A multi-resolution HEALPix data structure for spherically mapped point data | JoVE Visualize

Area of Science:

Astronomy and Astrophysics
Geospatial Data Science
Computer Science

Background:

Spherically mapped data requires specialized structures for efficient storage and querying.
Hierarchical Equal Area iso-Latitude Pixelization (HEALPix) is a common data structure for spherical data, but has limitations with sparse datasets.
Existing HEALPix implementations can suffer from inefficient memory use and suboptimal query performance for sparse point data.

Purpose of the Study:

To develop and evaluate a novel multi-resolution variant of the HEALPix data structure optimized for sparse point data on a sphere.
To address the limitations of the standard HEALPix structure, including memory inefficiency and query inaccuracies for sparse data.
To improve data handling for applications involving large-scale spherical datasets.

Main Methods:

Developed a Multi-Resolution HEALPix (MRH) data structure allowing variable subdivision levels across the sphere.
Implemented MRH and compared its performance against the standard HEALPix structure.
Tested both data structures using diverse spherically mapped datasets, including cosmic microwave background data, warhead trajectories, weather stations, and galaxy locations.
Executed four types of range queries on each dataset for both HEALPix and MRH.

Main Results:

The MRH data structure demonstrated significant memory reductions, using two to four orders of magnitude less memory than HEALPix for the same sparse data.
MRH exhibited a 72% average improvement in query execution speed compared to the standard HEALPix structure.
The new structure effectively handles sparse point data, overcoming drawbacks associated with HEALPix.

Conclusions:

The Multi-Resolution HEALPix (MRH) data structure offers a superior alternative to standard HEALPix for managing sparse, spherically mapped data.
MRH provides substantial benefits in terms of memory efficiency and query performance, making it highly suitable for large-scale astronomical and geospatial applications.
This advancement enables more efficient analysis of complex spatial datasets, particularly those characterized by sparsity.