My expertise is in Potential Field Methods, Gamma-Ray Spectrometry and Remote Sensing. Lately, I have been studying Machine and Deep Learning to integrate large datasets, common in these methods. Each project I work on proves these methods versatile and appliable to a considerable range of problems, from volcanic activity to environmental salvage. My passion, as a researcher, is volcanic activity, which is not quite common for someone who lived most of his life in a country with no volcanic activity since the Pleistocene.
New technologies are requiring a reassessment of not only data but of many Earth Sciences questions. For example, magnetic field data has been used for decades as a secondary method for volcano monitoring, gathering only intensity data – which changes according to geometric and temperature changes in the magma chamber. However, new technologies as the Internet of Things (IoT), Low Power, Wide Area (LPWA) networking protocols, and Arduino-like microprocessor units allow us to develop magnetometer networks to monitor all vector components of the magnetic field almost in real-time. Or using drones for high-resolution surveys at minimum costs.
In 2017, I joined as an assistant professor at the University of São Paulo. Two years later, I re-founded the GEOLIT Group, renaming it to Geoscience of the Lithosphere Research Group (GEOLIT | IGc – USP) to include fellow geologists from my institute. The group now counts with 30 students of different levels, four professors, and three geophysicists from the industry. A few months later, I co-founded the Artificial Intelligence in Geosciences Laboratory (Intelli+Geo), which promoted an online Artificial Intelligence workshop that reached more than 1500 people.
Today, my research has two fronts: the geohazard and the exploratory. The geohazard front relies on developing a low-cost Arduino-based magnetometer net, including IoT technology and LPWA networking protocols to monitor changes in the magma chamber from volcanoes. Changes in the magma chamber cause subtle contrasts in the magnetic field over it. Volcanic monitoring efforts have been using this methodology for a long time, but with a single magnetometer. I intend to employ low-cost technologies to comprehend variations in volcanic magnetization throughout time. These variations include the magnetic inclination and declination of the body. Seismological and InSAR data will complement the magnetic field data. The magnetometers are in the prototype phase, using two Arduino Uno protoboards, a low-precision magnetic sensor, and two LoraWAN® antennas.
In the exploratory front, the fieldworks developed in 2019, summed up to open-source data, allowed us to conduct a large-scale investigation in the Alta Floresta Gold Province (AFGP). We gathered geophysical, geomorphological, petrophysical, petrological, geochemical and isotope data to evaluate AFGP's gold prospects. The results support the search for new targets. The large dataset is also under scrutiny using Machine and Deep Learning algorithms, enhancing the interpretation. With the whole dataset in hand, the GEOLIT group will discuss the present theory of evolution of the province, the mineralization styles and, ultimately, the potential for gold production.
I consider myself a versatile researcher. My passion is, undoubtedly, volcanism. However, becoming a researcher in a country without active volcanoes made me learn to work better in other areas without losing sight out of my passion. It also made me behave better within a group, even if out of my area of expertise. Geophysics has its physical limitations, but creativity expands its application, and there is where I like to work. No matter where I research, I will seek to continue exploring new forms to understand the dynamics of active volcanoes and prevent losses from their eruptions.