Thomas' Learning Hub
Completefundamentalstutorial

Introduction to Cloud-Native

Foundational concepts and shifting paradigms.

Techniques Learned

Cloud ArchitectureParadigm Shifts

Tools Introduced

Next.js

Overview

Cloud-native geospatial (CNG) refers to geospatial data formats, software, and workflows designed to take advantage of cloud computing capabilities. These technologies enable efficient storage, access, and processing of geospatial data at scale, opening up new possibilities for Earth observation, climate science, and a range of other applications.

Key Concepts

What makes geospatial data "cloud-native"?

Cloud-native geospatial formats and technologies typically share these characteristics:

  1. Decomposable: Data can be broken down into smaller, independent chunks
  2. Directly accessible: Specific portions of data can be accessed without downloading entire files
  3. Cloud-optimized: Designed for distributed storage and processing in cloud environments
  4. Open standards: Based on open specifications that promote interoperability
  5. Format-aware: Tools understand the internal structure of data formats

Core technologies in the ecosystem

The cloud-native geospatial ecosystem includes several key technologies:

  • STAC (SpatioTemporal Asset Catalog): A specification for cataloging geospatial data
  • COG (Cloud Optimized GeoTIFF): A format for efficient access to geospatial raster data
  • Zarr: A format for storing and accessing n-dimensional arrays
  • GeoParquet: A columnar storage format for geospatial vector data
  • Apache Iceberg: A table format for large analytical datasets with geospatial support

Benefits of Cloud-Native Geospatial

  • Scalability: Process petabyte-scale datasets without downloading
  • Cost efficiency: Pay only for the compute and storage you need
  • Performance: Access only the specific data you need, when you need it
  • Collaboration: Share data through standardized formats and APIs
  • Democratization: Lower barriers to entry for working with geospatial data

From Fundamentals to Projects

The concepts in this tutorial (STAC, COG, Zarr, GeoParquet) are the "atoms" of modern cloud-native systems. Throughout the curriculum, you'll see how these are combined into larger projects:

  • Spatial Eval: Interactive Dashboard & Learning Module. An experiment framework for benchmarking LLM geographic knowledge against authoritative ground-truth data.
  • Scout: Live Lab & Learning Module. A "Chat-to-Map" application that uses Agentic GIS and LLMs to interact with geospatial data. NOTE: This feature is currently gated

Learning Resources

Practical Exercises

The best way to learn cloud-native geospatial is to write code. Each module in this curriculum ends with hands-on exercises using real datasets and tools like PySTAC, rasterio, xarray, and DuckDB. Use uv to manage your Python dependencies and run exercises from the shared src/exercises/ environment.

Exercise files coming soon.

Browse the Practical Implementation section below when exercises are available.

  • "Data Gravity Shapes the Architecture of Cloud Native Geospatial" by Howard Butler [Hobu, Inc]
  • "Cloud-Native Geospatial and ArcGIS" by David Wright [Esri]
  • "Minimum Viable CNG" by Brandon Liu [Protomaps]
Introduction to Cloud-Native | Cloud-Native Geospatial Tutorial