FI-TW: Open Dataset Links Weather to Train Delays in Finland
The newly released open dataset, FI-TW, merges Finnish railway operations with coordinated weather data from 2018 to 2024, designed to enhance the study of weather-induced train delays. This dataset fuses operational data from Finland's Digitraffic Railway Traffic Service with meteorological readings from 209 environmental monitoring sites, employing spatial-temporal alignment through Haversine distance. It features 28 engineered attributes encompassing both operational and weather-related variables, totaling around 38.5 million observations from a 5,915-kilometer rail network. Preprocessing incorporates advanced missing data management using spatial fallback algorithms. Marking a first for Finland, this dataset addresses the absence of combined weather and train data in Nordic areas. The research is detailed on arXiv under identifier 2601.16592v2, aiming to aid studies on the intricate relationships between operational, technical, and environmental influences on train delays.
Key facts
- FI-TW is the first publicly available dataset combining Finnish railway operations with synchronized weather data.
- The dataset covers the period from 2018 to 2024.
- Operational data comes from Finland's Digitraffic Railway Traffic Service.
- Weather data is sourced from 209 environmental monitoring stations.
- Spatial-temporal alignment is performed using Haversine distance.
- The dataset includes 28 engineered features and approximately 38.5 million observations.
- The rail network covered spans 5,915 kilometers in Finland.
- Preprocessing includes spatial fallback algorithms for missing data.
Entities
Institutions
- Finland Digitraffic Railway Traffic Service
- arXiv
Locations
- Finland