Friday 20 May 2022

Launching Weather Balloons!

 I have officially made my first foray into high altitude ballooning!



With the help of Nick Janzen and other friends, I have been able to partake in two launches so far. I produced a short documentary on the first adventure for The Weather Network:



May 7th, 2022:


For both launches, we launched from Nick's property in Carbon, Alberta, outside controlled airspace. Regardless, we had NOTAMs issued for both flights. 

Inside my homemade payload box, I had Nick's APRS radio transmitter, two GoPro cameras (a GoPro Session 5 pointing upwards, and a GoPro Max facing outwards), a 25800mAh battery bank as a backup for the cameras, and a SPOT device for backup GPS tracking. We used a 3-foot parachute by Spherachutes, and attached another micro-tracker to the flight train. 

We inflated a 1200g latex meteorological balloon (from Hoskin Scientific) with H2 gas, and released the 1550g payload around 1015AM. According to local forecast soundings, the minimum temperature reached was likely -52C. We'll have actual instruments on board for future launches.

The balloon reached a maximum altitude of 104,100 feet, before landing 71km NE of the launch site near Big Valley almost 3 hours later, or 4 miles west of Caprona on RR205. Watch the documentary for details on the entire journey, which was a nail-biter at times!


Trajectory of the May 7th flight. 



It was mind-blowing to see the GoPro footage from my first flight!


May 15, 2022:


I used a different payload box for this flight, so I could use my GoPro Max for the upward shot in the hopes of capturing the moment of burst. Turns out it worked yet again, and for a second time, the camera survived the entire journey on backup power from the battery bank. Both times, the bank (along with SPOT device) had a hand warmer taped to it. 

We put the APRS tracker in a separate box this time, but it didn't work for the majority of the flight - likely due to an antenna issue. The micro-tracker worked until shortly after the balloon burst, at which time the violent swinging motion caused the webbing attaching it to its carabiner to fail. The tracker then plummeted straight to Earth and stopped working.

We launched around 945AM and the flight lasted about 2.5 hours, reaching minimum temperatures below -60C (per forecast sounding) owing to a higher tropopause.

For the second time, the SPOT device worked perfectly and led us directly to the landing location, which was in an expansive tract of native grassland about 103km ESE of Carbon, or 3 miles north of Sunnynook on the west side of the road. 

The 600g balloon (from Hoskin Scientific) towing a 1650g payload reached a stunningly high maximum altitude of 92,900 feet before bursting. 


Distance between launch and landing, with a similar trajectory. 

The cirrus gave nice texture for this launch

A quick look back at most of the flight train before the blue tracker ripped off

The weather balloon peeling at the moment of burst!



October 1, 2022


Darren's first launch!

My third HAB mission brought lots more learning, on a fun (but stressful) day with most of the original launch team and my friend Darren. 


Our mission was to capture the sunset at or above 80,000 feet, so based on all parameters, we decided on a launch time of 550PM MDT. Sunset was at 710PM, and for every 1.5km of altitude above the Earth, we get one more minute of daylight. At altitude, this was 15-20 minutes later than sunset on the ground, and we wanted about 15 more minutes of ascent after the disc of the sun disappeared below the horizon. 



We used a 1200g Totex balloon, obtained from Hoskin Scientific. My payload as is typically runs about 1kg, which includes my Spot device, a GoPro Max, a battery bank, a couple of hand warmers, and the styrofoam box with harness. We typically add some other experimental radio tracking devices to the flight train, as we did in this case, ending up with a payload weight of 1585g. One could justify using a 42" chute for this weight (per Spherachute's descent rate chart), but we opted for the 36" chute once again (as we did on May 7th). We used hydrogen as a lifting gas. 


Northeasterly upper level winds were present to the NW of an upper low, so the trajectory was actually toward Calgary. At the surface, we had a breeze on the order of 10 knots from the south, so we needed all the help we had for handling the balloon while filling. Such a breeze on a large balloon also likely gave the impression that we had amply filled the balloon to lift our counterweight off the ground, but it quickly became apparent based on a 2.65m/s ascent rate out of Carbon that we had under-filled the balloon by a large margin. We accepted that this meant a much longer flight time was likely, and we'd miss our target of documenting the sunset at high levels. Based on the habhub calculator, our balloon was forecast to burst 3h38min after launch, at 932PM. We launched at 554PM. It was then forecast to land just NE of Calgary, which made us quite nervous.


