Welcome to our newsletter on everything related to impact data.
In this newsletter we will cover:
- Interesting variable: Energy Dependency
- Interesting tool: Google Earth AI
- Interesting read: Itiner-E
- Interesting dataset: Lumiere Cinema Admission
Interesting variable: Energy Dependency
Following recent geopolitical crises, European energy dependency is a critical variable to watch. In 2023, the EU’s average dependency rate was 58.3%, though this varied dramatically from Malta (97.5%) to Estonia (3.4%). The most significant story in the Eurostat data is the radical transformation of import origins: Russia, which was the top supplier of solid fuels (52.4%), natural gas (44.0%), and crude oil (25.3%) in 2021, was no longer a top supplier for solid fuels or crude oil by 2023, with the US, Norway, and Australia taking the lead. This high dependency is being partly counteracted by a steady rise in renewables, which now account for 24.6% of the EU’s gross final energy consumption (up from 16.7% a decade earlier), with Sweden leading at 66.4%.
Interesting tool: Google Earth AI
A powerful new platform I’m watching closely is Google Earth AI. It’s an analytical tool built on Google’s expertise of global modeling (think of Google Maps), now supercharged with Gemini’s advanced reasoning capabilities, designed to turn planetary information into actionable intelligence.
The core component is its Geospatial Reasoning framework. This allows users to ask complex questions in natural language, and the AI will automatically connect various models like weather forecasts, population maps, and satellite imagery to find the answer. This ability can provide breakthroughs in minutes that previously might have required years of research.
This is especially relevant for impact data, as GEAI is already being used by nonprofits and cities for critical work:
- Disaster Response: The nonprofit GiveDirectly uses GEAI’s flood and population data to determine who needs anticipatory cash assistance before a flood hits.
- Public Health: The WHO Regional Office for Africa uses the population and environment models to predict which areas in the Democratic Republic of Congo are at high risk for cholera outbreaks.
- Early Warnings: Its Flood Hub model provides life-saving riverine flood warnings up to seven days in advance, covering over two billion people.
It is already powering features across Google Earth, Maps, and Cloud, but its potential as a standalone tool for environmental monitoring and disaster response is a significant development to follow.
Interesting read: Itiner-e, The Digital Atlas of Roman Roads
For anyone who loves historical data, geospatial analysis or just Roman history, check out Itiner-e: The Digital Atlas of Ancient Roads. Its goal is to host the most detailed open digital dataset of roads spanning the entire Roman Empire.
What makes this project so compelling is not just the map itself, but the data-driven approach behind it. It’s a historical Geographic Information System (GIS) built by a scholarly community. The team digitizes the exact position of road segments by combining existing scholarship (like atlases and excavation reports) with modern aerial and satellite photography.
They categorize the reliability of each segment:
- Certain: Well-documented with high spatial accuracy.
- Conjectured: Based on less documentation.
- Hypothetical: Speculated roads with insufficient evidence.
The most engaging part is the route-finding tool (currently in beta). It allows you to calculate the distance and, more importantly, the estimated travel time between ancient places. The model incorporates a ‘passability’ multiplier for the landscape, even using Tobler’s hiking function to slow travel speeds on slopes above 6%
You can calculate your journey based on four terrestrial transport modes:
- Walking: 4 km/h
- Ox cart: 2 km/h
- Pack animal: 4.5 km/h
- Horse courier: 6 km/h
Future plans to include major rivers and sea lanes will make this an even more powerful tool for simulating historical movement and understanding the true interconnectedness of the ancient world.
Interesting dataset: Lumiere Cinema Admissions
An interesting resource for media or economic analysis is the LUMIERE dataset, maintained by the European Audiovisual Observatory. It is a systematic, title-by-title database of annual cinema admissions for films released in European cinemas since 1996. The dataset is comprehensive, covering 49 markets (35 in Europe and 14 non-European), allowing for detailed tracking of film performance, audience engagement, and market dynamics across different regions and genres.
Access to the LUMIERE database is provided as a free, public-facing online service rather than as a single downloadable file. Users can query the data directly on the Observatory’s website using a search interface to find information by film title, director, or year.
