Sending a balloon to space

And can anyone hear you scream when you get there?
09 July 2019


View over the Bristol Channel, England from The Naked Scientists Space Balloon


5, 4, 3, 2, 1... Yes! We have lift off!

This is not how I normally spend an average Saturday morning. But this year, the 29th of June was different: from a farmer's field to the east of Bath, watched by a herd of sedate, cud-chewing cows, we had just sent a large helium-filled balloon and its payload soaring skywards destined for the edge of space. 

The team begin to inflate the helium balloon that will carry the screams to the edge of space

We'd spent the last couple of months planning this. It all started with an email from Tim Pilgrim at the Brunel University press office. Omar Gad, One of their graduates, he reported, was planning on sending up a helium balloon to the edge of space at over 100,000 feet (30 kilometres) above the Earth. But, Omar wondered, rather than just take pictures from up there, would the Naked Scientists like to join in and suggest some experiments that we could do at altitude?

Device constructed by the Naked Scientists to play and record screams in space

We're a radio programme and a podcast, so for this to work for us it would have to have some kind of sound or acoustic element to it, and ideally some way for the audience to become involved as well. A telecon or two later, and we hatched the plan to test the claim that "in space, no one can hear you scream". Veterans of the "Alien" era will recognise the famous strap-line used to promote the Ridley Scott blockbuster when it first came out. But what's the physics behind this claim, and is it even true? Moreover, has anyone ever actually tested this scientifically? We didn't think so, and even if they have, reproducibility is the new buzzword on the scientific block!

To do this properly, we needed to develop a system that would mechanically isolate a speaker and a microphone. If we merely stuck both into a box, vibrations from the speaker body could travel through the box material to the microphone, like the sonic equivalent of a short circuit. Knowing that local physics and engineering demo guru Dave Ansell loves a challenge like this, I called him up. His clever solution: a pyramid structure fashioned from wooden dowels joined at the ends with cable ties and with the speaker and microphone suspended from the vertices by springs and elastics. To damp the microphone even more, so that extraneous vibrations coming through the frame wouldn't move it, he embedded it in a hefty chunk of brass.

A speaker and microphone mechanically isolated from one another to enable screams to be recorded in near-space

Admittedly, the first iteration of the apparatus wasn't perfect, but then that's what testing is for. The £8 microphone we bought off the Internet made more noise of its own than it picked up from the environment, so that was a non-starter. Thankfully, a costlier £80 model behaved better and was the item that we eventually sent aloft.

The next challenge to solve was how to play and record the audio in a standardised way during the ascent while simultaneously logging the altitude, air pressure and other parameters like the temperature. Cue Omar, who agreed to write the code for the Raspberry Pi we attached to the edge of Dave's sound device. We invited listeners to send in screams that they wanted played in space. The intention was to pick the best and mix them into a single sound file containing each scream played over a range of sound levels from very quiet to very loud.

The screams would be played out and recorded back by the computer every 5 minutes during the balloon's journey; we'd log at the same moment the altitude and air pressure. Recording at a range of volume levels meant that we could be sure that at least one set of the recordings would be at the appropriate level to compensate for any background noise or over or under-sensitivity in our equipment.

A chief concern during the build phase was whether our equipment would even survive the trip. Would a Raspberry Pi cope with pressures as low as 4mBar (less than 1% of the air pressure at sea level) and temperatures that might fall to zero? Omar's previous forays into near space had shown that the temperature inside the box ought to stay above 8 degrees Celsius, so that was less of a concern. But to check that our devices - and in particular the speaker, microphone and computer - could cope with a near vacuum we put them into a vacuum chamber in Dave Ansell's garden. We were able to play and record screams, and the equipment survived, suggesting that it should work in the real environment.

Meanwhile, we also made contact with the atmospheric chemistry team led by Rod Jones at Cambridge University. Rod's group have pioneered the development of miniature battery-powered sensors capable of detecting a range of atmospheric gases. They agreed to supply one of their gadgets to enable us to monitor carbon dioxide, nitrogen oxides and ozone. To make it work was simple - we drilled a hole in the bottom of the polystyrene box that was going to house our experiment and taped Rod's sensor on top!

To launch a balloon of this type requires Civil Aviation Authority (CAA) permission; the window we were given spanned a week, which was fortunate because the weather in the days beforehand was absolutely terrible. We had plumped for June 29th because it was a weekend, worked best for everyone involved, gave us the longest time to work on solving any problems, it was Omar's birthday, and it gave us the greatest opportunity to promote the project to the wider media.

In this respect, once we began talking things up, interest began to grow rapidly. Everyone has heard the claim about not being heard screaming in space, but the fact that we actually wanted to approach this scientifically, and have fun at the same time, really struck a chord. Several BBC and other radio presenters submitted screams and yells for us to send aloft, including BBC Radio 5 live's Tony Livesey. His rendition of the way he calls his pet dogs inspired a surge of animal owners to submit their own. The stand-out winner, though, was a lady from Cape Town who submitted the sound of herself screaming at her kids to clean up their bedrooms! You can hear the "mix" by clicking play...

Route followed by the helium balloon after take-off near Bath to recovery 100 miles away near BirminghamArmed with this cacophony, Omar loaded them onto the computer and we committed ourselves to launch after lunch on Saturday 29th June. But because the prevailing wind direction is towards the east, the launch site was near Bristol in the west of England, a journey of close to 200 miles from Cambridge. Taking off from this location would see the balloon ascend, reach its peak altitude at about 33 kilometres, pop, and then the payload return to Earth by parachute over a 3 to 4 hour period. This would also, hopefully, give us time to make it to the site where the balloon would descend, over 100 miles away, near Birmingham. 

