Microsoft and NASA have taken a substantial step toward making high-frequency Earth observation data easier to use at scale: the Harmonized Landsat and Sentinel-2 (HLS) dataset is now exposed through Microsoft's Azure Planetary Computer. This groundbreaking collaboration represents one of the most significant public-private partnerships in Earth science data accessibility, bringing together NASA's satellite observation capabilities with Microsoft's cloud computing infrastructure to democratize access to critical environmental monitoring tools.

What is the Harmonized Landsat Sentinel Dataset?

The HLS project represents NASA's ambitious effort to create a seamless, analysis-ready surface reflectance product by combining observations from the Landsat 8 and 9 satellites with those from the European Space Agency's Sentinel-2A and 2B satellites. This harmonization addresses one of the biggest challenges in Earth observation: creating consistent, comparable data across different satellite systems with varying spatial resolutions, spectral bands, and orbital characteristics.

By processing data from both satellite constellations through a unified atmospheric correction and cloud masking algorithm, the HLS product delivers surface reflectance data every 2-3 days globally. This high temporal resolution is crucial for monitoring rapidly changing environmental conditions, from agricultural growth cycles to deforestation patterns and urban expansion.

Technical Specifications and Data Quality

The HLS dataset operates at two spatial resolutions: 30 meters for Landsat-based products and 10-20 meters for Sentinel-2 products. The system processes approximately 12,000 scenes daily, covering the entire Earth's land surface. Each observation includes multiple spectral bands optimized for various applications:

  • Visible bands (blue, green, red) for true-color imagery and basic land cover classification
  • Near-infrared for vegetation health assessment
  • Short-wave infrared for moisture content and geological mapping
  • Thermal infrared from Landsat for surface temperature monitoring

Data quality assurance includes comprehensive cloud and cloud shadow masking, atmospheric correction using the LEDAPS and LaSRC algorithms, and geometric correction to ensure pixel-level alignment between different satellite observations.

Azure Planetary Computer: The Cloud Infrastructure

Microsoft's Azure Planetary Computer serves as the computational backbone for this massive dataset. Built on Azure's cloud infrastructure, the platform provides:

  • Petabyte-scale storage for the entire HLS archive and future updates
  • High-performance computing resources for large-scale analysis
  • Standardized APIs for data access and processing
  • Jupyter notebook environment for interactive analysis
  • Pre-built algorithms for common geospatial workflows

The platform uses the SpatioTemporal Asset Catalog (STAC) specification for metadata organization, making it compatible with a growing ecosystem of geospatial tools and libraries. Users can access data through Python, R, or REST APIs, with built-in support for popular geospatial libraries like GDAL, rasterio, and xarray.

Real-World Applications and Use Cases

Environmental monitoring stands to benefit tremendously from this enhanced data accessibility. Climate scientists can now track deforestation rates with unprecedented temporal resolution, while agricultural experts can monitor crop health across entire growing seasons. Urban planners gain tools to study heat island effects and green space distribution, and disaster response teams can access near-real-time imagery for flood mapping and wildfire assessment.

One particularly powerful application involves combining HLS data with machine learning algorithms for automated change detection. Researchers have demonstrated the ability to automatically identify new construction, mining activities, and land use changes by analyzing the dense time series provided by the harmonized dataset.

Accessibility and Computational Requirements

While the data itself is freely available, users need Azure computing credits to process large datasets. Microsoft offers various pricing tiers, including free access for educational and research institutions through grant programs. The computational requirements vary significantly depending on the scale of analysis – from individual scene processing that can run on a laptop to continental-scale analyses requiring distributed computing resources.

For users without extensive cloud computing experience, Microsoft provides extensive documentation, tutorials, and sample code to help researchers get started with common analysis workflows. The platform's integration with popular data science tools lowers the barrier to entry for researchers who may be new to remote sensing but experienced in data analysis.

Comparison with Alternative Platforms

The Azure Planetary Computer joins other cloud-based Earth observation platforms, including Google Earth Engine and Amazon Web Services' Earth on AWS. Each platform offers distinct advantages:

  • Google Earth Engine provides a mature ecosystem with extensive pre-processing and a large user community
  • AWS Earth offers integration with Amazon's broader cloud services and machine learning tools
  • Azure Planetary Computer emphasizes open standards and interoperability with the broader geospatial ecosystem

The addition of the HLS dataset gives Azure a competitive edge in temporal resolution, particularly for applications requiring frequent observations of rapidly changing phenomena.

Future Developments and Expansion

NASA and Microsoft have outlined an ambitious roadmap for the HLS project. Planned enhancements include:

  • Integration of Landsat Next data when the next-generation satellites launch in the late 2020s
  • Expansion to include additional satellite constellations
  • Improved cloud detection and atmospheric correction algorithms
  • Enhanced machine learning-ready data products
  • Real-time data processing capabilities for near-instantaneous access to new observations

These developments will further cement the platform's position as a cornerstone of global environmental monitoring infrastructure.

Getting Started with HLS on Azure

For researchers and developers interested in exploring the HLS dataset, Microsoft provides comprehensive getting-started guides. The typical workflow involves:

  1. Creating an Azure account and accessing the Planetary Computer
  2. Exploring available datasets through the STAC API
  3. Developing analysis scripts in Python or R
  4. Scaling computations using Azure's distributed computing resources
  5. Visualizing results using built-in mapping tools or external GIS software

The platform's documentation includes numerous examples covering common use cases like vegetation monitoring, urban growth analysis, and water resource assessment.

Impact on Scientific Research and Policy

The availability of harmonized, analysis-ready satellite data at this scale represents a paradigm shift in environmental science. Researchers who previously spent months downloading and pre-processing data can now focus on analysis and interpretation. The consistent data quality and frequent updates enable new types of research questions, particularly around the pace and pattern of environmental change.

For policy makers, the improved accessibility of Earth observation data supports evidence-based decision making for climate policy, conservation planning, and sustainable development goals. The ability to monitor environmental indicators at global scales with high frequency provides unprecedented transparency for international environmental agreements and conservation initiatives.

Community Response and Early Adoption

Early users of the HLS dataset on Azure have reported significant reductions in data preparation time and increased analytical capabilities. Environmental monitoring organizations praise the platform's ability to support operational monitoring programs that require regular updates. Academic researchers appreciate the reproducibility enabled by cloud-based analysis workflows, while commercial users value the scalability for large-area assessments.

The geospatial community has particularly welcomed the platform's commitment to open standards, which facilitates integration with existing tools and workflows. The use of STAC for metadata organization means that users can leverage a growing ecosystem of compatible software and services.

As climate change accelerates and environmental monitoring becomes increasingly critical, partnerships like the NASA-Microsoft collaboration on the HLS dataset represent essential infrastructure for understanding and responding to our changing planet. The combination of NASA's scientific expertise with Microsoft's cloud capabilities creates a powerful platform that promises to accelerate environmental research and support sustainable decision-making for years to come.