Traffic Accident Analysis in Poland: Integrating Weather Data and Sociodemographic Factors

This dataset provides a comprehensive resource for analyzing road traffic accidents (RTAs) in Poland by integrating accident records with high-resolution weather data and regional sociodemographic indicators. It covers traffic incidents recorded between 2020 and 2023, sourced from the national Accident and Collision Recording System (SEWIK). A key feature of this dataset is the spatial interpolation of meteorological data. Using universal kriging with elevation drift, point-based measurements from the Institute of Meteorology and Water Management (IMGW) were transformed into continuous weather surfaces. This process attributes precise, location-specific weather conditions (temperature, humidity, and precipitation) to each accident record, overcoming the limitations of sparse weather station networks. The archive consists of three primary CSV files: incidents.csv: The main table containing detailed records for each traffic accident. Each record includes incident characteristics (e.g., location, time, type, road conditions, speed limit) linked with its interpolated weather data and corresponding yearly socioeconomic indicators for the county. weather-data.csv: Daily weather data (average temperature, average humidity, and total precipitation) aggregated at the county level for the entire study period. socioeconomic-indicators.csv: Yearly socioeconomic data from Statistics Poland (GUS) at the county, voivodeship, and national levels. Indicators include population density, road infrastructure metrics (e.g., length and quality of paved roads), and vehicle fleet composition (e.g., counts of passenger cars, trucks, buses). This dataset is intended to serve as a foundational resource for researchers, data scientists, and policymakers for a wide range of analyses, including spatiotemporal risk modeling, causal inference studies, and the development of evidence-based road safety strategies in Poland. It is released under a CC BY 4.0 license and is accompanied by the descriptive paper to ensure full reproducibility and facilitate further research into evidence-based road safety strategies.本数据集为波兰道路交通事故（road traffic accidents, RTAs）分析提供了全面的研究资源，通过整合事故记录、高分辨率气象数据与区域社会人口统计指标构建而成。数据集覆盖2020至2023年间收录的交通事件，数据源自波兰国家事故与碰撞记录系统（Accident and Collision Recording System, SEWIK）。 本数据集的核心特色在于气象数据的空间插值处理。通过采用带高程漂移的泛克里金法（universal kriging with elevation drift），将波兰气象与水管理研究所（Institute of Meteorology and Water Management, IMGW）的点状观测数据转换为连续气象表面。该流程可为每一条事故记录赋予精准匹配具体位置的气象条件（气温、湿度与降水量），有效解决了气象站点网络分布稀疏的局限性。 本数据集档案包含三个核心CSV文件： incidents.csv：主数据表，收录每条道路交通事故的详细记录。每条记录包含事故特征（如事发位置、时间、事故类型、道路状况、限速等），并关联插值得到的气象数据以及对应县域的年度社会经济指标。 weather-data.csv：全研究周期内以县域为聚合单元的每日气象数据，涵盖平均气温、平均湿度与累计降水量。 socioeconomic-indicators.csv：源自波兰中央统计局（Statistics Poland, GUS）的年度社会经济数据，覆盖县域、省域及国家级层面。相关指标包括人口密度、道路基础设施指标（如铺装道路的长度与质量）以及车辆构成（如乘用车、货车、巴士的保有量）。 本数据集旨在为研究人员、数据科学家与政策制定者提供基础研究资源，支持包括时空风险建模、因果推断研究以及波兰基于实证的道路安全策略开发在内的各类分析工作。本数据集采用CC BY 4.0协议发布，并附带配套的描述性论文，以确保研究可完全复现，助力基于实证的道路安全策略相关后续研究。