Day 1 – Introduction to Earth Observation (EO) & Cloud Computing Fundamentals#
Earth Observation and Earth Modeling Data#
Earth observation and earth modeling data are integral to understanding and predicting Earth’s dynamic systems. Remote sensing technologies, such as satellites and drones, provide vast amounts of observational data, while computational models simulate processes like climate patterns, ocean dynamics, and geological activities. Together, these datasets enable comprehensive analysis, supporting environmental monitoring, resource management, and disaster mitigation efforts.
Copernicus Missions#
Satellite |
Description |
Status |
|---|---|---|
Sentinel-1A / 1B / 1C / 1D |
C-band SAR for all-weather, day-and-night Earth observation |
1A & 1C: Active |
Sentinel-2A / 2B / 2C / 2D |
Multispectral optical imaging for land, vegetation, water, and emergency response |
2A & 2B & 2C: Active |
Sentinel-3A / 3B / 3C / 3D |
Ocean and land monitoring (sea-surface topography, sea/land temperature & color) |
3A & 3B: Active |
Sentinel-4 |
Air quality monitoring (on geostationary MTG-S satellites) |
Planned (on MTG-S1, ~2025) |
Sentinel-5 Precursor |
Atmospheric composition (ozone, NO₂, SO₂, CO, CH₄, aerosols) |
Active (launched 2017) |
Sentinel-5 |
Advanced atmospheric monitoring from polar orbit (on MetOp-SG) |
Planned (~2025) |
Sentinel-6 Michael Freilich |
Sea surface height monitoring (continuing Jason altimetry missions) |
Active (launched 2020) |
Comparison of Optical Satellites#
Satellite |
Operator/Agency |
Resolution (m) |
Swath Width (km) |
Revisit Time |
Bands/Spectral Range |
Launch Year |
Main Applications |
|---|---|---|---|---|---|---|---|
Landsat 8/9 |
NASA / USGS |
15 (Pan), 30 (MS) |
185 |
16 days (8 days combined) |
11 bands (VIS, NIR, SWIR, TIR) |
2013 / 2021 |
Land use, vegetation, agriculture, hydrology |
Sentinel-2A/B/C |
ESA (Copernicus) |
10 (VIS/NIR), 20, 60 |
290 |
5 days (combined) |
13 bands (VIS, NIR, SWIR) |
2015 / 2017 / 2024 |
Agriculture, forestry, disaster monitoring, Land use, vegetation |
PlanetScope (Dove) |
Planet Labs |
~3–5 |
24–30 |
Daily (fleet-based) |
4 bands (RGB, NIR) |
2014+ |
Agriculture, environmental monitoring, urban change |
SkySat |
Planet Labs |
0.5–0.8 |
~6–8 |
<1 day (tasked) |
Panchromatic & RGB + NIR |
2013+ |
Infrastructure, precision agriculture, security |
WorldView-3 |
Maxar Technologies |
0.31 (Pan), 1.24 (MS) |
13.1 |
<1 day (tasked) |
29 bands (Pan, MS, SWIR, CAVIS) |
2014 |
Urban planning, defense, mining, agriculture |
Pleiades-1A/1B |
Airbus |
0.5 (Pan), 2.0 (MS) |
20 |
Daily (tasked) |
5 bands (Pan + 4 MS) |
2011 / 2012 |
Mapping, emergency response, urban monitoring |
SPOT 6/7 |
Airbus |
1.5 (Pan), 6.0 (MS) |
60 |
1–2 days (tasked) |
5 bands (Pan + 4 MS) |
2012 / 2014 |
Agriculture, cartography, land use |
Kompsat-3/3A |
KARI (South Korea) |
0.5 (Pan), 2.0 (MS) |
15 |
1–3 days (tasked) |
RGB + NIR |
2012 / 2015 |
Urban mapping, environmental monitoring |
ResourceSat-2 |
ISRO (India) |
5.8–56 |
Up to 740 |
5 days |
Multiple (LISS-IV, LISS-III, AWiFS) |
2011 |
Agriculture, land use, forestry |
Gaofen-2/6/7 |
CNSA (China) |
0.8–2 (Pan), 8–16 (MS) |
Varies |
Daily (tasked) |
RGB + NIR + others |
2014+ |
Environmental monitoring, precision agriculture |
SAR Satellites#
Satellite |
Operator/Agency |
Resolution (m) |
Swath Width (km) |
Revisit Time |
Bands/Spectral Range |
Launch Year |
Main Applications |
|---|---|---|---|---|---|---|---|
Sentinel-1 1A/1C |
ESA |
5-40 m (mode dependent) |
Up to 400 km (IW mode) |
~ 6 days |
C-band |
2014 (S1A), 2023 (S1C) |
All-weather imaging, flood monitoring, land motion, maritime surveillance |
TerraSAR-X & TanDEM-X |
DLR & Airbus |
Up to 1 m (Spotlight) |
Up to 100 km |
