Cover letter:

My mother was a writer: since I was a child I was surrounded by books and I am passionate about telling stories. After studying anthropology, I took a master’s degree in journalism and in 2018 I started writing in the Society section of the Argentine newspaper La Nación, one of the most important newspapers in Latin America.

I am interested in the environment, human rights, health, poverty, diversity and the challenges of the digital world. I participated in three team investigations – on the recognition of Argentine soldiers who died in Malvinas/Falkland, on sexual abuse in the Catholic Church in Argentina and on the 100,000 deaths caused by Covid in Argentina – which received national and international awards.

The arrival of Covid-19 marked a before and after in my professional life. When the government decreed a strict quarantine, I asked my editor to go out immediately. I was one of the few journalists who covered the crisis on the ground, from the very beginning. Week after week I traveled around Buenos Aires, talking to doctors working on the edge, homeless families, shopkeepers who lost everything and people who died alone. I wrote powerful stories that made visible the problems caused by both the virus and the measures taken to combat it.

Those experiences and the enormous work done by La Nación’s data team (LN Data) during the pandemic taught me that incisive journalism can really make a difference, but also the importance of data today. The datification of the world raises ethical, political and privacy dilemmas that democracies must address immediately. It also opens up a gigantic field of research for journalism.

To learn more about data, in 2021 I applied for and won a competitive UK government Chevening scholarship that allowed me to travel to pursue Birmingham City University’s Masters in Data Journalism, led by Paul Bradshaw. Under his generous mentorship, I developed strong data analysis, processing and visualization skills and learned a host of new tools, from R to QGIS. I was particularly interested in OSINT, AI, maps and satellite imagery. Among all the teams that inspired me, the work of Bellingcat is worth mentioning.

In 2022, I got the Google News Initiative internship to work for eight weeks in the visuals team of the prestigious The Guardian. For me, a journalist from the “end of the world,” it was a dream come true: the opportunity to learn from world-class journalists and developers and explore new narratives and innovative ways to reach audiences. During my internship, I published my own data- and map-based analysis of the Brazilian presidential election in The Guardian and collaborated on projects such as real-time visualization of the US midterm elections, among others.

I am someone who is just starting to walk the path of data and this is the first time I am participating in the Sigma Awards: the mere fact of doing so is already very gratifying for me. I was encouraged to apply, convinced that my work humbly tries to follow the best tradition of data journalism, documenting every step so that it can serve others to explore this field

Description of portfolio:

This portfolio was done entirely by me and originally published independently (and replicated on Birmingham Eastside and Online Journalism blog). It explores the use of artificial intelligence (AI) and satellite imagery in journalism. For that I looked for topics where I could put these techniques into practice. In this sense, it works the other way around from a traditional investigation, in which first the journalistic question appears and then the best way to answer it is chosen.

The use of AI and satellite imagery is a growing field with enormous potential in journalism, but still little explored due to lack of technical expertise, necessary infrastructure and the cost of commercial satellite imagery. My goal was to develop its basic fundamentals and produce pieces that could be useful to other journalists interested in getting into the subject.

This work is part of my final project for Birmingham City University’s Masters in Data Journalism. Two seminal pieces inspired me to work with AI and satellite imagery. Texty’s pioneering “Leprosy of the land” and Earthrise Media’s wonderful “Amazon Mining Watch”, which I met during the Perugia 2022 International Journalism Congress.

During this process I faced several challenges. The main one was the limited information available on how to use AI and satellite imagery in journalistic investigations. To overcome this, I interviewed the few journalists who had worked on similar projects. They recommended using ESA satellite imagery and using machine learning instead of deep learning, as it has a lower initial difficulty.

I experimented with supervised and unsupervised learning and became familiar with satellite data processing: how to work with image rasters, plot combinations of different bands and normalize images using attributes such as gamma and brightness values. Once I was familiar with the tools, I applied them to real cases.

Thus, I investigated the impact of Chinese military maneuvers on maritime traffic in the Taiwan Strait and the violation of the Forest Law in Argentina. And I also wrote other pieces as introduction and conclusion, with possible ideas for further research on these techniques.

The idea is that the five pieces can be read independently, but also function as a whole. To serve as a general overview of what can be done using satellite imagery and AI and, above all, to demonstrate that no great technical knowledge is needed to start exploring these tools.

Project links:

https://facostarainis.github.io/ml-satellites/index.html