ways for followers and supporters to contact the boat and these will be announced on the OAR Northwest website.

For data connections, the computer is connected to the phone with a mini to normal USB cable. This cable can be up to 10 ft long. The boat has two cables, one 3 ft and and extension of about 6 ft.

Which brings up an important detail of data transfer. When you install the drivers in the PC that link the phone to the PC software, it associates a specific USB plug on the computer with a com port, and that com port is then registered with the software. Thus you must always be sure to plug the phone into the same USB port you used when registering the device. JRH has on board two identical Panasonic Tough Books (thanks to Panasonic for that donation). They are called Thing-1 and Thing-2 (The names might have come about as they are quite imposing devices, that can be dropped 15 feet and still function. You would expect to see them on the fender of a army tank, rather than in a row boat, but we can be certain these are rugged!) Each accessory plug is in a sealed bay.

Thus we  have this reminder:  when using Thing-1 to send sat phone mail or wx data, the phone must plug into the 2nd USB port on the right side. This is the third bay from the back.

If the cable is not plugged into the right port you will get error msg 633 or 634, but it does not say wrong port. You will get the same error msg if you end up with an unconnected plug. It could just say you are not configured properly, or something like that.

Normally we would expect the phone to work inside the cabin and not be influenced by the fiberglass overhead. But with the top of the cabin covered with solar panels, this is not at all clear. They will have to test this once they get the phone on the boat.  And so, too, this brings up more nuts and bolts. If it is raining or rough conditions, one has the challenge of phone use keeping it dry. The phone can be put into a well sealed plastic bag and used outdoors in the rain. I have done this myself numerous times. But for data transfer we need the cable to the PC, and if conditions are rough we also have to have the cabin secured. I am not sure if there is an access port to the cabin that might serve to lead the cable through. 

So it could be there are times the data connections will not be possible due to weather.  If we are lucky, we might find a position inside the cabin that gets good signals. The advantage of being in the middle of the ocean (as opposed to at home with buildings and hills around) is you do have a full horizon to see satellites.

As for transfer efficiency, we have to accept that the transfer rate is not high. Nothing like we are used to on our wifi connections. I will add some more specifics here when i can, but first tests showed that sending one short email took 40 seconds, and a list of 5 very long mails, even to multiple addresses took only 52 sec. Clearly the  best procedure is to use Ocens Mail to store your outgoing mails, then send and collect new with a single log on.

Weather data via WeatherNet download will be mostly in GRIB format, and my rough estimates are 3 days of wind data over a 10º x 10º grid will take up 10 kb. This would be the same for waves or precip or pressure. You can scale that to anticipate the file sizes. But all of the data is stored so you can after downloads look at the data file to see actual file sizes. We will make another note later on about use of downloaded weather and sea state data underway. If all goes as planned there will not be a need for weather downloads, as all this will be provided to the boat daily using state of the art resources from the UW Atmospheric Sciences forecasting team, led by Angie Pendergrass. Nevertheless, it would be good to practice using these powerful resources that will be available underway.

Ocens Mail is the PC interface to the phone. Your actual email writing and reading will be with the program they install called iScribe. A very nice simple email program.  One pitfall is to write an email then send it from iScribe (which just sends it to Ocens Mail) then later decide you want to edit the mail.  If you open it in iScribe and edit and send, it will actually add a second mail to Ocens mail to be sent.  So to be safe, you can inspect your mail list before sending from Ocens Mail menu: Tools/Explore Mail to see what you are about to send.  This is also the best way to go in and change a mail without risking a duplicate.  Sending to multiple recipients does not increase air time.

Note we have numerous ways to track the vessel underway, but if it might ever be useful, you can plug a GPS into the PC and then configure Ocens Mail to send the Lat-Lon with every message. A nice feature of their software.

For our records, I just note that when we first got the phone the total call time was 5:32:44. This will let us check for the total time used to help plan time purchases as needed.

I should add that the Iridium 9555 system is a back up for the boat to its much more powerful Inmarsat Fleet Broadband SatLink system. This will take a note of its own. It can transmit large images and even video, and will serve as the workhorse for daily communications with and from the boat. It too has voice communications, but it it much more involved to turn it on and transmit. Like other high powered systems, you cannot even stand next to the antenna when it is broadcasting.  It is not like a cell phone!

We teach the use of weather data under way in our online course in marine weather, and you are welcome to follow this blog for many practical exercises in wind, waves, and ocean current forecasts.

PS. When the boat capsized the phone was inside the cabin, but not in its protective case, so it was likely destroyed immediately when the cabin filled up with water.  Had it been in its case, they could have called home from the life raft!  We address this point in the new, extended article on sat phones.






Wednesday, November 21, 2012

Row Boats in Ocean Currents

In routine navigation we teach the importance of tidal and ocean current flow on our progress. There are the obvious cases of narrow passes with strong currents that simply cannot be transited against strong flow. You do not even need to go to places like Seymour Narrows with currents of 14 kt. There are many passes in the Pacific NW with peak currents of 5 or 6 kts.

Often we get distracted on figuring the time of slack water to get through these passes safely and overlook the much bigger challenge of getting to the pass at the time we want to. It does not help us if we know when the pass is slack,  if we cannot predict our arrival time because we have 30 or 40 miles to go through weaker 0.5 to 1.5 kt currents that are changing with time. In short, the real nav challlenge is the navigation on the way to the strong current, not the pass itself.

