diff --git a/presentations/introduction/index.html b/presentations/introduction/index.html new file mode 100644 index 00000000..686dae10 --- /dev/null +++ b/presentations/introduction/index.html @@ -0,0 +1,1744 @@ + + + + + + Open Climate Service — Introduction + + + + + + + + +
+
+ +
+

Open Climate Service

+

+ Climate and earth observation data is distributed across dozens of + providers — each with different APIs, formats, and access + mechanisms. +

+

+ Open Climate Service is an open-source platform that cuts through + this complexity to deliver tailored data for climate-smart + decision-making. +

+
+ + +
+ The Challenge / Why many use Google Earth Engine +

A Deeply Fragmented Landscape

+
+
+ ECMWF / ERA5Temperature, wind, humidity +
+
+ CHIRPS / CHCRainfall estimates +
+
+ NASA / MODISElevation, land use, NDVI +
+
+ WorldPopPopulation distribution +
+
+ CopernicusSatellite observations +
+
+ DestinEERA5-Land hourly +
+
+ ENACTS / NMHSsLocal station data +
+
+ Sentinel HubOptical & radar imagery +
+
+ Global Flood DBFlood extent records +
+
+ Custom sourcesSurveys, health systems, IoT +
+
+
+ Each provider has a different API, data format, coordinate system, + and access mechanism. Making this accessible to every country — + without large technical teams — is the core problem we are solving. +
+
+ + +
+ Context +

What Is a Climate Service?

+
+
+

+ A climate service turns raw climate data into + actionable information for decisions — agriculture, health, + water management, disaster risk, urban planning. +

+

+ The World Meteorological Organization (WMO) defines climate + services as "climate information developed and delivered to meet + a user's needs." Open Climate Service is designed to make this + possible for any country, regardless of technical capacity. +

+

+ Our entry point is health — but the platform is intentionally + generic. The same service can support disease modelling, food + security early warning, or national climate reporting. +

+
+
+
+
+
🏥
+

Health

+

Malaria, dengue, cholera risk

+
+
+
🌾
+

Agriculture

+

Drought, yield forecasting

+
+
+
💧
+

Water

+

Flood risk, water stress

+
+
+
🏙️
+

Urban

+

Heat stress, air quality

+
+
+
+
+
+ + +
+ 2½ Years of Learning +

What We Have Learned

+
+
+
01
+

The field is more complex than expected

+

+ Data formats, projections, temporal gaps, and provider-specific + quirks demand a flexible architecture — not a rigid pipeline. +

+
+
+
02
+

Countries want to own their data

+

+ Strong demand for local data sources, national hosting, and + independence from global platform dependencies like Google Earth + Engine. +

+
+
+
03
+

Innovation comes from the countries

+

+ The best ideas have come from HISP groups and country teams. A + bottom-up approach consistently beats top-down product design. +

+
+
+
04
+

Resolution and local data matter

+

+ National teams need higher-resolution data and the ability to + blend global datasets with local station observations. +

+
+
+
+ + +
+ The Journey — Stage 1 +

Climate App + Google Earth Engine

+
+
+ 2021 – 2023 +
+
+
ECMWF
+
CHIRPS
+
NASA
+
WorldPop
+
+
+
+
+
+
+
+ Google Earth Engine + External platform · Google-owned +
+
+
+
+
DHIS2 Climate App
+
+
+
+
DHIS2
+
+
+
+
+
+ ⚠️ +
+ Single point of dependency. Everything ran + through Google Earth Engine — a powerful platform, but not + country-owned. Access could be restricted. Customisation was + limited. Local data sources were hard to integrate. +
+
+
    +
  • Rapid to get started
    • +
  • Limited data sources (what GEE provides)
    • +
  • No offline or local deployment
    • +
  • Google controls access and pricing
    • +
  • Hard to blend local station data
    • +
+
+
+
+ + +
+ The Journey — Stage 2 +

Opening the Toolbox

+
+
+ 2023 – 2024 +
+
+
ECMWF
+
CHIRPS
+
NASA
+
WorldPop
+
+ Local data +
+
+
+
+
+
+
+
+ Google EE + Fading out +
+
+ Climate Tools + Open Python library · New! +
+
+
+
+
DHIS2 Climate App
+
+ Analysis & research +
+
+
+
+
DHIS2
+
+
+
+
+
+ 🔓 +
+ Opening the toolbox. Climate Tools gave HISP + groups and researchers direct access to the same data sources, + without GEE. Capacity building across Africa and Asia. But + still no packaged service. +
+
+
    +
  • Open-source Python library
    • +
  • Works without Google credentials
    • +
  • More data sources accessible
    • +
  • Still requires technical expertise to run
    • +
  • No standardised service layer for countries
    • +
+
+
+
+ + +
+ The Journey — Stage 3 +

Open Climate Service

+
+
+ 2024 → +
+
+
+ ECMWF +
+
+ CHIRPS +
+
+ NASA +
+
+ WorldPop +
+
+ Local data +
+
+ Custom +
+
+
+
+
+
+
+
+ Open Climate Service + Country-owned · Open-source · Extensible +
+
+
+
+
+
+
+ Climate App +
+
+ CHAP +
+
+ QGIS / Python / R +
+
+ Any tool +
+
+
+
+
DHIS2
+
+
+
+
+
+ ✅ +
+ A complete, country-owned climate data service. + Packages Climate Tools into a deployable service. Replaces the + GEE dependency in the Climate App. Integrates directly with + DHIS2 and CHAP. +
+
+
    +
  • Deploy on your own infrastructure
    • +
  • Works with any compatible tool
    • +
  • Built-in DHIS2 and CHAP integration
    • +
  • Extensible: add new data sources as plugins
    • +
  • Open standards — no vendor lock-in
    • +
+
+
+
+ + +
+ The HISP Way +

