Dataharvest 2025 - the European Investigative Journalism Conference

This interactive session blends insights from professional experience and academia. Participants will walk away with a better understanding of what audiences (including you) see in data visualizations, and what you might need to do to make your message clearer.

Getting European data from EUROSTAT, the statistical office of the European Union, can be tricky, clicking through ever-increasing menus and confusing download buttons - but there is an easier way. Instead of manually downloading spreadsheets, we will learn how to use the API provided by Eurostat. Once you know how this works you will be able to adapt your knowledge to query data from other organizations including the OECD and the World Bank. We'll also learn how generative AI can help simplify querying data in this format by ensuring we use the correct syntax. Bring your own laptop: we will be using some Python: you will need a gmail account in order to access a copy of the script for the session. No other software is required.

The frequency of irregular migrant crossings, and dangers faced by thousands of people continues to attract headlines across Europe. In this session we will take a deep-dive into the best data sources to use, what they show, how they differ, their publishing schedules and their limitations. We will also look at how to access the data efficiently, to visualise it, and to combine it with other sources to produce data-driven stories. The session is for those just starting to look at this topic, or veterans looking for inspiration for new story ideas. There will be some mention of scraping in R and google sheets, but no knowledge of these is required to attend.

Today, bad data holds the power to distort reality, spread misinformation, and manipulate public opinion. This session dives deep into how inaccurate, incomplete, and even deliberately manipulated data fuels the growth of information disorder. Participants will learn how to spot and verify flawed data, understand its far-reaching impact on trust in the media, and explore practical strategies to combat its spread.

The opioid crisis and the role of Purdue Pharma might seem like a US-based story. But there is much profit being made from pain in Europe, and new addicts are being created in the process. Paper Trail Media, in collaboration with The Examination, uncovered the opioid business on this side of the Atlantic, leading to a documentary aired by ZDF in Germany, and a front page story in the Washington Post. Find out how they did it – and how the story is still being uncovered in an area where data is sparse.

Satellite imagery is a powerful tool for data-driven investigations, but getting started can be tricky. In this session, you'll learn how to navigate the U.S. Geological Survey's portal to access free satellite imagery from around the world - as well as common limitations when working with the data format. Then, using the *terra* package in R, you'll be guided through loading and analysing this data to identify urban heat islands. You will leave with practical skills, fresh inspiration and the confidence to tackle your own spatial analysis projects.

Participants who want to follow along should ideally have a working knowledge of R.

We will be using a cloud instance of RStudio accessed via your browser, so you won't need to install any software or download any files. If you want to use another language, or have RStudio installed on your computer already, then we will share any necessary data on a stick.

If you are interested the source for the data is USGS Earth Explorer portal. 

OpenAleph is an open-source investigative tool, developed and maintained by the Data and Research Center, based on the original software created by the OCCRP. In the last month, Alex and Simon (https://dataresearchcenter.org/) have developed new features for search, including making new file types searchable. OpenAleph allows users to upload large quantities of documents and makes them fully searchable in several languages. In this hands-on session, we will show attendees ways of using the search feature as power users.

Data is the beating heart of many cross-border investigations. But how do you build a data team capable of delivering groundbreaking work?
In this panel, we will discuss how to manage complex, multinational collaborations, juggle deadlines, and push back on unreasonable demands. We'll also explore best practices for leading data-driven investigations across multiple jurisdictions while navigating technical and editorial challenges.
This panel will also explore the relationship between a data editor and their team, as well as their role in embedding data journalism in newsrooms of different sizes.
This session is geared toward data journalists, editors (of all stripes), newsroom decision-makers, and cross-border organisations seeking to explore new investigative data stories or strengthen their existing offerings.

Floods are among the most frequent and damaging natural disasters in Europe, yet understanding their true impact remains difficult. As of now, there is no harmonized, official European database that systematically tracks flood damage across sectors and regions. Satellite data from the Copernicus Emergency Management Service (EMS) offers a powerful resource for addressing this gap. This session will introduce participants to the methodology developed by the Troubled Waters project to retrieve, clean, and structure Copernicus flood data from 2023-2024. It will show how to calculate flood extent and impact across agriculture, infrastructure, ecosystems, and population areas—and how to link the impacted areas to official EU statistical regions using NUTS classifications. It will also demonstrate challenges in the data, such as overlapping observations and irregular boundaries, and the approach to resolving them. Additionally, the session will explain the process of building a decade-long flood dataset combining inconsistent data from Copernicus, Hanze and EM-DAT. Participants will also receive ready-to-use code for both Python users and those without a coding background. Finally, the session will demonstrate how these insights can be transformed into a visual, scrollytelling story format that helps communicate complex findings to broader audiences.

In Serbia, BIRN's four-year investigation into public fund allocation revealed a network of phantom NGOs embezzling millions of euros, orchestrated by state and city officials. Building a comprehensive database took new journalistic effort, involving activists, IT experts, and journalists. Together, they exposed corruption, leading to award-winning stories that sparked widespread public attention and protests. How did they do it? Come and find out!

If you have not yet tried investigating in this part of the world, we will be looking at why it’s worth investigating property owners in Dubai, and how to do it. We will share tools and tricks we used in the Dubai Unlocked collaboration, and also discuss ways of identifying leads and developing a solid story about how the Emirates are used as a safe haven for criminals.