Kan man lita på Navionics kartografi?

Navionics digitala sjökort
Navionics sjökort används i många olika fabrikat av navigatorer. Exempel på det är Lowrance, Raymarine och Simrad.
Ursprungliga underlaget för dessa digitala sjökort kommer från avritade svenska sjökort för drygt 20 år sedan. Det skedde i Indien där Navionics har ett kartografiföretag. När så Google Earth började publicera bilder över världen började Navionics jämföra dessa bilder med sina sjökort. Då fann man att många strandlinjer i sjökorten inte stämde överens med verkligheten.

Rättar efter Google Earth
Historiskt ritades sjökort för hand av Sjöfartsverkets kartografer. När så satellitsystemet GPS kom och kvalificerad flygfotografering användes upptäcktes, helt naturligt, att "karta och verklighet" inte stämde överens. För att få öar, uddar och kustlinjer i sjökorten att stämma med verkligheten samarbetar idag Sjöfartsverket och Lantmäteriverket i projektet "Nationell strandlinje."
Navionics har valt att göra många av justeringarna via Google Earth. Det är en metod som fungerar rätt långt, men långt ifrån ända fram.

Feltolkar satellitbilder
Ett problem är att satellitbilder ibland feltolkas. Det kan vara en båt som låg förtöjd vid en klippa när satellitbilden togs och vips blir det ett litet skär. Det finns andra fall där båtar kört mellan två öar och Navionics kartografer har tolkat båten som en kobbe.

Navionics säger att de köper kartografiskt underlag S57 från Sjöfartsverket sedan ungefär ett år tillbaka i tiden. De säger också att de jämför Google Earth med S57 och deras egna kartbilder. Allt läggs samman. Frågan jag ställer mig är då hur det kommer sig att ön på bilden ovan finns i Navionics sjökort, men inte i verkligheten? Något annat svar än "alla sjökort har sina fel och brister" får jag inte.

Frågan är, köper verkligen Navionics underlag från Sjöfartsverket eller säger de bara så för att bemöta Garmin och BlueChart? Garmin och BlueChart använder från våren 2011 S57 underlag till sina sjökort. Om även Navionics gjorde det borde de väl ha minst samma utseende som S57? Hur kan annars viken nedan skilja sig så mycket åt?

Sjöfartsverkets försäljningsombud heter Primar och ligger i norska Stavanger. Det är genom Primar som försäljning av S57 underlag sker. När jag frågar Sjöfartsverket om de sålt några S57 underlag till Navionics får jag veta att det finns en sekretessklausul i avtalet mellan dem som säger att Sjöfartsverket inte får säga huruvida Navionics verkligen köper underlag eller inte.

Korrigerad strandlinje - vad händer med 3-meterskurvan?
Navionics får med många detaljer i sjökorten när de jämför med Google Earth. Bryggor, pirar, nya hamnar etcetera kommer med i sjökorten. Här är de i många gånger steget före Sjöfartsverkets S57.
"När strandlinjer flyttas och korrigeras, vad händer då med den mörkblå tremeterskurvan", frågar jag under vårt studiebesök i Italienska Viareggio där Navionics har sitt huvudkontor. Jag får ett svar som jag inte riktigt kan tolka, men det verkar som om den flyttas parallellt med strandlinjen. Här är en viktig skillnad. Om en strandlinje flyttas inåt eller utåt går det inte bara att flytta med tremeterskurvan har Sjöfartverkets kartografiska avdelning lärt mig. Tremeterskurvan mäts med ekolod och det finns inget givet samband med att den kan flyttas lika mycket som strandlinjen. För att tremeterskurvan ska bli rätt i sjökortet i förhållande till verkligen måste den positioneras och lodas genom sjömätning.

Kan man lita på Navionics sjökort?
Navionics ligger i framkant när det gäller sjökort och kartor och hur de används och hur de kan rättas. Idag kan vi köpa Navionics sjökort till våra mobiler, läsplattor, PC och navigatorer. Mer om det i kommande artikel.

Huruvida det kartografiska underlaget är så bra som de själva säger tar jag med en nypa salt. Alla sjökort har fel någonstans. Vi kommer nog aldrig att få uppleva sjökort som är 100 % rätt. När det gäller grundflaks positioner litar jag mera på S57 underlag från Sjöfartsverket än på Navionics. När det gäller strandlinjers konturer, bryggor och hamnars utseende ligger Navionics i många fall steget före. Viktigare är dock för mig att kunna lita på grundområdens positioner. Där har jag för vana att alltid ha god marginal, ibland långt över 100 meter, och det lär jag få fortsätta med under många år framåt. Landkonturer ser jag ju med ögat och om en grupp synliga stenar inom ett ytterst begränsat område är fyra eller fem till antalet, eller om en bränning mera liknar en lite kobbe tycker inte jag spelar så stor roll. Jag ser ju dem, men jag ser inte ett grund på 1,2 meters djup om jag inte är mitt över det och där vill jag inte vara, därav marginalen.

