THE CHARTMAKING PROCESS WHERE DID THE REEF GO? MIXED MESSAGES
The saloon is quintessentially Azimut; with luxe materials and high levels of finish, no amount of comfort was spared. Far left: Winning features of the lower helm: sliding glass panel to bring in the sea air and comfortable helm seats. Left: Keeping beat with the excellent furnishings, the lower helm sports twin Raymarine MFDs.
It’s a dark and squally night in November 2014. Vestas Wind, one of the boats participating in the roundthe-world Volvo Ocean Race, is romping along at 16 knots in the Indian Ocean. There is a loud crack—the daggerboard has sheared off. The boat pivots hard to port; the keel strikes the bottom and Vestas Wind is aground in 6-foot breaking seas on the Cargados Carajos shoals, 240 miles northeast of Mauritius in the Indian Ocean.
As waves wash over her, Vestas Wind pitches and rolls violently, and starts to break up, losing both rudders, the keel, and large sections of the hull. Water floods belowdecks and the ship’s lithiumion batteries begin to smoke, threatening to go up in flames. In the early hours of the following morning, the crew scrambles off the back of the boat onto the sheltered side of the reef from where they are rescued later that day. All are safe and unharmed.
No one aboard was aware the reef existed, in spite of the fact that the area had been resurveyed from 2008 to 2010 by the Indian Hydrographic Office. The electronic charts in use were accurate and current, and a highly experienced navigator was in charge of route planning. So how could this happen? This accident dramatizes a fundamental weakness in many of the electronic charting systems in use on recreational boats. To understand it, we have to look at how charts are made. Hydrographers take soundings by running straight lines and accurately positioning these and any submerged landmasses. In shallow and inshore waters these lines are closely spaced to ensure that no intervening features are missed. The deeper the water and the farther offshore, the survey lines get more widely spaced, because missing the features between them is less relevant.
Up until the widespread use of sidescan sonar (starting in the 1960s), there was little or no lateral vision when taking soundings, and prior to 1939, none at all, especially if the soundings were collected with a lead line. Prior to the advent of satellite navigation systems, the positioning accuracy used to plot soundings was significantly less than what any off-the-shelf GPS can deliver today. Over half of the soundings on current U.S. charts were collected using pre-sidescan-sonar and non-satellite-positioning techniques. In other parts of the world, it can be anything up to 100 percent (for example, much of the Pacific, Bahamas, and Caribbean; even substantial sections of the UK’s coastline).
Hydrographers typically conduct surveys at a larger, more detailed scale than those used to produce the charts derived from the survey. As many soundings as possible are collected and crammed onto the survey sheet, to the point that traditional, pre-digital survey sheets can be almost illegible. The cartographer has to reduce this mass of data to a legible level without losing the essentials.
For obvious reasons, the first priority of cartographers has always been to display areas of shallow water together with hazards, such as rocks and reefs. Large-scale charts that cover a small area in great detail have room to add plenty of soundings, such as narrow channels between shoal areas. Small-scale charts (covering a large area in little detail), on the other hand, do not have the space to legibly add fine detail once the low-water areas have been inked. As a result, the more you zoom out with paper charts, the larger the shallowwater areas get. There is absolutely no misunderstanding concerning hazards. There are two types of electronic charts: raster and vector. A raster chart is an electronic photograph of a paper chart, so the same shallow-water features are reproduced: Shallow water is always clearly identified, regardless of scale. A vector chart is an entirely different animal. All of the data on the most detailed (largest scale) chart or survey sheet is captured in a database and is available to be displayed at any chart scale. When you zoom in, at some point you get all this information. When you zoom out, if this information were to be retained, the chart would rapidly become so cluttered that it would be unreadable. This is where the traditional cartographer removes detail to make the chart legible, preserving the shallow-water soundings. Reducing the clutter on vector charts is done in the software.
It is not uncommon for vector software to remove the shallowwater soundings. Given the zoom level at which the navigator aboard Vestas Wind planned his route, and at which the digital charts were being used while at sea, the reef simply disappeared from view. The shallowest water displayed was 130 feet. The navigator plotted a course directly across the reef. Had the navigator consulted a paper chart at any scale, the reef would have been clearly visible.
Note that most of the current crop of multifunction display devices (MGD) can now display worldwide raster charts in addition to vector charts, which was not possible in the past due to higher memory requirements of raster charts). While a benefit, raster charts often still don’t include access to important paper chart elements, such as the source map. On a dark night in January 2013, the minesweeper USS Guardian powers at 7.5 knots directly onto the Tubbataha Reef in the Philippines in 4- to 6-foot seas. All efforts to free her fail. Later, in rising seas she swings broadside onto the reef, breaking her keel and punching numerous holes in her wooden hull. Waves sweep over the deck and drive her farther on, and she is at risk of breaking up. The crew is evacuated. Guardian is subsequently cut into three pieces, each small enough to be lifted off the reef by a giant salvage crane.
How could this happen on a navy vessel with its multiple navigational checks and balances, especially given the fact that the reef was clearly displayed more or less in its correct position on the small scale (general) vector chart? The problem here was the navy’s large-scale (coastal) vector chart had the reef 7½ miles out of position, based on erroneous satellite imagery. This had been known since 2011 but had not been corrected due to “human error.”