Digital Sovereignty

+

+ We are not replacing Google Earth Engine with another global service + that countries depend on. The goal is a service that is + owned and operated by the countries themselves. +

+
+
+
🏛️
+

Country-owned

+

+ Deployed on national or regional infrastructure. Countries + retain full ownership of their data and can operate the service + independently. +

+
+
+
🔓
+

Open-source

+

+ Fully open code, open standards, open data formats. No + proprietary lock-in at any layer of the stack. +

+
+
+
🧩
+

Extensible by design

+

+ HISP groups and country teams can add new data sources, custom + processing, and local integrations without modifying the core + service. +

+
+
+

+ The HISP network — the same organisations that have sustained + national DHIS2 implementations for decades — are the natural + operators of Open Climate Service instances. +

+
+ + +
+ Data & Integrations +

A Generic Climate Service for Countries

+
+
+
🌡️
+

Climate & Weather

+

+ Temperature, precipitation, humidity, wind speed, heat stress + indices. Daily to monthly aggregates. +

+
+
+
🌿
+

Environment

+

+ Land cover, vegetation index, air quality, elevation, flood + extent. +

+
+
+
👥
+

Population & Society

+

+ Population distribution by age and sex, urbanisation, + socio-economic indicators. +

+
+
+
🏠
+

Local & National Data

+

+ Station observations, ENACTS data, national surveys. Used for + bias correction and filling gaps. +

+
+
+
💉
+

Interventions

+

+ Vaccination campaigns, bed net distribution, and other programme + data that interact with climate. +

+
+
+
🔌
+

Custom sources

+

+ Any new data source can be added as a plugin — no core code + changes needed. +

+
+
+
+ + +
+ Data Quality +

The Power of Gridded Data

+
+
+
📍📍📍
+

Station data

+

+ Sparse coverage. Gaps in remote areas. Can't aggregate to + arbitrary boundaries. Requires interpolation. +

+
+
+
+ + + + + + +
+

Pre-aggregated districts

+

+ Fixed to one boundary. Uniform value per polygon. Loses spatial + variation within districts. +

+
+
+
+ + + + + + + + + + + + + + + +
+

Gridded data

+

+ Continuous coverage. Aggregate to any boundary — now or + in the future. Richer spatial patterns for modelling. +

+
+
+

+ Gridded data lets you re-aggregate to new district boundaries, + combine with facility catchment areas, or zoom into sub-district + patterns — without going back to the original data source. +

+
+ + +
+ Architecture +

Built on Open Standards

+
+
+

+ Open standards mean no vendor lock-in — data can be accessed + from QGIS, Python, R, or any other tool without custom + integration work. +

+
+
+

Zarr + GeoZarr

+

+ Cloud-native storage for multidimensional climate arrays. + Each time step is independently accessible — no need to + download entire datasets. +

+
+
+

STAC — Spatial Temporal Asset Catalog

+

+ Standard way to describe and discover geospatial datasets. + Works with any STAC-compatible catalogue browser. +

+
+
+

OGC API

+

+ Standard API for querying and accessing geographic data, + supported by major GIS tools. +

+
+
+
+
+

+ FAIR Principles +

+
+
+
F
+
Findable
+
STAC catalogue
+
+
+
A
+
Accessible
+
Open HTTP API
+
+
+
I
+
Interoperable
+
Open formats
+
+
+
R
+
Reusable
+
Open licence
+
+
+
+

Best practices built in

+

+ Climate normals and anomalies, climate indices (xclim), + WHO/WMO guidelines for aggregation and reporting. +

+
+
+
+
+ + +
+ What's Next +

Roadmap

+
+
+
+
+

Streaming ingest & resumable downloads

+

+ Large historical datasets (35+ years) ingest incrementally and + can resume after interruption. No more starting from scratch + on failure. +

+
+
+
+
+
+

Async job execution & progress reporting

+

+ Long-running ingestion and processing jobs run in the + background with live progress updates and retry on failure. +

+
+
+
+
+
+

Spatial aggregation to DHIS2 org units

+

+ Automated pipeline from gridded climate data to DHIS2 data + elements — aggregated to district or facility catchment + boundaries. +

+
+
+
+
+
+

Local data & bias correction

+

+ Integrate station data and ENACTS observations to correct + systematic biases in global reanalysis products. +

+
+
+
+
+
+

Cloud deployment & scalability

+

+ Support for S3-compatible object storage and containerised + deployment on national cloud infrastructure. +

+
+
+
+
+ + +
+

The Team

+

+ Built in close collaboration with HISP groups and country teams + across Africa and Asia. +

+
+
+
A
+

Abyot

+
+
+
+ B +
+

Bipin

+
+
+
+ Y +
+

Yambanso

+
+
+
+ D +
+

Diana

+
+
+
+ B +
+

Bjørn

+
+
+
+ Documentation + GitHub + DHIS2 Climate +
+

+ Open-source · MIT licence · Contributions welcome +

+
+
+ +
+ + + + + +