The mission of the Climate, Weather and Water Forum (CWWF) is to facilitate an annual dialog among scientists, engineers, students, public and private enterprises and government entities on pressing issues related to climate change, weather extremes, water availability, and sustainability. We aim to:
TBC
Accurate prediction of precipitation and its extremes at subseasonal-to-seasonal timescales remains a central challenge in meteorology, with direct implications for flood preparedness, water-resource management, agriculture, and climate-risk reduction. This talk will highlight recent advances in understanding and improving precipitation forecast skill from three complementary perspectives: physical dynamics, atmospheric energetics, and artificial intelligence. First, I will discuss how the stability of tropical and extratropical wave propagation governs subseasonal predictability, creating identifiable “opportunities” when precursor signals propagate coherently and “barriers” when they do not. Second, I will show how emerging atmospheric total-energy signals offer a physically grounded route to extending early-warning capability for extreme precipitation in a warming climate. Third, I will introduce recent progress in AI-based subseasonal forecasting systems, which are beginning to outperform conventional dynamical models for precipitation prediction and provide new tools for identifying precursor signals. Together, these studies suggest a shift from predicting precipitation directly to diagnosing the dynamical, energetic, and data-driven sources of predictability that can support next-generation forecasting and early-warning systems.
Methane (CH₄) is a powerful greenhouse gas, over 20 times more potent than CO₂ over a century. Vast amounts of CH₄ are stored under the ocean floor, slowly leaking at sites called methane or cold seeps. Yet, a critical knowledge gap exists. Scientists lack actual measurements of this global methane flux because 1) seeps in global ocean are largely unmapped — especially in the Global South, 2) unknown total area of ocean seeps which is likely much larger than estimated, and 3) unquantified methane flux from most known seeps. This undermines accurate climate predictions. Furthermore, as future natural gas extraction targets the deep-sea seabed, the impact on the unique chemosynthetic ecosystems and biodiversity in seep field remains poorly understood.
MThe UN Ocean Decade Program “Global Climate Impacts of Methane Seeps (CliMetS)” and the UN Science Decade program “Mysteries of Ocean Cold Seep Interfaces (MOCSI)” are launched to address these blind spots. These are a global, collaborative effort to find, study, and assess methane seeps and their ecosystems. Its work will have profound impacts on 1) Climate Strategies: Quantifying seabed methane flux into the water column and atmosphere is essential for accurate climate models, 2) Ecosystem Safety: The programme will map these deep-sea oases, establish a biodiversity baseline, and create conservation guidelines, and 3) Policy & Management: Findings will be translated into practical tools for ocean governance under frameworks like CBD, UNCLOS, BBNJ, and the work of the IPCC.
MIn this presentation, I will introduce these UN Decade programs and welcome world experts in climate study to join our programs to fill up the knowledge gaps in the impacts of ocean methane seeps on future climate.
The Loess Plateau of China has witnessed a remarkable greening trend due to vegetation restoration in recent decades. However, the precipitation response to greening remains unclear, and the hydrological effect of greening is controversial. Here, we revisited biophysical effects of greening on precipitation over the plateau during 2002–2015 using the state-of-the-art water vapor tracer embedded in a regional coupled model. We find that greening can promote the growing season precipitation (0.45 mm·day−1), with 15% and 85% of the precipitation increment resulting from increases in the local evapotranspiration and the water vapor inflow from outside the plateau, respectively. As a consequence, the enhanced precipitation can compensate for the terrestrial water loss driven by the increased evapotranspiration, leading to a slight increase in the water yield. This study highlights the dominant role of the nonlocal effect in precipitation responses to greening in this region.
TBC
TBC
Indian Summer Monsoon forecasts generated by three operational sub-seasonal ensemble prediction systems are calibrated against Indian Meteorological Department gridded observations using a UNET convolutional neural network (CNN). The UNET is trained using ensemble-mean precipitation hindcasts from three GCMs, to calibrate 1-degree resolution tercile probabilities of weekly rainfall at lead times of 1 to 4 weeks for the June-September season. Hindcast skill and real-time forecasts are compared against a baseline of extended logistic regression (ELR). Both the UNET and ELR are evaluated firstly for each of the three GCMs individually using the same cross-validation approach. The UNET is shown have higher Ranked Probability Skill Score (RPSS) than the ELR, although its variance is larger with negative skill at some gridpoints. An equally-weighted multi-model combination of 2-3 models improves the UNET skill further, and reduces the inter-gridbox variance, providing positive RPSS almost uniformly. Finally, we evaluate refinements to the UNET by including additional predictor variables, including univariate antecedent observed MJO and ENSO observed indices.