Launch-day dawned with a perfect weather forecast; in fact, it was scheduled to be super hot: over 30 degrees, which made the drive to the remote farmer's field, somewhere off the M4 motorway near Bath, relatively unpleasant. But waiting for us - among the cows - were Omar, his brother Adham, Brunel engineer and balloon veteran Konstantinos Banitsas, and Tim Pilgrim.

Space balloon payload apparatus to record screams in hear space as well as temperature, pressure, atmospheric gas composition and locationApart from roasting would-be balloon team members, high temperatures can pose a problem for balloon launches, because warmer air can limit the rate of ascent and the ultimate maximum altitude of the balloon. Nevertheless, a clear blue sky promised the possibility of stunning aerial photographs from our experiment during its journey. To make the most of this opportunity, we had cut two phone-sized recesses into the sides of the polystyrene box, and drilled a camera-lens-sized hole through the box in each. Two mobile phones set to continuous video mode, were taped into position.

One by one we also taped in and turned on the various components that we assembled: the screaming system, the Cambridge atmospheric monitoring system, a system to record temperature and pressure, and a telemetry system. This latter gear was going to be beaming back to us a constant stream of data logging the location, altitude and ambient conditions around the payload.

The next step was to inflate the balloon. Standing in the middle of the field was a large cylinder of helium; the balloon was laid on a groundsheet to prevent contact with any sharp objects, and gloved-up to protect the balloon various team members helped to make sure nothing took off too soon. Fully inflated, and with a diameter of more than a metre across, the balloon was generating about 4.5kg of lift, so hanging onto it was quite demanding. Which is why we gave the job to Naked Scientists intern Matthew Hall. Luckily he survived the ordeal, and we successfully sealed the neck with a series of cable ties and tape. In turn, these were attached to a line connected to a parachute, which would deploy and control the descent of the balloon once it had burst at maximum altitude, and beyond that was our polystyrene box payload.


Amelia and Matthew are given the job of hanging onto the inflated balloon until we were ready to launch

With everything secured, we stood back and counted down; released, the balloon shot up and within minutes was far too small to see. Avoiding the cows, and what they'd deposited on the ground around us, we packed up. But what to do with the remaining traces of helium. What a silly question to ask a bunch of scientists armed with a microphone...

But now the real work began, because we had to pursue the balloon across the country as it was pushed northeast by the prevailing winds. Armed with a giant antenna resembling a television aerial, which made Omar look a bit like a Ghost Buster sitting in the car, but the good news was that the telemetry data, and the air pressure and temperature readings were beaming down to us safely. 


The balloon reached an ultimate altitude we had planned - about 33km. At this point, the low ambient air pressure (at less than 10mBar) causes the balloon to stretch to such an extent that it bursts, triggering the payload to descend at a speed checked by the parachute. But the images and video we shot from up there are stunning.

Here we are at the edge of space, 33 kilometres up, at the moment the balloon was about to burst...

And then as we crash land into a field, not far from Birmingham...

Finding the balloon and payload again proved slightly tricky because, when it landed, the telemetry antenna buried itself in the ground, so we couldn't track the signal. Luckily, the "Find my Phone" app was running on one of the mobile phones inside the box, which enabled us to track it down to the field in which it had landed, with all of our experiments intact!


The atmospheric carbon dioxide profile we detected as we ascended towards the edge of space

We dropped the atmospheric sensor back to the Cambridge Chemistry Department, where Rod Jones's team decoded the data for us and plotted this graph, which shows the concentration of carbon dioxide on the x axis against altitude (in kilometres) on the y axis. We must caution that this is uncalibrated data, but the skew at the lower end of the graph suggests a CO2 plume at lower altitude. We suspect that this might be Bristol and the M4, which were just upwind of us at launch!

So what about the sounds we recorded? We anticipated that, with increasing altitude and hence dropping air pressure, the sounds picked up by the microphone, although they are being played at an identical volume every time, should become progressively quieter and possibly totally inapparent. This is because sound is a compression wave: as it moves, the speaker accelerates air molecules in one direction. These cannon into adjacent molecules, pushing them along; they, in turn, hit adjacent molecules and so on, and the sound energy propagates through space. When these air molecules bump into the microphone, they transfer their momentum to the microphone, applying a force. The size of that force is proportional to the loudness of the sound. So as the number of air molecules falls as the pressure drops, there are fewer molecules to transfer momentum to the microphone, so the force applied is lower. This means that the microphone moves less and detects a quieter sound. 

This is exactly what we detected.

Here is a recording close to the ground, where the air pressure is roughly 1 atmosphere:

And here is the same sound, played at the same volume setting on the speaker but at 33 kilometres where the air pressure is less than 1% of what it is at sea level:

So our prediction is that, had we continued to ascend to a place where the air was even more rarified (thin), eventually we would reach a point where there were so few air molecules being moved by the speaker and then applying a force to the microphone that the sound would become inaudible.

Hence, we can conclude that, in space, in all likelihood, you could certainly scream (at least for a short while), but no one would be able to hear you!

Thanks to the team who made this such a fun and memorable experience, including Omar Gad, Dave Ansell, Tim Pilgrim, Phil Sansom, Konstantinos Banitsas, Matthew Hall, Amelia Smith, and Rod Jones's group in Cambridge. Now we need to think up an experiment for next time!



The payload slung beneath the helium balloon containing cameras, pressure, temperature and telemetry sensors, and our screaming test system.








Hi Team

Just wondering if there is a way to get the raw data for students in data science class?

What data are you looking for, Kevin?

thank you!

Add a comment