Part 1 – Mapping My Air Travel Adventures

I love travelling. Whether it’s the excitement of exploring new places, the inevitable collection of random souvenirs that always seem like a good idea at the time (I am looking at you "random rocks from the Lima shore"), or just the feeling of temporarily escaping the daily grind, it’s something I truly enjoy. (I even made it into a game for myself by searching for a "Travel-Score" to dunk on friends and family) But let’s be real—travelling isn’t exactly an environmentalist’s best friend. Every flight I take contributes to the massive, invisible CO₂ cloud floating, and that got me thinking: How bad is it, really?

So, in the spirit of both curiosity and eco-guilt, I decided to dive into the numbers. I also wanted to have my facts right before I get "flight shamed". I documented every single flight I’ve taken since I was born. Yes, every flight—because I’m nothing if not thorough. And for those of you wondering, flights with layovers count as separate ones for this calculation. This resulted in digging into my old emails and photos to recollect any trips and also of course directly from memory. I wrote them down in a simple file with the year, departure and arrival city name.

I then used the city names to find their exact latitude and longitude with the help of geopy and an old script. Once I had those, I calculated the distance between the cities using the haversine formula. The result was a shocker: I've flown far enough to make a small dent in the ozone layer and I have taken 139 flights as of today!

Next, I used the transport_co2 library to estimate the CO₂ emissions based on the distance for each flight. After calculating all that, I thought, why not slap it on a map? Despite not being my favourite type of projection, I used the Mercator projections as the others produced ugly corners in the visualization. I always wanted to create a map like this to try and recreate the old Facebook friends map. I got the ideas for the color-coding from this blog where a power law scaling is recommended for small data sets (which it is in this case, thankfully!). Of course, the color coding is done based on the CO₂ emissions: the darker the color, the lower the emissions. You can zoom into the map and also hover on each flight path to get an estimate of the tons CO₂ equivalent emissions (tCO₂e).

Part 2 – What Does This Carbon Emission Actually Mean?

Okay, so now that I’ve mapped out all the flights I’ve taken (and probably scared myself with those numbers), I wanted to figure out just how much this 111.65 tCO₂e actually means in real-world terms. Because, let’s face it, throwing a massive number out there is one thing, but unless I’m visualizing it in relation to something I know, it’s just another abstract figure. So, I dug deeper.

The easiest way I thought of visualizing this is to do some sort of one-factor-at-a-time sensitivity of analysis comparing the total cost of my flights against various lifestyle factors such as transport methods, food choices, number of trees required to offset this and also the carbon cost of this. Of course, this is not a scientific study but only something based on a few hours of doing "internet research". So take this with mountains of salt!

Note: I am only comparing the carbon footprint of flights against other factors, but the total carbon footprint of a person would involve all these factors combined.

How Far Could I Have Gone in Other Transport Modes?

First up, I thought, "How does flying compare to other modes?" What if I had chosen other means of transportation? How far could I have traveled using other modes? To keep a standard I measured each mode of transport in terms of number of Earth’s equatorial circumference which is about 40,075 km (Hover over each bar to view the distance as a multiple of Earth's circumference)

I used the same transport_co2 for the other modes of transport (except walking and bicycle, more on those below). From this library, I can calculate the emitted CO₂ for a given distance travelled, however, I wanted to calculate the other way around i.e., calculate the distance for a given CO₂ emitted. In addition, I did not want to find out the inner workings of the function but made the assumption that a larger distance would result in a larger CO₂ emission. With that assumption, I was able to implement a simple bisection method to find out the distance for a given CO₂ emitted, specifically the total CO₂ emitted by all my flights taken so far.

As mentioned before, the distances for walking and biking were not from transport_co2 library instead calculated based on calorie consumption and CO₂ emissions associated with that energy. To estimate the CO₂ offset by walking or biking, I assumed an annual vegan diet's carbon footprint—around 1.5 tCO₂e per year as mentioned here. Based on an average daily intake of 2600 calories.

