Fallen Picoballoon recovery attempt

Introduction

Wednesday, 17 August 2022, Cezar YO8TLC told us that a picoballoon (0x17) launched from France by Stefan DK3SB might have landed somewhere in northern Romania, relatively close to us. Cezar gave us all the info available and judging by the telemetry data the balloon landed in a mountainous region 65 kilometers away from our city, at an altitude of around 1000 meters. After crashing, the balloon seemed to continue to transmit telemetry data. In the following days, somewhat reluctantly, we tried locating it - without success. This is the report and how we did it.

Resources:

1. Grid Square SHP - A collection of all QTH grid squares in shapefile (SHP) format, as used by QGIS or ArcGIS
https://sourceforge.net/projects/maidenheadshpfi/
2. Digital Elevation Model - A collection of digital elevation model files (DEM) for GIS
https://earthexplorer.usgs.gov/
3. Quantum GIS - A highly capable GIS software, great for map creation and spatial data analysis - https://www.qgis.org/en/site/

All these resources are completely free and extremely useful!

Creating the maps

Once the telemetry data was analyzed we realized that the balloon crashed somewhere in KN27qp. Because of the low location accuracy reported by the system we had to take the entire grid square into consideration. Then we looked at the reported altitude, which had an accuracy of 100 meters and was reported for several days after the crash, at 998 m.

I've found the QTH grid squares in shapefile format and imported them into QGIS. In order to get the terrain information inside the KN27qp and the surrounding area I used USGS Earth Explorer from which I downloaded the Digital Elevation Model on top of which I built the maps. The DEM allowed me to extract the contours and thus get the elevation in any given point. Applying some QGIS basic functions I separated an altitude range between 950 to 1050 meters (+/- 50 meters of the balloon's reported altitude) and exported the maps in raster format, along with the KML and GPS files that would later enable us to track our movement in real time.

Listening for the WSPR

We have decided to give it a try on Saturday, almost a week after the crash. We had no equipment ready at hand and so we had to improvise. Andrei YO8SSQ came up with the idea of using a SDR dongle connected to our phones. Unfortunately, the dongle he managed to find was crashing our phones a lot and was not on the right frequency anyway, so we asked Cezar YO8TLC for help. He couldn't come along for the ride, but gave us his SDR dongle and antenna. After a few hours of scrambling about, looking for long lost equipment and buying replacements (an OTG cable), we set out for Breaza on our motorbikes.

In Breaza we stopped to scan for a signal. We connected the SDR dongle to our phones and used the SDR++ app with minimal results. The app crashed often and stopped detecting the dongle, requiring several reboots. There was a signal, but it faded in and out and we couldn't say for sure whether it was the balloon or not. Most probably not. We couldn't decode it and so we relied on the transmitting pattern of the payload (TX bursts of 110 seconds with a 10 second gap in between), but unfortunately it was too faint to pin it. On our two next stops there was total silence. The last spot might have been the better of the three, being on a hill, but by the time we reached it the sun was obscured by ominous clouds - the balloon is powered by solar energy. The approaching thunderstorm chased us off.

Even if the data is correct and the balloon indeed crashed inside the plotted area, locating it is still like looking for a needle in a haystack. The payload is not of a significant value, the balloon was not built with recovery in mind, but it would have been fun to find it nonetheless.

Thanks to YO8TLC and YO8SGB for lending the SDR equipment.