WMO Artificial Intelligence for Nowcasting Pilot Project (AINPP) Workshop
The AI for Nowcasting Pilot Project (AINPP) aims to support developing countries through international collaboration in developing and evaluating AI-based nowcasting technologies. The workshop aims to share the outcomes of regional intercomparison studies, compile user requirements from developing countries, and to discuss the operational use of AI for nowcasting.
Following the successful kickoff workshop in Guangzhou, China in November 2024, AINPP has made significant progress toward its goals. Intercomparison activities in Asia, Africa, and Latin America have been launched, and technical subgroups have been established to advance data standardization, evaluation methodology, visualization tools, and public-private engagement strategies.
The WMO AINPP workshop, hosted by the National Institute of Meteorological Sciences (NIMS) of KMA, aims to consolidate these achievements, foster knowledge exchange, and assess the mid-term strategic direction for 2025-2026, with a focus on the application of AI-based nowcasting in developing countries.
Preceding the main workshop, a two-day seminar on 22-23 September will focus on Foundation Models for Weather and Climate. This pre-workshop aims to enhance participants’ understanding of emerging foundation model techniques and explore their potential applications - from demonstrating seamless Earth system prediction from minutes to centuries to supporting early warning and consecutive actions like crisis management and risk prevention.
The provisional concept note is here.
The local information can be found here and further updated one for the local information.
The templates of documents can be downloaded from here when it needed to present in the workshop.
Workshop Agenda
(updated on 15 September)
all times in Korea Standard Time UTC+9
Day 1, Sept 22nd, 2025 | ||
Time | items | Chair |
10:00-10:30 | 1. Open Ceremony 1.1 Opening Remarks and Welcome Address 1.2 Logistics Information | Jeong Hoon Cho, NIMS/KMA |
10:30-12:30 | 2. Keynote I 2.1 Strategies for Utilizing Large AI Weather Models 2.2 Foundation model for the Earth system | Hyesook Lee, NIMS/KMA |
12:30-14:00 | Lunch break | |
14:00-17:30 | 3. Meteorology-AI Hackathon Presentations and Evaluation 3.1 Hackathon Project Results Presentations (4 Teams) 3.2 Open Evaluation 3.3 Break 3.4 Results & Awards Ceremony | Jeong Hoon Cho, NIMS/KMA |
Day 2, Sept 23th, 2025 | ||
Time | items | Chair |
10:00-12:00 | 4. Keynote II 4.1 Explainable AI for Weather and Climate 4.2 Current Status of NVIDIA’s CBottle Development 4.3 AINPP Latin America: Nowcasting Validation and Web Platform Testbed Update (Alan Calherios, Brazilian National Institute for Space Research) | TBD |
12:00-14:00 | Lunch break | |
14:00-16:00 | 5. Foundation Model for Weather and Climate 5.1 TBD (Yoo Geun Ham, SNU) 5.2 Efficient Adaptation of Weather Foundation Models (Hyunwoo J. Kim, KAIST) 5.3 Uncertainty aware prediction in weather forecasting systems 5.4 The Weather Generator Foundation Model (Christian Lessig, ECMWF, Online) | TBD |
16:00-16:30 | Tea Break | |
16:30-18:00 | 6. Panel Discussion 6.1 Reference Architecture for designing a Foundation Model for Weather and Climate 6.2 Practical Value of a Foundation Model for Weather and Climate 6.3 Challenges and Issues in Developing a Foundation Model for Weather and Climate (Haiyu Dong, Microsoft) 6.4 TBD | Seok-Woo Son, SNU |
Day 3, Sept 24th, 2025 | ||
Time | items | Chair |
9:00-9:50 | Session 1: Open ceremony 1.1 Welcome speeches - NIMS/KMA (Young Youn Park) - AINPP (David John) - WMO WIPPS (TBD) 1.2 Logistics Information | Jeong Hoon Cho, |
| Group photo | ||
9:50-10:20 | Tea Break | |
10:20-12:00 | Session 2: AI for weather forecast 2.1 TBD (Hyesook Lee, KMA) 2.2 WIPPS plan to incorporate AI technologies (Yuki Honda, WMO) 2.3 R&D on AI for nowcasting and forecasting at CEMC/CMA(Yong Wang, CMA) 2.4 AI Weather Prediction across Scales at US NSF NCAR (David John, NCAR) 2.5 AI for Weather Forecasting at Google, Progress Across Nowcasting, Medium Range and More (Shreya Agrawal, Google) | David John Gagne, NCAR |
