Weather radar networks represent one of the most valuable but underutilized sources of atmospheric observations. With thousands of radars operating globally, we generate petabytes of data annually. Yet much of this data remains locked away in inaccessible archives, stored in formats designed for 1990s technology.
Read the full blog post at Earthmover →
Date: October 20, 2024
The Scale of the Opportunity¶
Consider the numbers:
2000+ weather radars operating globally
15+ years of continuous observations for many networks
Petabytes of archived data
Sub-kilometer resolution in space
Minutes-level temporal resolution
This data could revolutionize our understanding of:
Precipitation processes and extremes
Storm evolution and predictability
Climate change impacts on rainfall patterns
Flood forecasting and water resources
Agricultural water management
The Accessibility Problem¶
Despite this potential, radar data remains difficult to access and use:
Technical Barriers:¶
Proprietary and legacy data formats
Files scattered across multiple archives
Lack of standardized metadata
Poor documentation
No unified discovery mechanism
Infrastructure Barriers:¶
Data stored on tape or offline systems
Slow or unreliable download mechanisms
No cloud-based access
Limited compute resources co-located with data
Knowledge Barriers:¶
Steep learning curve for radar data processing
Limited training materials
Fragmented software ecosystem
Few examples of modern analysis workflows
A Vision for Open Radar Data¶
Imagine a future where:
All radar data is FAIR - Findable, Accessible, Interoperable, Reusable
Cloud-native formats enable efficient analysis without downloads
Standardized APIs provide consistent access across radar networks
Rich metadata supports discovery and quality assessment
Analysis-ready datasets reduce barriers for new users
Machine learning can leverage decades of observations
Real-time streaming enables operational applications
This isn’t just a dream - the technology exists today.
Making It Reality¶
Through projects like raw2zarr and the Open Radar Science community, we’re working to:
Convert historical archives to cloud-optimized formats
Develop standardized processing workflows
Create accessible documentation and tutorials
Build analysis-ready datasets for specific applications
Foster collaboration across institutions and countries
Read the full article on Earthmover →
The Path Forward¶
Unlocking weather radar data requires:
Investment in data infrastructure and cloud storage
Standardization of formats and metadata
Open-source tools for data access and processing
Community building across disciplines
Training for the next generation of researchers
The promise is enormous. The time to act is now.
Interested in contributing to open radar science? Check out the Open Radar Science community or explore raw2zarr for cloud-optimized radar data workflows.