~11 days (individually) |
X-band |
2007 / 2010 |
High-res imaging, DEM generation, land use, disaster response |
ALOS-2 |
JAXA |
~1–10 m (Spotlight) |
Up to 490 km (ScanSAR) |
14 days |
L-band (PALSAR-2) |
2014 |
Disaster monitoring, agriculture, forestry, infrastructure |
Gaofen-3 |
CNSA |
~1–10 m |
Up to 650 km |
< 4 days (global avg) |
C-band |
2016 |
Ocean monitoring, environment, disasters, maritime surveillance |
Satellite Data Access Platform Comparison#
Platform |
Provider |
Data Available |
Access Method |
Notes |
|---|---|---|---|---|
ESA / T-Systems |
Full Copernicus Sentinel missions + others |
Web, API, OGC, STAC |
Replaces SciHub, direct cloud access and streaming |
|
NASA |
MODIS, VIIRS, Landsat, ECOSTRESS, GEDI, etc. |
Web, API, Earthdata Login |
Powerful search, lots of U.S. satellite products |
|
USGS |
Landsat, Sentinel-2, NAIP, MODIS, ASTER |
Web Interface |
Longstanding archive, easy bulk download |
|
Landsat, Sentinel, MODIS, NAIP, etc. |
JavaScript & Python API |
Analysis-ready data, powerful cloud computing platform |
||
Microsoft + Partners |
Landsat, Sentinel, NAIP, STAC-compliant catalogs |
STAC API, Python SDK (pystac) |
Analysis-ready datasets in the cloud (Azure) |
|
Amazon Web Services |
Landsat, Sentinel-2, MODIS, NAIP |
S3 Access, STAC, pystac-client |
Raw cloud-optimized geospatial data (COGs, Zarr) |
|
Community-driven |
Sentinel, Landsat, MODIS |
Python SDK |
Flexible EO analysis in time-series with datacubes |
|
Planet Labs |
PlanetScope, SkySat (high-resolution, commercial) |
Web, API |
Commercial data, requires subscription (limited trial access) |
|
Descartes Labs |
Landsat, Sentinel, MODIS, and commercial imagery |
Python SDK + API |
Commercial access, focus on AI-ready analytics |
|
EU-funded consortium |
Sentinel-1, -2, MODIS, Landsat |
openEO API (Python/R/JS) |
Cloud computing platform for EO analytics |
|
Airbus |
SPOT, Pleiades, Pléiades Neo, WorldView |
Web, API |
Commercial data, requires subscription (limited trial access) |
Earth Modeling Data Access Platfrom#
Data Source |
Operator / Agency |
Access Portal / API |
Data Types / Models |
Spatial / Temporal Resolution |
Applications |
|---|---|---|---|---|---|
Copernicus Climate Data Store (CDS) |
ECMWF for Copernicus Climate Change Service (C3S) |
Reanalysis (ERA5), seasonal forecasts, climate indicators |
Global; up to 0.25° (~30 km); hourly/daily/monthly |
Climate monitoring, impact studies, model evaluation |
|
Copernicus Atmosphere Data Store (ADS) |
ECMWF for Copernicus Atmosphere Monitoring Service (CAMS) |
Atmospheric composition, greenhouse gases, aerosols, forecasts |
Global; up to 0.4° (~40 km); 3-hourly to daily |
Air quality, emissions, health impact, UV radiation |
|
NOAA Climate Data Store |
NOAA/NCEI |
Reanalysis (e.g., NCEP/NCAR), satellite, in-situ, model data |
Global; varies widely |
Climate trends, historical analysis, environmental change |
|
NASA Earthdata / GES DISC |
NASA |
Reanalysis (MERRA-2), land surface models, atmospheric data |
Global; ~0.5° to 0.125°; hourly/daily |
Climate and weather modeling, research, impact analysis |
|
CMIP (Coupled Model Intercomparison Project) |
WCRP / ESGF |
Climate projections (CMIP5, CMIP6), multi-model ensembles |
Global; ~1°; monthly/daily/hourly |
Climate change scenarios, IPCC assessments, model comparison |
Cloud Computing#
Cloud computing revolutionizes the way earth observation and modeling data are processed and analyzed. By offering scalable computational power and storage, it facilitates the handling of massive datasets and complex simulations. Cloud platforms also enable real-time data sharing and collaboration, empowering researchers and policymakers to make informed decisions efficiently and effectively.