The Madden-Julian Oscillation (MJO) is the dominant source of subseasonal-to-seasonal (S2S) predictability and underpins early warning of extreme weather and disaster risk reduction. Traditional S2S models such as IAP-CAS v1.3 suffer from inadequate ensemble spread, which fails to fully capture atmospheric uncertainty and constrains MJO prediction capability. This study integrates the Second-Order Exact Sampling (SOES) scheme into ensemble initialization to refine perturbation generation using large historical samples at low computational cost. Sensitivity experiments based on winter MJO cases from 2019 to 2023 optimize the configuration, lifting the real-time MJO forecast skill by up to 6 days and exceeding 30 forecast days in operational predictions. The upgraded IAP-CAS system better reproduces MJO moisture structures and thermal stratification, improving convection representation and substantially advancing precipitation forecasts across China. Beyond MJO prediction, the SOES-optimized platform supports practical S2S applications, including 2026 Asian monsoon outlook and decadal-scale projection of future westerly jet variations.
Land surface processes provide an important source of weather predictability from days to weeks, because soil moisture anomalies can persistently influence the lower atmosphere. Exploiting this predictability requires accurate land surface states, but current land data assimilation systems are computationally expensive and depend heavily on expert-designed workflows. Here we present AI-Land-DA, a fully data-driven framework that integrates remote sensing observations with a neural land surface model to predict land states across diverse hydroclimatic regimes. AI-Land-DA suppresses long-term error growth and improves estimates of soil moisture and surface energy fluxes, especially at longer lead times. Compared with the operational GEFS, it substantially improves flash drought predictability, including more accurate detection of drought onset and intensification.
TBC
Assessments of extreme weather events such as droughts and floods have traditionally focused on physical severity, often overlooking human-centered impacts due to limited timely data. This study addresses this gap by developing a socially informed, data-driven framework that integrates physical and societal impacts. Using case studies of droughts in California and Texas and flooding associated with Hurricane Helene, we leverage crowdsourcing, data mining, and Internet-of Things inputs to rapidly generate datasets from social and news media. These datasets capture socialphysical interdependencies and the evolving dynamics of societal impacts. Results demonstrate that reliable insights can be extracted from inherently noisy social mediabased data, enabling more human-centric impact assessments and mitigation strategies. We further develop online diagnostic platforms that provide synthesized insights to stakeholders while enabling affected communities to report experiences and damages
TBC
TBC
TBC
Recent AI weather models such as Pangu-Weather, Aurora, GraphCast, GenCast, AIFS, FuXi, FourCastNet3, and SFNO now rival operational numerical prediction, yet no single architecture dominates across all variables and lead times, and continued scaling of any one model yields diminishing returns. We present FTAE-Weather, a deep reinforcement learning framework that coordinates these pretrained models rather than training yet another foundation model. A tactical Weight-Agent reads the atmospheric state and emits continuous, state-conditioned weights over the expert pool; a strategic Evolve-Agent uses the Weight-Agent's revealed preferences to prune redundant models and absorb newly released architectures. Across ten variables and lead times from 24 to 360 hours, FTAE-Weather lowers RMSE by up to 15% over the best individual model and 22% over equal-weight averaging, while adding fewer than 0.1% parameters. The result is a forecasting system that improves automatically as the field releases new models — with direct implications for early-warning systems and operational disaster preparedness.
Climate and weather extremes — storms, heat waves, floods, droughts, and compound events — have become the defining natural hazard challenge of the 21st century. Their growing frequency and intensity are overwhelming engineered infrastructure, disrupting global supply chains, and propagating risks across societies through teleconnections that no single country can insulate itself against. While climate change mitigation through decarbonization remains an urgent priority, even optimistic emissions trajectories leave us facing decades of increasing exposure. Climate adaptation efforts — improved infrastructure design, financial instruments, early warning systems — are essential but are constrained by limited data, deep uncertainty in future projections, and the diffuse question of who bears responsibility for action.
This talk argues that a third pillar is emerging and demands serious scientific and institutional attention: Climate Stabilization, or the deliberate modification of developing weather and climate extremes to reduce their societal impact. Rather than waiting for disasters to unfold and recovering afterward, this paradigm asks whether the physical dynamics of the atmosphere offer leverage points — windows in time and space — where strategically placed, small perturbations could redirect the trajectory of an extreme event. This is the core idea of Weather Jiu-Jitsu: exploiting the inherent instabilities and nonlinear sensitivities of atmospheric circulation to achieve large-scale redirection of an extreme using energy borrowed from the circulation itself, not brute-force external forcing. Related efforts, such as Japan's Moonshot Goal 8 program, are exploring similar scientific and technological frontiers.
