Goose Bay Ionosphere Radar Evaluation Data

Radar data was collected by a specialised system situated in Goose Bay, Labrador, specifically designed for evaluating the ionosphere. The powerful system uses a phased array of sixteen high-frequency antennas, transmitting power on the order of 6.4 kilowatts. The signals targeted free electrons in the atmosphere. Returns are classified as "Good" ('g'), indicating evidence of structure within the ionosphere, or "Bad" ('b'), meaning the signal passed through without interaction.

The dataset features 35 columns, which include 34 float attributes and one categorical target. The attributes atr1 through atr34 represent the raw radar data derived from the complex electromagnetic signals. These numerical values result from processing the signals using an autocorrelation function, providing two values for each of the 17 pulse numbers inherent to the Goose Bay system. The attributes typically range between -1 and 1. The target attribute is categorical, labelled 'g' for good returns and 'b' for bad returns.

The dataset contains a total of 351 rows, representing individual instances of radar returns. It is provided as a CSV file (ionosphere.csv), weighing approximately 100.18 kB. Data quality is excellent, with 100% validity across all attributes and zero reported mismatched or missing values. The expected update frequency for this dataset is never.

This data is ideally suited for academic and practical applications within Atmospheric Science. Key uses include developing and benchmarking binary classification models aimed at accurately distinguishing between structured ionosphere returns and non-interacting signals. It is highly relevant for research into high-frequency signal processing and applying machine learning to geophysical data.

The data collection is geographically tied to the area monitored by the Goose Bay, Labrador radar system. The scope focuses strictly on evaluating ionospheric activity and structure.

Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0).

Machine learning engineers focused on classification problems. Researchers studying atmospheric phenomena and geophysics. Students and practitioners of complex signal processing and radar technology.

Original Data Source: Goose Bay Ionosphere Radar Evaluation Data