Obviously, the lower the power of your boat the more important this is. So this is more an issue for sailing vessels that go only 5 kts or so, or any kayak or row boat. But to some extent it affects even larger vessels if the strong current pass is a long way off. Even cruise ships take Seymour Narrows seriously.

As the OAR NW team gets closer to pushing off from Dakar on the way to Miami, 3,600 nmi across the tropical Atlantic, we face the fact that this same issue of currents becomes a dominant factor in efficient navigation, even in the middle of the ocean.  This is not like the Gulf Stream, as complex as that is–we did that back in 2006 when they rowed across the North Atlantic–this is mid-ocean without any well defined rivers of current like the Gulf Stream, which is well forecasted and the data easy to come by.

First, it is not even well known that current could be a factor where the boat will be now. If you look on a Pilot Chart, you will see that the predicted (climatic) general flow of the current is right behind them all the way across, at about 12 miles a day drift on average. But like most environmental data, the average is rarely what you get. It is even worse with currents than with wind.

The surface of the ocean, well away from large global patterns like the Gulf Stream or equatorial currents, is still subject to transient patches and eddies of strong current. In fact, the surface of the ocean, underneath the undulations of the swells and waves, is a turbulent surface, like the top of a big washing machine.  Each of these hills and valleys of water give rise to geostrophic current flow as gravity pulls the piles back down.  We can get very good data on the ocean surface heights from radar altimetry from satellites, and this data is then used in mathematical models to predict the currents, which are then either enhanced on diminished by the prevailing winds, temperature gradients in the water and other factors.

In short, with the help of many satellites (not just one) and with sophisticated computer modeling the flow of surface currents can be predicted in great detail. An example is shown below.  These data are from the RTOFS-Atlantic model, operated by the US National Center for Environmental Prediction.  A summary of all models is given at this GODAE site. We use the program Panoply (written by Robert B. Schmunk of NASA) to view these data.

Here we see an ongoing eddy just south of the rhumbline route of OARNW, marked with a red ellipse. The picture below it is zoomed in, and the one below it zoomed in more.



The speed scale in m/s. To get kts multiply by 2 (actually 1.94).  Thus we see in the middle of the ocean, a patch some 30-40 nmi across and some 300 long that varies from NE to E,  all of which is flowing against the boat (which is headed west) at current speeds higher than can be rowed and peaking at some 5 or more kts. This is faster than most of the Gulf Stream!

If the boat wandered into this, or it wandered under the boat, they would be stuck. Their GPS would tell them they were going backwards, but they would not know why nor what to do about it.  These systems could last weeks.  Thus we  have a wonderful opportunity to study these patterns.

So you see the fun we have ahead of us, which is two fold. First use the predictions as best we can to direct the boat to the favorable side of such eddies, and second to use all of the wonderful data we are gathering and transmitting back to UW hourly to measure these currents to see how accurate the predictions are. This data will be available to the public as well.

There are numerous models, so we will try to learn as much as we can about this and report our progress here as we proceed. We will return with another note on how we unfold the current speed and direction from what we measure on board. It will take input from compass heading, knotmeter speed, COG and SOG from GPS, plus true wind speed and direction to account for wind drift, and some (fudge) factor to account for the affect of the waves on boat speed. (We monitor the pitch and roll of the boat at all times, and when it surfs down a wave, the bow will go down at a steeper trim than when just crossing a wave without surfing, and the acceleration shows up in the speed records.)

We can also now follow drifting buoys to get more info on the currents as well. See note on our OARNW Ocean Currents page.

We can now overlay these data onto Google Earth, along with the location of the boat, winds and waves, to get a full dynamic, near live view of the conditions. The data below are from Nov 18, 2012 1400z. It shows strong eddies north and south of the Cape Verdes. Color code for speed is the same as above, ie red is about 5 kts! These test data are compliments of mercator-ocean.fr (thank you), again processed and displayed using Panoply.



For further information on mesoscale ocean currents see the text and many great links in this article from Aviso.  It is an excellent overview with many specific examples.

One of the things you will learn from surfing around the Aviso site is that these anomalous current eddies are not common (could we guess that from the name?) and not uniformly distributed around the world's oceans. There are places they occur more often that others. A couple interesting ones are the Gulf of Tehuantepec and the Cape of Good Hope, neither of which is related to our trip. The Aviso links have the discussion and maps.

To see how these statistics might affect our trip, we first note the terminology: cyclonic eddies are those spinning like Lows, that is counterclockwise in the NH; and anticyclonic eddies spin clockwise, like Highs. The example we show above is an anticyclonic current eddy. Here are pictures that compile location statistics over many years for each type. These pictures are extracted from larger pictures at Aviso.

To show up in these data below, the eddies had to persist for at least 18 weeks, which is quite a long time.  Both types drift slowly to the west at 4 to 5 nmi per day. We may well end up chasing some, just has we had to chase Gulf Stream meanders in 2006, which typically move rather faster than these eddies.  A few days of extra hard work could give the boat a nice ride for a week if we find a good one.

Anticyclonic eddy statistical locations and tracks.
Cyclonic eddies.

These just give us some idea of the probabilities. The one we showed above, for example, is in a region that does not show any activity at all.  We can only conclude that we are more likely to encounter these on the second half of the voyage more than on the first half.  However, a row boat is strongly affected by any of them, so we will be watching at all times.

We still have no direct information that these forecasts are valid and practical on the level we care about. This will be one of the interesting outcomes of the voyage.

We have more information on ocean currents at www.starpath.com/currents.