The talk will address the foundational questions this agenda raises for a forecasting and Earth science community: What physical mechanisms enable or constrain atmospheric steering? How can ensemble prediction systems, adjoint methods, and emerging AI tools be harnessed to identify intervention points and compute impact outcomes with spatial specificity? What are the data and modeling gaps? How do we frame the ethical and governance dimensions as this moves from laboratory curiosity to potential operational deployment and commercial application? In the context of CWWF's themes of seamless prediction, physical modeling, and AI for Earth science, I will sketch a research roadmap integrating chaos-informed perturbation theory to AI enabled adaptive control optimization that builds on AI-accelerated impact forecasting to provide the foundation for Climate Stabilization as a rigorous scientific enterprise and, within a decade, a viable business with measurable returns to investors and societies alike.
TBC
TBC
TBC
Climate services are scientifically based information and products that support improved decision-making by enhancing understanding of how climate variability and change influence risks and impacts. As a rapidly growing field, climate services sit at the interface between scientific research and user needs, with increasing recognition of their critical role in disaster risk reduction and sustainable development.
Effective climate services can help societies anticipate and manage climate-related hazards, such as droughts, floods, and extreme weather, thereby reducing vulnerability and strengthening resilience in climate-sensitive sectors. However, ensuring that climate information is accessible, relevant, and actionable remains a significant challenge. Evidence shows that climate services are most effective when they are co-developed and co-produced with users. This talk will explore how collaborative approaches and sustained user engagement can enhance the uptake and impact of climate information, drawing on experiences from the Climate Science for Service Partnership (CSSP China). Focusing on the agricultural sector in the UK and China, it will highlight how co-produced climate services can support risk-informed decision-making, improve preparedness for climate-related hazards, and contribute to more resilient and sustainable development pathways.
Flash floods are among the most critical climate-related hazards in arid and semi-arid regions, especially in Egypt where climate variability, urban expansion, and water scarcity increase disaster risks. Although rainfall is generally limited, extreme storms have become more frequent and intense in recent decades, causing severe damage to infrastructure, communities, and economic activities. At the same time, floodwaters represent an important non-conventional water resource for a country highly dependent on the Nile River.
Egypt has adopted integrated approaches that transform flash floods from destructive hazards into valuable water resources within disaster risk reduction and sustainable development frameworks. These approaches combine structural and non-structural measures, including protection dams, storage lakes, artificial channels, hydrometeorological monitoring networks, flood forecasting and early warning systems, and updated hydrological risk assessments under changing climate conditions.
Successful rainwater harvesting experiences in South Sinai, particularly in Saint Catherine, show how low-cost mountainous lakes and groundwater recharge can support flood mitigation, groundwater sustainability, agricultural development, and community resilience. Egypt’s experience demonstrates that integrating flood management, water harvesting, scientific research, and stakeholder participation can strengthen climate resilience, water security, and sustainable development in arid regions.
TBC
TBC
Professor, Provost of HKUST, & CAE Foreign Academician, Hong Kong
Professor & CAS Foreign Academician, Tsinghua University, China
Professor, Arizona State University, United States
Professor, Arizona State University, United States
Professor, HKUST, Hong Kong
Senior Research Scientist, Columbia University, United States
Professor, Nanjing University of Information Science and Technology (NUIST), China
Professor, University of Illinois at Urbana-Champaign, United States
Professor, University of Colorado Boulder, United States
Professor, Yunnan University, China
Professor & Vice Dean of School of Atmospheric Sciences, Nanjing University, China
Professor, Institute of Atmospheric Physics-CAS, China
Director, the Water Resources Research Institute, Egypt
Senior Scientist, Met Office, United Kingdom
Water Policy Specialist, ICIMOD, Nepal
Professor, Dalian Marine University, China
Professor, Tianjin University, China
Associate Research Fellow, Lanzhou University, China
Forum Chair / Director of CCRS, HKUST
Forum Co-Chair / BNU
Forum Co-Chair / Shanghai Meteorological Service
Event Coordinator / HKUST
Secretariat / HKUST
Secretariat / HKUST
Website / HKUST
Registration / HKUST
Transportation / HKUST
Scheduling / HKUST
Scheduling / HKUST
Onsite Helper / HKUST
Onsite Helper / HKUST
Onsite Helper / BNU
Onsite Helper / BNU
Onsite Helper / BNU
