Rowing to the Azores: Tracking a Transatlantic Trek
The Azores. These volcanic islands in the Atlantic have enthralled me for years, offering stunning hikes, beautiful beaches, and an all-around tropical experience for those not quite wanting to be in the tropics. The small Portuguese island chain has topped my list of travel destinations for nearly a decade.
And, as I write this, I’m just about halfway there, rowing on calm sea waves about 1,000km off the coast of Virginia. The sun is shining through lightly overcast skies, and occasionally I see a sea bird far overhead. Or at least that’s what I imagine it’s like, as I write this in a DC coffee shop. Let me explain.
A Path to Home
Finding an exercise routine has also been a bit of a challenge for me. After years of lapsed gym memberships and uninspiring classes, I finally found a stylistic match. OrangeTheory Fitness, with their data-rich tracking systems, was inspiring- the data nerd in me was so excited that I’d found a challenging exercise routine that had so many numbers involved! It was genuinely fun tracking my progress from class to class. That was, until March 2020, when the absolute last thing anyone wanted to do was stay in an enclosed space with others huffing and puffing.
Without those classes, it was back to the drawing board trying to find a sustainable exercise routine. Well, almost- what I’d learned from OrangeTheory was that, for me to make exercise take hold, I needed to gamify my exercise, and make it all about the data. Something I could track; something I could visualize; something I could set goals for and progress towards. With that as a foundation, I began laying out exercise routines with measurable metrics, and tracking them in a couple rudimentary spreadsheets. Eventually I expanded on this, writing scripts that would generate a weekly workout routine based on a variety of workout types, lengths, and styles.
That worked well enough for a year or so, but eventually I saw even this strategy waning. While I was working on improving specific metrics, such as my 5-mile bike time, none of these laddered up into a larger objective I could measure. Then I had a revelation. I’d always wanted to visit the Azores, but also wanted a reason to make such a big trip. What if I were to row to the Azores, and upon reaching Ponta Delgada I’d buy a plane ticket out to the islands? Less than a week later, a rower had arrived at my door.
Traveling in Comfort
This is the latest update on my rowing journey, after nearly two years. This dashboard is powered through a combination of Google tools- Forms to collect data points, Sheets to act as a database and perform important calculations, and Looker Studio to visualize the journey. While my current location is somewhere in the Atlantic, the trip has been performed comfortably from my own home.
The Data Below the Waves
No sooner had I ordered my machine than I began planning how I’d track this journey. Obviously, we’d have some sort of spreadsheet involved- Google Sheets is my preferred tool, especially without any sensitive data being collected. Easy enough. Of course, with Sheets as a data source it was just a hop, skip, and a jump to building out a primitive Looker Studio dashboard. By the time the rower was delivered, I already had both of these in a prototype phase, capable ot tracking the distance I rowed, my average strokes/min, and a few more data points for each of my future workouts.
And this worked well enough for my first few rows. I’d finish rowing, log on to my computer, input my data, and continue with my day. The process was simple, but hoping I’d remembered all the data correctly after each row was just a bit too much. To solve this, I employed Google Forms, creating a bare-bones submission form that could collect key metrics and then add them to my existing Google Sheets database. With a link to the form on my phone, I could submit my performance after each exercise in just seconds, and without even having to get off the rower!
With the trifecta of tools set up, I was confident I had created the core data ecosystem I’d need for my journey. However, this was the easy part. Visualizing the journey was the real challenge.
Traversing the North Atlantic Current
While a straight line from Washington, DC to The Azores would’ve been a perfectly fine abstraction, I wanted this to be both a physical and mental exercise. Pretty quickly, I realized that the biggest challenge for me would be determining how to approximate my journey along the North Atlantic Current, an arc that spanned thousands of miles and possessed strong currents to support my rowing efforts.
Tide Lines
One of the benefits Looker Studio provided me was a built-in integration with Google Maps, allowing for a geographic visualization of any location data I provided it. Of course, at that moment, I didn’t have location data- I only had my distance rowed each trip. To map out my location, I needed to translate my distance rowed to a portion of my journey, and map that along a geocoordinate grid. I did this by finding the Latitude and Longitude of both my beginning and end points, calculated the distance between those points, and used that to calculate the change in Latitude and Longitude per kilometer rowed. Simple! I plugged in the coordinates for Washington, DC and Ponta Delgada, and I was… rowing somewhere on the National Mall.
While going in a straight line would be the simplest way of tracking my progress, I’m a bit of stickler for accuracy, even in a symbolic act. I began looking at maps of the region, and after some quick calculations, realized that beginning in Annapolis, MD would be effectively the same water distance as starting in DC. From there, I could head down the Chesapeake Bay to Virginia Beach, do a straight shot to Ponta Delgada, and be done with all this geocoordinate mess!
…Or so I thought, until I realized just how big of an endeavor I’d undertaken. The Azores are most of the way across the Atlantic, and my initial calculations put my total travel distance at about 6,000km. Rowing 5km per session (which was well above what I was managing in those first few rows), I’d need to be “on the water” for 1,200 days- more than 3 years of rowing without missing a day! I was shocked at how long it would take, and knew with that level of dedication I would give in before I made it far. I had to come up with a plan- if Ben Lecomte could swim the Atlantic in less than 100 days, surely I could make it to the Azores in under 1,200!