For walking, assuming 80 calories burned per kilometer as mentioned here, the distance required to offset a CO₂ target can be estimated using:

\[ \text{Distance walked} = \frac{\text{CO}_2 \text{ target} \times \text{Calories per year}}{\text{CO}_2 \text{ emissions from diet} \times \text{Calories burned per km}} \]

Similarly, for biking, where the calorie burn is about 23 calories per kilometer as mentioned here, the distance required is:

\[ \text{Distance biked} = \frac{\text{CO}_2 \text{ target} \times \text{Calories per year}}{\text{CO}_2 \text{ emissions from diet} \times \text{Calories burned per km}} \]

Where: \[ \begin{array}{rl} \text{CO}_2 \text{ target} & = \text{The total emissions from all my flights} \\ \text{Calories per year} & = 2600 \times 365 \, \text{kcal} \end{array} \]

Of course, these are just fun numbers to calculate and don't have a rigorous scientific basis. However, what was interesting to see was that some modes of transport turned out to be more efficient than walking or biking. This perhaps concurs with what is already known

Years of Diet Comparison – Meat Lovers, Take Note!

We all know that our food choices can make a big difference in reducing CO₂ emissions. I wanted to see how my CO₂ emissions from flying stack up against the emissions saved by different diets. For example, in how many years would a meat eater emit the same CO₂ as a vegan? Although I’ve mostly been vegan for a few years, this comparison helps put the impact of flights into perspective against everyday choices like diet.

Using the numbers from here and the calculations shown below, we can estimate the number of years for each diet to offset my total flight CO₂ emissions:

\[ Y_{\text{diet}} = \frac{\text{Total CO₂ from flights}}{\text{Yearly CO₂ from each diet}} \]

We see below the equivalent years (represented by number of birthday emojis 🎂) of different diets to offset the total CO₂ emitted by my flights so far:

It is good to know that the difference is substantial and well within a single person's lifetime. So, this gives me hope that our lifestyle choices can realistically impact our CO₂ emissions. (Note: I realize that diet is an emotionally charged topic and I promise to keep my (very mild) judgment in check! 😄)

Number of Trees to Plant – Start Digging Holes

We know that planting trees is one of the effective methods to offset our emissions. How many trees would it then take to offset my flight emissions alone? According to this resource a mature tree absorbs around 10 kg of CO₂ every year. So, then to offset my emissions, I’d need to plant a whopping 11165 trees. That's an entire forest! Guess I’d better start clearing my weekends if I want to make up for all those flights. Furthermore, one cannot just go around planting a single species of tree over acres of land leading to monocultures. Various restorative methods yield varying results. Further, they really start contributing to CO₂ removal after they are mature meaning, we need to grow them for years before we start seeing results.

Apart from actually contributing to NGOs that do good work, some slacktivists like me can help by using something like Ecosia which is a search engine that uses their profits to plan trees (I am in no way associated to them!). In addition, one can directly contribute to their cause to plant trees. In my case, this would cost me around 27912€. This is definitely something I hope to offset!

Cost to Offset – Can I Pay My Way Out of This?

I know carbon offsets are controversial but while writing this, I had a chance to read up a little bit about them. They seem to be some good number of high quality offset opportunities while at the same time one has low-quality cheap offset methods which are questionable at best (looking at you Ryanair!). I believe, we are beyond the point where just "avoiding deforestation" (or REDD+) is sufficient and we actually need positive action towards offsets. Of course, this market is also quite fluctuating and there isn't a fixed number to tell how much it costs and what is the quality of this. You can read up more about this here. We of course, need both technology based and nature based solutions for this as well.

All in all, this is still a fun blog post and not a serious scientific research article so I want some number to calculate, so I just ended up using the one from here and from that my flights would cost me between $ 2791 to $ 3908. That’s more than the cost of a few round-trip tickets! Sigh—I guess the price of this first world hobby also comes with a hefty environmental tax.

Final remarks

This blog started as a fun conversation with a friend about plotting flight paths in Python and ended up taking me on a philosophical journey, reckoning with my CO₂ emissions.

