Turning the hottest day on record into an experiment


Effects of global warming that so far have been neglected, are the impacts on research projects that study global warming solutions. This blog describes the first experiment to study such effects. The preliminary outcomes are worrying. Sudoku-experiments indicate that researchers seem to perform poorly when temperatures soar, making their research less effective. Several solutions are discussed, including foreseeing unlimited amounts of ice-cream and reducing the number of managers.


GeoConnect³d is funded under the theme geo-energy. Studying geo-energy means accrediting that climate change is real. As a scientist, you don’t take any one’s word for that. Instead you download raw data and do the stats. Others may need to pick someone trustworthy for advice. If you now think of Donald Trump, your definition of trustworthy differs a bit from mine.

But that is not the topic of this blog-paper. This is about the first experiment of its kind: determining the impact of climate change on a project that deals with climate change. After all, if those researchers can’t think anymore when it gets hot, then all hope may be lost.


Take a hot day. For this experiment, we’ll use 25 July 2019, the day that the Brussels heat record from 1947 was forecasted to be broken.

Take an office building without insulation or airco. Fortunately, we have such infrastructure available. Our own offices are fit for purpose.

Look for a suited and disposable test subject. The heat warning is real, so only try this on yourself if you are at least a moderately mad scientist (easy check in my case).

Monitor the subject throughout the day. Check regularly for life signs, and measure office temperature. We’re using our in-house developed Niphargus temperature sensors with accuracy better than 0.01°C, just for show.

Test the fitness of the subject regularly by the sudoku-test. Measure the duration for solving such puzzle and register the number of mistakes and retries made. Testing material was obtained from sudoku-download.net (www.sudoku-download.net/sudoku_9x9.php).

Test subject

Apart from being geologist and male, the test subject is average in all important aspects: he is less healthy than he should be, and older than he wants. Even if his species was originally indigenous to Africa, this subject was sampled from a population that underwent significant genetic changes when adopting to the temperate and wet conditions that until recently prevailed in Belgium. A female specimen was not available for this test. The last one left on a last-minute midweek to the coolest place in Europe, which was forecasted to be Kiel in Germany (no, I’m not kidding).

Blue and left are (official) outside temperatures, red and right those inside the office. Measurements for 25 July 2019.


The official temperature in Brussels (Uccle) increased steadily to peak between 15:00 and 18:00 with a maximum temperature of 39.5°C. The office temperature increased about 2°C during working time to reach 28.887°C.

The outcome of the sudoku-test is summarized in the following table:

Sudoku ID Start End Duration Remarks
1 07:19:00 07:27:48 00:08:48 One mistake, immediately spotted
2 08:57:00 09:07:31 00:10:31 Was ready after about 7 minutes, but last number did not fit. Was able to correct, four numbers were wrong. Also, one earlier mistake that was immediately corrected.
3 10:57:00 10:04:44 00:07:44 No mistakes.
4 12:56:00 13:04:14 00:08:14 Two mistakes, immediately spotted
5 14:57:00 fail > 20 min After 6 minutes when it became evident that something was wrong. Timed-out after 20 min when already 8 mistakes were found without getting closer to solution
6 16:56:00 17:05:29 00:09:29 Slow but no mistakes.
8 18:37:00 18:43:47 00:06:47 One mistake, immediately spotted

The test subject shared following experiences as signs of life during the testing period:

Entering the office at 7:13, after a short and restless night and a morning routine that was too slow. The building was hermetically sealed, so did my morning round of opening windows and doors. After putting on the kettle, did first sudoku. Still sweaty from biking in Brussels, so my arms and hands kept sticking to the paper.

First time in years that I’m largely on top of my inbox. That’s not due to me functioning better, there is just no mail traffic. Had to close the windows before 9 o’clock, battening down the hatches so to speak. Temperature inside immediately feels much warmer, and I’m starting to feel light headed.

11 o’clock. Hand palms feel sweaty, but with help of my ‘arctic’ fan blowing on my feet and up my shorts, I’m still largely dry and functional. Belgium must be crisp dry, otherwise I’d be sweating like a pig. These sudoku’s are not very hard. They are all based on the same algorithm, so practice seems to pay off. Well, that and the energy of breakfast number four. I’ve only brought one lunch, so the afternoon will be a more hungry experience.

Around 13:00 time for another sudoku. Must have been a really easy one, because it felt really hard to keep going while the time is still ok. While my brain is feeling like it is being cooked, I do wonder whether it is really that warm. Outside it must be, Brussels is too quiet, no one is active.

3 o’clock in the afternoon. Quite terrified to see how many mistakes I was able to make in the sudoku before I started noticing them. I was clearly working slower as well. Before that I was reviewing a paper, but I think I will do the authors a favour by finishing that tomorrow instead of today.

5 o’clock update: I started tidying up the office, which is a bad sign. It is the activity that I find least useful and most frustrating. Equally boring are management tasks, which I’ve been doing for most of the day until my brain got too spongy. The few attempts at science failed, not out of lack of interesting stuff, but I found myself simply staring at a problem without making progress. I have no idea what the temperature in the office is, the sensor does not have a display. Difficult to say whether I’m complaining with reason, or because I’m a test subject that cracks easily. Outside it has to be hot. My office faces north, but even without direct sunlight the heat radiates back when I approach the windows.

18:45: Last sudoku of the day, really fast because really easy. Who rates these things as ‘difficult’? Must admit that I stopped trying to make this day useful, instead I switched to some light reading that has been piling up on my desk for months now, with some background music from the forties and fifties. That made the conditions quite a bit more bearable, dizziness and light shivering have stopped.

Discussion and conclusion

The test subject survived. However, instead of improving sudoku scores throughout the day due to increased practice, there is a significant performance drop in the tests around 15:00 and 17:00 in terms of number of mistakes or slow solving time. This corresponds to the time of outside temperatures reaching over 39°C and office temperatures approaching 29°C.

The negative correlation between temperature and scores of the sudoku-test clearly demonstrates that mental fitness degraded, up to a point where research becomes impossible or unreliable. The subject seems to be aware of this, and switched from research activities to the less demanding management tasks.

If these preliminary results are confirmed, then studying how to counter climate change is significantly hampered by global warming itself. Such feedback loop may indicate that a tipping point exists after which the impact of rising global temperatures will always outpace that of increased research efforts.

This needs to be prevented at all costs. The portfolio of climate mitigation measures may need to be extended. Feeding them ice-cream on a daily basis should be tried. During hours of low productivity, they can take over the tasks of managers or other low-effort activities, which will free up money for research.

Acknowledgement and waiver

This work forms part of the GeoERA project GeoConnect³d, which has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme under grant agreement No 731166. The maximum length of a working day at the RBINS is limited to 9 hours and working longer is considered to be a hobby. Therefore, no tax payer’s money was spent on solving sudoku puzzles. The test subject declares that apart from normal wear and tear, no lasting harm was experienced and looks forward to the next experiment if it effectively involves ice-cream.

Kris Piessens, Geological Survey of Belgium (RBINS)

Note: This blog is optimized for viewing in Chrome or Firefox. Click on images to display in full resolution.

Leave a Reply