Real time, the balloon actually ended up right over NE Calgary at about 50Kft, before drifting SE and going offline. Our radio tracker had frozen and stopped transmitting, and my Spot device sent its last ping for a while at 742PM. Based on this fact, we estimated (based on a truncated descent trajectory, assuming a much faster descent) that the balloon should land somewhere near Chestermere. We headed to Chestermere BP's, thinking our balloon would burst and come back online sometime after 930PM, but upon arrival, my Spot device had sent two pings over Scenic Acres and then NE of Springbank Airport, so we assumed it was already on descent (possibly due to the radio tracker separating from the flight train, causing the ascent speed to increase, leading to an earlier burst). We jumped in the cars and drove west.


However, perplexed, the Spot device once again stopped transmitting (after 852PM and 902PM). We looked around near some private property in Springbank before resolving to give up and resume the search during daylight. Feeling dejected, I had come to the conclusion the payload sustained severe damage or landed in an area the GPS would not work. Then, we got another ping, this time near Chestermere, where we'd originally thought it would land. We blasted back across Calgary and looked at the location where the ping was, but found nothing. Then, we received more pings well to the east, so while there was some relief from knowing the Spot device was still alive, we figured there must be some kind of electromagnetic interference causing all the wildly different locations around southern Alberta. All of this was operating under the assumption the balloon must have burst by then, and that GPS devices do not transmit above 60,000 feet. 



We opted to call it for the night, and hope the GPS would sort itself out by morning, pinging a location more than once. It then occurred to us that a significant, M8.8 solar flare likely disrupted our GPS device right in the middle of the flight. It peaked at 822PM MDT, which likely explains the blackout between 742PM and 852PM as the balloon was over Calgary.



In the morning, I woke up and checked my Spot device. The last pin I had seen was south of Strathmore at 1239PM, and initially while looking at the map I saw no new pins, so my heart sank as I had believed the payload was lost. However, upon closer inspection, I noticed that there were actually a pile (over 50) pins at the spot south of Strathmore. The Spot was 3 for 3 - we had found the landing location of the payload. Thankfully, the its receiver hadn't been nuked by the solar flare - but will be considering solar activity for launches going forward, via solarham.net.


17 hours after turning on the Spot, and after sending it up in a box to temperatures near -60C for several hours, it was still functioning. Those Energizer lithiums, which I replace before every flight, are great! My camera card had filled up after 5h20min, which included 4h57min of flight time, meaning it was recording until 1051PM. Based on the clear view of the Calgary/Airdrie/Okotoks lights from well to the east, the balloon was still rising at this time, which, given a roughly constant ascent rate of 2.65m/s (159m/min), would mean the balloon was rising through 47223m AGL. I received pins at 1052PM and 1102PM, before receiving one to the southwest at 1132PM. I believe the balloon burst between 1100-1115PM due to the position of the pins relative to upper level winds, which we looked at after the fact.


It turns out that due to an unexpected ascent to such high altitudes (above the level of forecast soundings and stuck within the parameters of the calculator, that assumed the (Kaymont) balloon envelope would fail at 35km) meant the trajectory would be much different. On raw GFS data, we found that a strong westerly jet existed above 130Kft, which explains the rapid eastward motion prior to burst as it ascended to near 50,000m (164Kft). This is a ridiculously high altitude that can only be inferred and sadly not confirmed by tracker data, but it is consistent with the fact the newer Totex balloons seem to over-perform (which may also be due to under-filling). It is also consistent with the trajectory, which was amazingly (though spottily) being transmitted well above 60Kft by my Spot device. 


Given the position of the 1132PM pin, the balloon was on descent at an unknown altitude, before hooking NW on southerly/southeasterly low level winds, coming to rest at its landing location. It wasn't until 1239AM that we began receiving pings from its landing site in a field south of Strathmore. The payload was undamaged, and the backup battery bank still had 2/3 power (GoPro still full power, though 128G SD card full). This is something to keep in mind. The Spot Device works awesome (using fresh lithiums, with a hand-warmer - unknown to what extent this helps), transmits spotty locations above 60Kft, and may take a while to begin transmitting once on the ground. 


Next few flights we want to nail the sunset, and try shooting video at 4K resolution - both during the day, and at night. I had opted for 1080p/60fps because of the turbulent motion of the box in the wind, so I could slow down the footage to being smooth. It works well for this purpose in this setting - but is a poor low light performer. 


Landing site, after a flight that lasted well over 5 hours.


The view from about 50Kft, not long after sunset. Looking south toward the moon, with Belt of Venus to the E/SE, a thick layer of haze, and the city of Calgary below.