Some of you might wonder: won’t the numbers and images here change as I take more flights? They will! So, I made this blog dynamic (thanks to Jinja2 for my fellow nerds) and set it to update whenever I add new flights to my database.

In the end, writing this has reminded me of my privilege to travel and the not-so-small environmental cost that comes with it. Yes, it’s a bit hypocritical to say I care about the planet while booking a flight for my next adventure. But I’m doing my best to keep a lighter footprint where I can: I’ve lived vegetarian most of my life and switched to veganism, managed without a personal car so far thanks to good public transport, and try to rely on renewable energy for electricity. It doesn’t fully justify flying—I know that—but it’s a challenge I’m aware of and working to improve. Who knows, maybe someday I’ll trade air miles for virtual ones and pioneer ‘sustainable wanderlust’!

Over the past month, many of you have reached out to me about my Ironman triathlon experience. So, naturally, I decided to humble-brag and turn it into a blog series. In this series, I’ll be documenting the ups, the downs, and everything in between—covering my training (yes, I survived), my vegan nutrition, and a few coding hacks that helped me train smarter (because of course, there’s code involved).

In 2014, I watched a friend’s friend (you know the type) complete an Ironman, and it sparked a wild idea: What if I could do that too? There was just one minor hitch—I could barely swim a lap without gasping for air, didn’t own a bike (unless that rusty €30 bike counts), and could only manage a 2-3 km run, with plenty of snack breaks. But hey, why let little details like that stop me?

Back then, my goal was simple: survive the run part of a triathlon. So, I signed up for a half marathon. Naturally, the next logical step was a full marathon a year later. And to top it off, I started taking swim lessons at my university, because apparently, I love to suffer. Shoutout to my friend Aadu Huli, who unofficially took on the role of my motivator—dragging me to the pool until I upgraded my swim style from "drowning" to a somewhat decent breaststroke.

By 2017, I bought my first road bike and signed up for a sprint triathlon. That’s when I faced my first open water swim—a slightly terrifying experience where I was proud just to come out of the water still breathing. Fast forward to 2018, and I’d tackled three Olympic-distance triathlons, still holding onto my trusty breaststroke, like an old friend.

Then came the pandemic—because nothing says “pause your ambitions” like a global lockdown. Fast forward to 2023, and I found myself in a new city with fresh opportunities—and a sweet perk from work: freestyle swim coaching. Plus, joining my local triathlon club’s summer camp kicked my training into high gear. The cherry on top? A brand-new bike, which made me feel like a pro (or at least look the part). With everything in place, I decided it was time to tackle the big one—the Ironman.

Now, what kind of blog would this be without some nerdy data? I’ll be sharing my training stats (thanks to the intervals.icu API) and some slick JavaScript visualizations to track my swimming, biking, and running progress. You’ll get insights into some race prep—and maybe a few graphs showing when I hit the “why am I doing this” phase. But if I can do it, so can you—with the right support, a solid plan, and a high tolerance for discomfort.

The Most Common Question I Get: "How Often Did You Train?"

This really depends on three things:

1. My experience level (beginner with big dreams).
2. My competitive spirit (firmly at the back of the pack).
3. My training plan (Phil's Intermediate Full Distance Triathlon + Free Email Access to Coach: 36 Weeks because, why not?).

Most plans suggest 2-3 sessions per sport each week, plus some weight training. As you can see from my activity stats, biking took the spotlight—because running the full marathon distance in training is brutal. My longest run capped at 44 km, while my longest bike ride stretched to 155 km. Biking clearly became my favorite because, well, let's just say I wasn’t exactly born with wheels.

Fun fact: I ran a full marathon while on vacation in São Paulo, making my longest run a 44 km beast (I forgive Garmin for the 1.804 km error). Because who doesn’t relax by running a marathon, right?

The hardest part? Cue the dramatic music. The month leading up to the race. July 2024 was my most intense month, with over 70 hours of training! My craziest week? 21 hours of pure exercise. My most intense day? A 7.5-hour simulation: 97 miles of biking, a 2.4 km swim, and a 26 km run—all in one day. Sleep? Who needs it?

