Seamless Prediction and Prediction Services

This Research Topic centers on multiscale climate predictions spanning subseasonal-to-seasonal (S2S), seasonal-to-annual (S2A), and annual-to-decadal (A2D) timeframes—termed “seamless prediction”—coupled with climate prediction service optimization, and integrates the research, development, and iteration of prediction systems. Its core mission is to eliminate the disconnect between cutting-edge scientific progress and real-world societal demands via an equitable, transparent, and two-way “research-to-operation” (R2O) framework.



Seamless prediction seeks to dismantle the traditional silos of time-scale segmented prediction, addressing cross-scale interactions that conventional approaches overlook. The S2S scale focuses on forecasting extreme weather events (e.g., heatwaves, heavy rainfall) and anomalous atmospheric circulation to support short-to-medium-term disaster preparedness. The S2A scale targets interannual climate anomalies driven by key teleconnection patterns (ENSO, AO, IOD), guiding agricultural planning and water resource management. The A2D scale prioritizes slow-changing components (e.g., ocean heat content, permafrost dynamics) critical for long-term climate resilience strategies. The research will develop integrated systems fusing multiscale physical models with machine learning algorithms, optimize multi-source observation data assimilation (satellite, in-situ, reanalysis), and strengthen cross-scale uncertainty quantification and propagation analysis.



Prediction outcomes will be translated into tailored services for governments, key industries (agriculture, energy, transportation), and the public, prioritizing information asymmetry resolution to ensure vulnerable groups and underdeveloped regions access equitable resources. The iterative R2O mechanism converts research into operational tools while integrating practical feedback for system refinement, enhancing prediction accuracy, reliability, and applicability to support climate-resilient societies and sustainable development.



Key themes for this Research Topic include:

• Development of seamless prediction systems: Design principles for integrating physical models and machine learning, multi-source data assimilation optimization, and cross-scale uncertainty handling techniques.

• Verification and comparison with observations: Standardized evaluation frameworks for multiscale predictions, performance metrics across S2S/S2A/A2D scales, and validation against long-term observational datasets.

• Examples of use of seamless predictions: Case studies including agricultural drought early warning, coastal flood risk management, energy sector climate adaptation planning—with a focus on new energy (wind and solar power) and ship navigation support.

• Equitable and transparent R2O for science-society alignment: Strategies to leverage the R2O mechanism for bridging scientific advancements and societal needs, with a focus on equity and transparency. It emphasizes prediction services’ role in advancing sustainable development, aiding community climate adaptation and disaster risk mitigation.

• Two-way interactive R2O for prediction system optimization: Operation of the bidirectional R2O mechanism to translate research findings into practical operational tools, while integrating on-the-ground feedback to refine prediction systems. Its core goal is to enhance the accuracy, reliability and applicability of multiscale climate predictions, further supporting climate-resilient societies and sustainable development.