Sjöfartsverket utför ständigt sjömätningar och korrigeringar sker löpande på grundflaks positioner med mera. Navionics gör inga egna sjömätningar i Sverige utan där använder de sig av bland annat Google Earth, bilder från Lantmäteriet, underlag från Hydrographica och sina gamla underlag från avritade papperssjökort. Huruvida de köper S57-underlag från Sjöfartsverket och hur dessa underlag i så fall används för korrigeringar har det inte gått att få ett säkert svar på. Det beror på vem man frågar.

Navionics har ett system där vi användare kan påverka innehållet i sjökorten och påpeka felaktigheter. Det kallas för UGC (User Generated Content). Via mobilen eller till exempel en iPad kan vi meddela fel eller tala om att en naturhamnen "NN" har en fin klippa att förtöja mot eller annan information som vi gärna vill sprida. All data samlas in. Rättelser i sjökorten kontrolleras av Navionics och därefter uppdateras sjökorten. Där ligger verkligen Navionics i framkant av utvecklingen. Mer om det i kommande artiklar.

//

I regret to observe that this article is inaccurate and may induce to conclusions that are potentially dangerous for the reader.
The point is that ALL charts have errors: this applies to Navionics as well as official paper charts, S57 and any other charts by other manufacturers.
Navionics alone has a billion features in its charts, which means that even with an error rate of 0.1% (which few products ever achieve) we still have a million errors. Again this applies to Navionics as well as any product made by anybody else, whether it is the Swedish HO paper charts, or the Swedish HO S57 data, or the product of another company.
To write an article in which one picks an error of Navionics which others do not have, and not pick errors that others have but Navionics does not have (and you can rest assured there is plenty of that) is not only unfair, but also dangerously misleading, because it induces readers to blindly trust products that do contain errors.
Which chart is the safest?
Is the safest chart the one with the least number of errors?
Not long ago in Norway a ship hit a rock that was not reported on the official HO chart; as a consequence the ship sunk and the entire crew died. Even if that happened to be the only error in that specific chart, can we say that it was the safest chart for the area because it only had 1 error, while other charts had more errors, but not that specific error and no accidents happened with their use?
So the bottom line is that NO chart is error-free and to imply that other products have no errors, or less errors, or are safer than Navionics, is materially inaccurate and potentially very dangerous to the reader.
The big difference between Navionics and all others is that Navionics has empowered its users to make corrections to their charts, and those corrections are instantly distributed to the entire community.
Sweden has millions of boaters, which means millions of potential surveyors who can make corrections to the nautical charts, and share such corrections in real time with the entire community.
And if one user makes a mistake and introduces instead of correcting an error, immediately other users can correct it and even report an abuse if need be, and Navionics can disable that user from doing any further damage; but I am happy to report that so far only one such person out of hundreds of thousands of users worldwide has existed.
No chart making organization, whether a government HO or private company, can deploy as many surveyors at sea and make as many corrections and make them available to the entire community so fast and so easily: to add a rock to your chart takes about 30 seconds of work, and from that moment onward your own chart is correct, plus the chart of the entire community is correct, plus you will benefit from all the corrections made by all the other members of the community.
Nothing can beat that: well worth the €10 investment in an app that does that plus is a full backup to your main GPS.
===========
I also have specific comments:
===================
1) Navionics uses 20-year old paper charts as source.
I do not know how this was invented.
Navionics uses all available sources which includes S57, Satellite pictures (not only Google but also more up to date pictures as we have shown to the Author in Viareggio), Hydrographica where available, and of course paper charts.
The paper charts we use are not 20 years old, but the most up-to-date available, further updated with Notices To Mariners. These are the ONLY charts legally available to boaters in Sweden: Swedish boaters by law must use and rely on official paper charts or on ECDIS, which is the only legal electronic chart but is so big and expensive that cannot be used on boats. S57 data used outside ECDIS is NOT a legal replacement for the paper charts: using S57 on a GPS plotter in lou of paper charts means violating the law.
Therefore it is totally inaccurate to say that Navionics uses 20-year old paper charts, and totally wrong to imply that boaters should rely more on S57 on plotters than on paper charts, because that amounts to instigating a violation of the law.
2) comparison with Garmin
Garmin's charts have errors, just like the charts made by anybody else. To pick examples of errors that Navionics has, but Garmin does not, and not do also the opposite, which means show errors that Garmin has but Navionics does not, is not only unfair, but misleading and ultimately bad journalism.
We do not engage in mud slinging, therefore we do not engage in the nonsense of pointing out the errors that Garmin has.
We think that for the greater benefit of everybody it would be best to focus on what instruments are available to minimize errors and to make sure that corrections are instantly available to the entire community with no effort.
3) boats mistaken for rocks or land features