How Many Kilometers Did I Actually Cover During Training?

Let’s get into the nitty-gritty: the actual distances during training. As the graph shows, I tried (emphasis on tried) to keep up with around 10 km of swimming, 50 km of running, and 300 km of biking per month. But... vacations happened. The dip in training during November and December aligns with my trip to India for a friend’s wedding, where I did absolutely nothing. You can see the steep decline in all distances during that time. However, I redeemed myself during my second vacation in April, as shown by a similar drop in training but with a redeeming event: I ran a marathon in São Paulo and even hiked to Machu Picchu. Clearly, vacations aren’t always an excuse to slack off—they’re sometimes just another opportunity to push limits.

Training Time Breakdown: Biking Dominates, Obviously

The donut chart highlights how much time I dedicated to each sport starting October 16, 2023 (the day I officially kicked off the training plan). As expected, biking takes up the largest chunk—no surprises there. However, running wasn’t far behind, thanks to the extra weekly run I did with a good friend outside of my plan. You can also spot the gym sessions, which made up a smaller slice of the pie but were essential for foam rolling, stretching, and sometimes sneaking in a yoga pose. Swimming had its moments too, especially with that weekly hour of coached swimming from my office perk. All in all, it's clear my schedule stayed pretty packed!

In Conclusion...

This post was equal parts humblebrag and learning journey. I got to flex my JavaScript skills and hack together some Python scripts to make sense of my training data, all while documenting my Ironman training adventure. Figuring out how to use the Intervals.icu API and tweaking the data to fit my needs was a fun (and sometimes frustrating) challenge. For those of you who stuck around this long—thank you! In a future post, I’ll dive into the costs of doing an Ironman, the lessons I learned from doing it on a plant-based diet, and plenty more. Stay tuned!

Calculating a travel score for myself

I know, I know—travel is supposed to be a fun activity, enjoyed without the constraints of metrics and measurements. But there’s a part of me that loves seeing improvement in what I do, especially when it involves some numbers moving up or down towards a goal.

After a recent trip, I started wondering how I could measure my travel experiences. The most obvious metric is the number of countries visited. While this is a decent measure, I wanted a more comprehensive view of my travels.

While googling, I stumbled upon this amazing tool by Patrick Stotz. I absolutely loved it because it combined two of my interests: travel/maps and discretization. However, there was a tiny hiccup—I had to manually click boxes based on the cities I have visited. This assumes two things about me:

1. I enjoy clicking a lot.
2. I know exactly where the cities I’ve visited are located on a map.

Probably neither of these assumptions is correct. But, I do know a bit of code. With my interest in learning JavaScript, I saw this as a micro-challenge to automate the process. I recycled some old code and modified the original Travel Score calculator to accept a new JSON object with all the places I’ve traveled to. Now, not only can I calculate my travel score, but I can also keep it updated as I roam the globe. Check it out here.

The original tool by Patrick Stotz is under an MIT License.

Future Improvements

I’m excited about the potential to enhance this tool further. Here are a few ideas I have in mind to better measure “how much one has traveled the world”:

1. User Uploads: Allow others to upload their own JSON files to calculate their travel scores. (I’m still figuring out how to do this, so if anyone has tips, feel free to share!)
2. World Heritage Sites: Include extra information about UNESCO World Heritage Sites visited to add another layer of cultural significance to the score.
3. Country-Specific Scoring: Make each country’s score different. For example, visiting a rare or less-visited country would give more points than visiting a popular tourist destination.
4. Duration of Stay: Incorporate the length of time spent in each location, with longer stays contributing more to the travel score.
5. Repeat Visits: Acknowledge multiple visits to the same country or city, reflecting the depth of exploration rather than just the breadth.
6. Regional Diversity: Measure the diversity of regions visited within a country to capture a more detailed picture of one’s travel experiences.
7. Off-the-Beaten-Path Locations: Award additional points for visiting less well-known or remote locations that might be more challenging to reach.