Generating Data
The mathematical work I had ahead of me required that I could abstract the Atlantic into a simple XY axis. To do this I needed to map locations based not on Latitude and Longitude (which are not equidistant), but by relative distances in kilometers.
To do this, I found a free GIS map using ESRI, which I’ve used in the past for work projects. This map let me layer a map of ocean currents on top of Latitude and Longitude lines, and could also quickly provide the distance between any two locations. These distances and corresponding coordinates were critical in making the final calculations that powered my future path.
Keeping Current
The answer lay in currents- specifically, the North Atlantic Current (NAC) and Gulf Stream. These produce of the world’s strongest surface-level currents; a rubber duck could reach the Azores in months once in the thick of the current.
With this information in hand, I went about figuring out how to best utilize these currents for my journey. Utilizing online maps, I found where I could most quickly intersect with the current from Virginia Beach. Even though this approach took me slightly further south, the speed of current quick neutralized any lost ground. Then came, to my surprise, the hardest part of this whole process: mapping my location, in geocoordinates, based on my location along an arc.
Let’s take a step back. The NAC, flows roughly along the East Coast of the US, curving eastward into the Atlantic and towards Europe. If I wanted to maximize my utilization of the current, I couldn’t just travel in a straight line- I’d find myself somewhere in the center of the North Atlantic Gyre, needing to traverse thousands of kilometers with little to no support from the surface current. Instead, I needed to follow the arc of the current, following until the point where best I could diverge from the main current and head south toward the Azores. But how could I do that?
Looking at the path of the NAC along my route, I noticed it formed a roughly parabolic curve- if I could calculate a formula for the arc of the current, I could then determine its length. To do this, using oceanographic maps, I plotted out a handful of points along the NAC that roughly represented the vertex of the parabola, where I would first and last row directly along the current, and a couple of points between those. Using this, I built a calculations table where I could control for the variables:
y = (ax+d)^c * b
The formula was simple- a, b, c, and d were all constants that could be adjusted as needed, while X and Y were the distances, in KM, from the vertex of the parabola. As I adjusted the constants in the formula, I compared the outputs with the points I had already plotted out along the path, looking for a set of constants that managed to most accurately align with all those points, Through trial and error, I slowly found the function I needed for my parabolic path, and much to my surprise it was actually quite simple:
y = (x)^¹/³ * 100
With the function found, and knowing my beginning and end points, I could now calculate where I would be along the arc after traveling a certain distance east- an improvement from before, but still not nearly enough for me to map my journey along the arc. Calculating distance along an arc requires performing a derivation, which wasn’t something my little spreadsheet couldn’t calculate formulaically each time I submitted a form. Instead, I took a more straightforward approach. First, I determined the total distance I would be traveling along the arc- my calculus skills are a bit rusty, so instead of doing the equation myself, I let ChatGPT crunch the numbers for me. I then built a table, calculating the Latitude and Longitude for a point every 1km along the route, and updated my mapping formula to lookup from this table whenever my row was along the route.
Utilizing AI to Crunch the Numbers
The formula for calculating distance along a parabola requires a derivation which, while manually calculable, was something I entrusted to AI. Providing the formula I’d approximated, alongside the grid coordinates I’d generated from the GIS map, I was able to give a clear and concise prompt for ChatGPT.
Despite its increasing prevalence, the ability for AI to confidently hallucinate a response is always a worry of mine, and so I also asked for a graph showing my request. This let me compare the arc in the generated graph to the one I had in Google Sheets, minimizing the likelihood that ChatGPT had misunderstood my question and provided a plausible but inaccurate answer.
Finishing Touches
With all the hard math behind me, my final task was to determine just how much support the current would provide on my journey. The Gulf Stream is one of the strongest surface currents in the world, and while the NAC is not as powerful, it is no slouch. From averages found online, I went with what I felt was a reasonable 1.5km/hr. I’d apply this in much the way Mr. Lecomte did during his swim, and utilize the current for 24hrs during each day that I rowed, effectively adding 36km to any row occurring during that section of my trip. While that may sound like a lot, and it is, the arc I am traveling is over 4,300km- I’m still on a 100+ day journey when fully supported by the current.
Overall, this project has enriched my mind and body- finding a way to map geocoordinates to distance along an arc, fully automated through Sheets formulas, was an exhilarating challenge. I’m always at my happiest when I get to combine different data sources and references in my work, and pulling together maps of currents, calculus functions, and seafaring data made this a fun passion project. And while the end result of the arc along the Atlantic isn’t particularly dramatic in appearance, I’m glad to know I went with a solution just a little more complex than drawing a straight line and estimating a lower amount of support from the prevailing currents.
In total, I expect that I’ll reach the shores of Ponta Delgada after around 300 days of rowing. For someone who has struggled to maintain an exercise regimen for more than a few months at a time, I’ll take that as a win.