12:00-14:00 | Lunch break | |
14:00-15:20 | Session 2: AI for weather forecast 2.6 The AI for weather forecasting in National Meteorological Center of CMA (Kanghui ZHOU, CMA) 2.7 AI Weather Operations at Microsoft (Haiyu Dong , Microsoft) 2.8 NVIDIA AI Model Developments for SR Forecasting (Jeff Adie, NVIDIA) 2.9 Global data for local tornado detection: AI-driven tornado detection using Dual-Pol Radar 2.10 Data-driven regional weather forecasting, | Wang Yong, |
15:20-15:50 | Tea Break | |
15:50-17:30 | Session 2: AI for weather forecast 2.10 AI Developments at ECMWF: AIFS and beyond (Mariana Clare, ECMWF, Online)(UTC+2) 2.11 TBD (Jaesik Choi, KAIST) 2.12 MLCast-A clean and simple API for running pre-trained ML-based nowcasting models (Aitencia Aitor, Online, UTC+2) 2.13 AI-based nowcasting in NCM Saudi Arabia (Albaraa Khayat, NCM) 2.14 AI based precipitation nowcasting from satellite (Richard Muller, DWD ,Online) | Haiyu Dong, Microsoft |
Day 4, Sept 25th, 2025 | ||
Time | Items | Chair |
9:00-11:00 | Session 3: AI based nowcasting 3.1 AI-based Nowcasting in Extreme Weather Prediction (Wai Kin Wong, HKO) 3.2 Application of AI regional model-ZHIJI 3.3 Overview of the thunderstorm nowcasting research at the German Aerospace Centre (Tobias Boelle, German Aerospace Center, Online, UTC-4) 3.4 NowAlpha: Nowcasting with Generative Bidirectional Transformers (Jaehoon Yoo, KAIST) 3.5 AI Model for Radar-Based Extreme Rainfall Nowcasting (Changhoon Song, SNU) 3.6 AI-Based Data-to-Data Translation for Generating Ground Radar-Equivalent Rainfall Estimates from GK-2A Satellite Observations | Hyesook Lee, KMA |
11:00-11:30 | Tea Break | |
11:30-12:30 | Session 4: Regional intercomparison updates 4.1 Africa Nowcasting testbed (Abhilash Singh, University of Leeds) 4.2 Asia Nowcasting intercomparison (Weiwei Wang, CMA-Shenzhen) 4.3 Latin-America Nowcasting intercomparison (Alan James Peixoto Calherios, Instituto Nacional de Pesquisa Espaciais) | Kanghui Zhou,
|
12:30-14:00 | Lunch break | |
14:00-17:00 | Session 5: Technical Tour | |
18:00-20:00 | Dinner banquet | |
Day 5, Sept 26th, 2025 | ||
Time | Items | Chair |
9:00-10:40 | Session 6: Requirements from developing countries. 6.1 The status and requirements of AI (TBD, any suggestion?) 6.2 Advancing Nowcasting with Deep Learning Techniques in West Africa (Samuel Koranteng AFFUL, Kwame Nkrumah University of Science and Technology) 6.3 The status and requirements of AI in Mexico (Raúl Aquino Santos, Universidad de Colima) 6.4 AI for Nowcasting in RA VI Europe: Opportunities, Needs and EW4ALL(Martin Benko, Czech Hydrometeorological Institute) 6.5 From Research to Operation: Experience in Deploying AI Nowcasting Model in KMA | Yuki Honda, WMO |
10:40-11:10 | Tea Break | |
11:10-12:30 | Session 7: R2O Challenges and PPE
7.1 An overview of survey results on current status and needs for AI-based nowcasting implementation (Alan James Peixoto Calherios, INPE)
7.2 Round table discussion: Q1: How can the accessibility of AI-based nowcasting products be improved for the developing countries in support of forecasting local extreme weather?
Q2: How can AI-based nowcasting products or techniques be integrated in the forecast operational process in the developing countries?
Q3: What are the major challenges and technical difficulties, and how AINPP’s outcomes or areas of future development be leveraged to address such issues?
Q4: Which areas of support can public private engagement (PPE) in AINPP be promoted to enhance the developing countries in using AI-based nowcasting products or technical development?
| Wai Kin Wong, HKO |
12:30-14:00 | Lunch break | |
14:00-17:00 | Session 8: Steering group meeting (Kanghui Zhou) 8.1 Evaluation: How to connect evaluation efforts across testbeds and unify efforts 8.2 Visualization Subgroup: What visualizations are working well and how to exchange visualization tools across testbeds 8.3 Hosting/archiving AINPP datasets and code 8.4 Academic paper and mid-term report 8.5 The implementation plan for 2026
| David John Gagne, NCAR |