What is Cloud Computing?#
Cloud computing is the on-demand delivery of computing resources (such as servers, storage, databases, networking, software, analytics, and intelligence) over the Internet (“the cloud”) with different pricing policies.
It allows users to:
Scale resources dynamically
Avoid upfront infrastructure costs
Access services globally and remotely
Key Concepts#
Concept |
Description |
|---|---|
Virtualization |
Abstracting physical hardware into virtual machines for flexible use |
Scalability |
Automatically scaling resources up/down based on demand |
Elasticity |
Ability to dynamically allocate and deallocate resources |
On-Demand Self-Service |
Users can provision resources without human interaction with providers |
Pay-as-you-go |
Pay only for what you use |
High Availability (HA) |
Systems designed to ensure uptime and resilience |
Multi-Tenancy |
Multiple users sharing the same infrastructure securely |
Resource Pooling |
Providers pool resources to serve multiple consumers efficiently |
Latency |
Delay in processing data; minimized with edge/cloud regions |
Cloud Deployment Models#
Model |
Description |
Example Use Case |
|---|---|---|
Public Cloud |
Services offered over the internet and shared across clients |
Hosting websites, SaaS products |
Private Cloud |
Infrastructure dedicated to a single organization |
Enterprises with strict data control |
Hybrid Cloud |
Mix of public and private, allowing data & app sharing between them |
Enterprises with both local & cloud needs |
Multi-Cloud |
Use of multiple cloud providers for redundancy or feature diversity |
Disaster recovery, cost optimization |
Cloud Service Models#
Model |
Description |
Examples |
|---|---|---|
IaaS (Infrastructure as a Service) |
Virtual machines, storage, networks |
AWS EC2, Google Compute Engine |
PaaS (Platform as a Service) |
Runtime environment, development frameworks |
Heroku, Google App Engine |
SaaS (Software as a Service) |
Fully functional applications via the web |
Gmail, Google Docs, Microsoft 365 |
FaaS (Function as a Service) / Serverless |
Execute code in response to events |
AWS Lambda, Azure Functions |
Major Cloud Providers#
Provider |
Description |
Key Services |
|---|---|---|
Amazon Web Services (AWS) |
World’s largest public cloud provider |
S3 (storage), EC2 (compute), Lambda |
Microsoft Azure |
Integrates well with Windows and enterprise services |
Azure VMs, Blob Storage, Functions |
Google Cloud Platform (GCP) |
Focuses on data analytics, ML, and open-source support |
BigQuery, Compute Engine, Cloud Run |
IBM Cloud |
Focus on AI/ML, hybrid cloud, and enterprise solutions |
Watson AI, Cloud Foundry |
Oracle Cloud |
Popular with database-intensive applications |
Oracle DBaaS, Autonomous DB |
Benefits of Cloud Computing#
Lower operational costs
Faster innovation and time-to-market
Global scale and reach
Better security and compliance
Focus on core business rather than infrastructure
Why Cloud Computing for EO?#
Earth Observation missions generate massive volumes of data — petabytes from satellites like Sentinel, Landsat, MODIS, etc. Traditional data processing on local machines becomes inefficient due to:
Large data volume (e.g., Sentinel-2 produces ~1.6 TB/day)
Need for high-performance computing (HPC) for analysis
Collaborative access across institutions and borders
Cloud computing provides the infrastructure, scalability, and accessibility required to efficiently handle and analyze EO data at scale.
Key Benefits for EO and EM#
Benefit |
Description |
|---|---|
Scalability |
Automatically scale computing power for processing large EO datasets |
Accessibility |
Global access to EO archives and tools via the internet |
Collaboration |
Share notebooks, processing pipelines, and data in real-time |
Storage Optimization |
Store large EO archives on-demand (e.g., cloud buckets) |
Integration with AI/ML |
Easily integrate EO pipelines with machine learning models |
Pay-as-you-go |
Cost-effective for large but irregular workloads |