Skip to content

Development Roadmap

Jump to bottom
Gibies George edited this page Jan 28, 2026 · 8 revisions
⬅️ Back 🏠 Home Next ➡️

🗺️ AIESDA Project Roadmap

The Artificial Intelligence based Earth System Data Assimilation (AIESDA) project follows a tiered evolution strategy to integrate AI into NCMRWF's core forecasting systems.

🚀 Development Phases

Phase Milestone Background Source DA Component
Level 0 Physics Baseline Dynamical Model (Mithuna/BharatFS) Conventional JEDI
Level 1 AI Forecast Integration AI Model Background Conventional JEDI
Level 2 Native AI-DA AI Model Background AI-Specific DA
Level 3 Hybrid Earth System Dynamical Hybrid AI-Based DA

📊 Process Flow

This diagram illustrates the data transition from Level 0 to Level 3.

graph TD
    subgraph L0 [Level 0: Baseline]
    A[Dynamical Model] --> B[JEDI Conventional DA]
    B --> C[Initial AI Forecast]
    end

    subgraph L1 [Level 1: AI Background]
    C --> D[AI Background State]
    D --> E[JEDI Conventional DA]
    end

    subgraph L2 [Level 2: Native AI-DA]
    E --> F[AI Specific DA Solver]
    F --> G[Full AI-DA Loop]
    end

    subgraph L3 [Level 3: Hybrid]
    G --> H[Hybrid Physics-AI Coupling]
    end

    style L3 fill:#eef6ff,stroke:#0056b3,stroke-width:2px
Loading
⬅️ Back 🏠 Home Next ➡️

Clone this wiki locally