Hey everyone, my name is Stefan Herdy, and I work at ATTO as a Data Scientist and Software Developer. I’m based at the University of Graz in Austria in the group of Bettina Weber. In my research, I typically focus on developing and applying advanced deep-learning methods for remote sensing. But I traveled to ATTO earlier this year in a slightly different role. My goal was to enhance our understanding of the forest’s health and its role in global climate regulation with cutting-edge technologies. Specifically, we’re using innovative tools to monitor the rainforest’s microclimate and atmospheric dynamics. I’m excited to share some insights from my work here.
During my time at ATTO, I set up and managed small stations deep within the forest to measure the microclimate. At these stations, we continuously collect data on temperature, humidity, and radiation from various levels of the forest. This ranges from the forest ground up to the top of the canopy. This real-time monitoring provides valuable insights into the dynamic interactions between the rainforest and the atmosphere.
Managing these systems in such a remote and challenging environment is not an easy task. Beyond the technical aspects, we also had to ensure, that our measuring stations would communicate reliably and transmit the data across vast, dense forest areas. This requires us to look for innovative solutions: the LoRaWAN (Long Range Wide Area Network) for data transmission. These allowed us to monitor these independent stations from afar without relying on more established systems that are typically either power-hungry or only work over short distances. LoRaWAN’s low-energy consumption and extended range are particularly useful in the Amazon, where the terrain and canopy can obstruct traditional wireless signals.
On the data management side, we stored and processed all collected data with InfluxDB, a time-series database well-suited for handling the massive streams of environmental data generated by the stations. In combination with Docker containers (a lightweight, self-contained environment that packages your software), we created a flexible and scalable system to deploy our software and manage our databases efficiently. They ensure that we can maintain the system remotely and update it as needed. For data analysis, we relied heavily on the programming language Python. The versatility of Python, along with its powerful scientific libraries such as NumPy, SciPy and pandas, allows us to quickly process and visualize the data.
Maintaining these systems in a remote location like ATTO can be challenging. The setup of such a measuring system is not just about technical know-how. It also requires careful planning and adaptability, especially in the unpredictable conditions of the Amazon rainforest. Whether dealing with equipment malfunctions, power outages, or environmental wear and tear, each day in the field presents unique challenges that demand creative solutions and persistence.
You can find more about my work and related software projects at https://stefanherdy.com/ or https://github.com/stefanherdy.