While I cannot exclude that this may have happened, it seems a gratuitous accusation:
· First, we have not seen any such occurrence
· Even if it occurred, it is safer to report a boat moored to a rock as an additional rock than not report the rock at all, and we have seen examples of such occurrence, therefore instead of making fun of Navionics it would be more correct to highlight the fact that it has corrected many errors by way of satellite imagery, and if anything erred on the side of safety
· We have seen examples of ships that have been turned into museums and have become a fixed feature, albeit a floating one, and it is correct to report them in the chart as a fixed feature. Infact they are reported as a fixed feature by the HO paper chart, by the HO S57, by Garmin and by Jeppesen
4) A tiny island shown in Navionics chart that does not exist on S57 nor in reality
Once again, aside from the fact that for fairness and accurate journalism, examples of the opposite sign should also be made; and aside from the fact that this is an error in the direction of safety, we have explained that when comparing different sources we find something like a rock that is not reported in S57, knowing that S57 in some cases does miss rocks, for the sake of safety we do not remove it from the chart unless we have total evidence that the rock does not exist. This is the wise and safe course of action, not at all an "embarrassing" fact as the article reports.
5) Making fun of the use of satellite imagery
Aside from the fact that we have shown examples where we use satellite imagery that is more up-to-date than Google, I again take exception at the fact that the article makes fun of Navionics for introducing few errors in interpreting satellite imagery, and does not point out the huge mass of errors that have been corrected and detail improvements that have been made thanks to satellite imagery. This is a gross misrepresentation of reality, and quite frankly tantamount to arguing that not using satellite imagery is better than using it, which is archaic.
6) rubbersheeting
The article says that we were vague in telling how we make corrections when we find cartographic features that are shifted compared to reality, and says that we do it wrong.
I find that offensive and incompetent.
Rubbersheeting is the most advanced technology available today to make such corrections. I challenge the Author and anybody else to find a better technology.
While this may or may not yield 100% correction, it is the best that humankind is able to achieve today, and again I challenge anybody to prove me wrong.
I also challenge the statement that I have been vague in answering the question (implying that I had something to hide). I fully accept that the author is not a GIS specialist and may not well understand such technical details, but for sure cannot accuse me of being vague and dodging an answer. Throughout the meeting I have always been more than keen to answer any question and to repeat whenever needed. This one is no exception and I am more than happy to spend as much time as needed to fully clarify the technology. If the Author did not understand, should have asked again instead of writing that I was hiding something.
7) use of S57
The article basically portrays Navionics as lying about the use of S57, which I find insulting.
First of all, let me make a clear statement: FMA and not Navionics requires a confidentiality clause in its licensing agreements. I made that clear during the meetings when I said that if you ask FMA whether or not we have a contract they might decline an answer, but cannot say that we do not have a contract, because it would be untrue.
Let me also say that we received evaluation samples of S57 back 2010, therefore it is correct to say that we started using it about a year ago even if we signed a contract at a later stage. To imply that we are liars because we said we started about a year ago even if we signed the contract only 6 months ago is a gratuitous insult and a journalistic ethic violation: before writing his personal conclusion the Author should have asked us the reason for the apparent inconsistency of dates.
Let me also take this opportunity to explain once more what we do with S57.
Navionics has been using S57 from many countries for many years, so we have a lot of experience on the assets and liabilities of using S57.
At the beginning when we received S57 we just erased our database and replaced it with S57.
Furor from the market ensued, because we lost so much detail and value added accumulated over the years, that multitudes of customers were infuriated by the regression.
As a consequence of our negotiations with FMA, and then our contract, we first received samples and then full set of S57, but before erasing our database and replacing it with S57 we invested quite some time comparing the two.
The result of the investigation was once more that we would have gained a lot, but also lost a lot, and we would once again be flooded by infuriated customers.
As we did not want to repeat the same mistake done in other countries, we therefore made the conscious decision of retaining our database, value-added with corrections from satellite imagery, from other sources like Hydrographica and others, from the UGC received from the community, and cross check it with S57 for further corrections.
Can we still find places where S57 is better than our charts? Of course yes.
Can we find places where our charts are better than S57? Of course yes, plenty.
Is it correct to tell readers that they should buy Garmin instead of Navionics-based GPS because it is the only way to have safe charts? This is a gross misrepresentation of reality, and one that can induce boaters to blindly rely and navigate in a hazardous way, which amongst other things is also in violation of the law.
Best regards
Giuseppe Carnevali
President
Navionics