Guardian’s skipper and navigator noted the discrepancy between the small-scale and large-scale charts and simply assumed that the more detailed chart must also be the more
ocean, survey lines can be many miles apart, missing massive details in between.
These sorts of things don’t just happen in remote corners of the world. In 1992 the Queen Elizabeth II ( QE2) cruise ship, at that time the flagship of the Cunard Line carrying over 2,800 passengers and crew, struck an uncharted ledge off Cuttyhunk Island in Rhode Island Sound when traveling at 24 knots. A long gash was cut in the ship’s starboard side. The QE2, with a draft of 32 feet, was sailing in waters charted to a depth of 40 feet. The ledge fell in between survey lines.
A couple of other factors may have been at work in the QE2 grounding. We have a serious anomaly on all U.S. charts (paper and electronic). The displayed soundings are based on something known as Mean Lower Low Water (MLLW). This does not equate with extreme low-water events, such as those that occur during spring tides. The water level can, on occasion, be significantly less than the charted depth, sometimes by several feet. Most of the rest of the world uses a more conservative low-water chart datum, known as Lowest Astronomical Tide, or LAT, so this problem does not occur. Many U.S. paper charts have a table somewhere on the chart that has an Extreme Low Water column that informs you by how much the water level can go below the charted depth. These tables do not show up on vector-type electronic charts, and are likely to be hard to find on raster charts.
The other circumstance that may have been relevant in the QE2 grounding is the fact that when you power a boat with little water beneath the keel, the propeller tends to suck the water out from under her, lowering the boat relative to the surrounding water. This is known as squatting, and if the clearance beneath the keel is small enough, you can suck out the water until you are dragging the bottom, at which point the faster you try to go the more you drag, and the slower you actually go. We learned this long ago when we built and lived on canal boats in the U.K.’s shallow inland waterways.
We experienced the squatting effect dramatically one night when heading up the Mississippi Gulf Outlet Canal (now closed, because of the disastrous role it played in the flooding of New Orleans during hurricane Katrina). A fully loaded container ship passed us dragging the bottom and moving just fast enough to maintain steerageway. We edged over to the side of the channel to get out of the way. As the ship came past it sucked the water out of the channel leaving us temporarily hard aground and heeled way over. It’s the summer of 2003, the sun is shining and a fresh breeze is blowing—a great day for a cruise. My neighbor has just bought a catamaran. He calls: “Nigel, let’s go boating.” I am working on the page proofs of a new book called How to Read a Nautical Chart (published by McGraw-Hill and now in its second edition). “I’m sorry Mike, I can’t. I have to finish these page proofs.” His swift reply, “Bring them with you.” So here I am, at the wheel of Mike’s brand-new pride and joy, reading the page proofs of How to Read a Nautical Chart, when I run his boat onto a rock at 7 knots. We are both thrown across the boat. Subsequent investigation reveals a crushed keel.
The fault here? Mike did not have an electronic charting system on board so we were using paper charts. The first time I had sailed these Maine waters (years before) I had made a mental note that if I stayed outside of a line between Haddock Island and Western Egg Rock, I would clear the rock with room to spare. Subsequently, I did not bother to look at the chart, which was unfortunate, because I had memorized the rock on the wrong side of the line. For years I had sailed over this same rock and never struck it because the tide had been high enough to clear it. On this day, we caught it at low tide. In all likelihood, if we had electronic charts onboard and displayed at the helm, we likely would not have run aground.
My point here is that ultimately, most groundings are the result of human error rather than chart error. Like most recreational navigators today, to avoid the kinds of mishaps illustrated above, I now rely heavily on electronic charts both for route planning and for navigation. There are numerous benefits as compared to paper charts, including real-time display of the boat’s position (typically to a greater degree of precision than the display of the chart data) without the possibility of navigator error, overlaid AIS and radar displays, and various “added value” features, such as the ability to pull up instant tide tables. I have evolved a set of processes to minimize the associated risks.
We always carry paper charts, typically at a scale of somewhere between 1:80,000 and 1:150,000. This is not detailed enough for difficult harbor entries, but is good enough for most other navigational purposes and as a back-up to the electronic charts without breaking the bank. It also ensures that all shallow-water areas are clearly displayed and gives me additional data, such as notes and source diagrams that are unlikely to show up on the electronic charts.
Because of the relatively small size of even the largest electronic chart display as compared to a paper chart, and also the fact that the screen resolution on electronic displays requires you to zoom in to some extent to make any chart legible, further reducing the coverage area, I do my initial route planning for all but short and uncomplicated passages on the paper charts. This determines whether or not I have a relatively straight shot between the departure and arrival points. Let’s assume it is a straight shot.
I zoom out the electronic chart until I have my two waypoints, and then I lay down a track between them. Then I zoom in and look for hazards and other features that might require an adjustment to the track, scrolling from one end of the track to the other, moving the track and adding waypoints as necessary. The final step is to zoom in to the level at which all available data is displayed. I scroll along the entire track once more, rechecking all the soundings and other features along the track and in its vicinity. I make adjustments as necessary, highlighting potential dangers, and reviewing alternative tracks (which I often add to the chart) and emergency anchorages in case conditions change. This route becomes our primary navigational tool.
There are increasingly sophisticated software packages that automatically perform many of the functions just described, with Navionics’ Dock-to-dock Autorouting being the best-developed at this time. The Coastal Explorer PC charting program has a