Abstract The Madden–Julian Oscillation (MJO) is the dominant mode of intraseasonal variability in the tropical atmosphere and a central source of subseasonal-to-seasonal (S2S) predictability. Despite decades of progress, theoretical understanding of its forecast horizon remains elusive and contested. This workshop-inspired review synthesizes insights on MJO predictability, highlighting advances in theory, modeling, and machine learning applications, while critically assessing persistent challenges. We revisit the layered concepts of intrinsic, potential, and practical predictability, with intrinsic predictability set by the flow itself, potential predictability defined by idealized initial conditions, and practical predictability constrained by real observations, data availability, and forecasting methodologies. We examine the physical sources and statistical approaches underlying prediction skill, and evaluate current metrics and verification methodologies, emphasizing that evaluation metrics must be responsive to user needs and that observing systems should be designed in concert with model requirements. We then assess limits imposed by physical processes such as MJO initiation and the Maritime Continent propagation barrier, as well as their potentially unclear modification under climate change. Finally, we outline future research opportunities—spanning observations, hybrid modeling, and user-centered forecasting—and argue for a more integrative prediction framework in which dynamical, statistical, hybrid, and machine-learning approaches are treated not as competitors but as complementary components of a unified MJO forecasting system, while identifying open questions that will shape the next decade of MJO research.
Abstract Supporting decision-making in hydrological forecasting depends on aligning scientific outputs with the practical needs and timelines of operational centers. This study documents the collaborative development of decision calendars with forecasters from the Alaska-Pacific River Forecast Center and the Yukon Government River Forecast Center in the transboundary, high-latitude Yukon River basin, where hydrology is dominated by cold region processes. Decision calendars were used as a tool to synthesize shared operational knowledge; map recurring forecasting modes to an annual cycle; and identify critical timing for monitoring, forecasting, and communication. The process highlighted challenges in transitioning research advances into operational practice, including reconciling data sources, communication needs, and the integration of new models and methods. Decision calendars inform the development of open-source forecasting workflows using the existing operational software framework. They also provide a basis for developing user-informed evaluation that assesses forecast performance based on operational needs. Findings emphasize the importance of knowledge sharing and structured methods to exchange information in developing actionable hydrological science. Lessons learned from the Yukon River basin may be transferable to other regions seeking to improve forecast usability and uptake through structured collaboration between researchers and operational forecasters. Significance Statement Turning new scientific ideas into practical tools for river forecasters is a challenge. In this study, the Alaska-Pacific and Yukon River Forecast Centers worked with researchers in a collaborative process to create decision calendars. This exchange of knowledge mapped out practical forecasting needs to better direct research toward the realities of forecasters’ work. This approach and case study offer a model for improving the transition from research to operations in river forecasting, particularly in cold regions.
Abstract To enable informed decision-making, effective weather risk communication requires visualizations that make scientifically accurate information understandable and that address the public’s information needs. However, many current communication products struggle to achieve this goal. Using participatory design workshops and prototype development, we engaged vulnerable communities in South Florida to reimagine how NOAA’s Hurricane Threats and Impacts (HTI) graphics communicate hurricane risk information. These sessions revealed information needs and visualization preferences of those directly impacted by hurricane threats. These critical, community-driven insights, combined with established visualization best practices, can be used to inform the development of new designs based on HTI graphics. Key design improvements focused on the representation of threat levels, impact communication, and spatial orientation. Specific modifications addressed product layout, color encoding, accessibility, and language barriers. The resulting prototypes illustrate how workshop findings translate into specific design decisions for hurricane impact risk graphics. This workshop-based approach can be adapted to support the design of community-informed risk communications for hurricane and potentially other weather hazards.
Abstract This study expands upon previous efforts to more fully explore the impact of NOAA’s high-altitude Gulfstream IV-SP (G-IV) on tropical cyclone (TC) forecasts. We employ a multi-model approach that uses four different releases of NOAA’s Hurricane Analysis and Forecast System (HAFS) to better understand the sensitivity to model configuration choices and yearly version upgrades. This approach is fairly unique in literature, as almost all similar evaluations have involved a single modeling release. Overall, the results suggest that the G-IV is not being optimally used. Its data marginally improves TC track forecasts, mainly for tropical storms, while impacts for other variables are mainly neutral. This stands in stark contrast to recent work that shows that the entire suite of reconnaissance data available from all platforms substantially improves both track and intensity forecasts. In the context of prior peer-reviewed literature, the results here suggest that the G-IV should be brought radially inward and focus more on the TC vortex to the extent possible. NHC plans to test these modified flight patterns in collaboration with NOAA’s Advancing the Prediction of Hurricanes Experiment (APHEX) starting in the 2026 hurricane season. We also discuss the implications of our results and highlight the value of systematic evaluations of reconnaissance-data impacts for guiding model use and resource allocation.
Abstract The third Wind Forecast Improvement Project (WFIP3), a U.S. Department of Energy and National Oceanic and Atmospheric Administration sponsored investigation, sought to improve understanding of the physical phenomena in the atmosphere and ocean that dictate the structure and variability of wind and thermodynamic fields within the marine atmospheric boundary layer. WFIP3 focused on mesoscale and submesoscale flows – including sea breezes, low-level jets, low-level clouds, and coastal storms – and the ability of advanced numerical model parameterizations to represent them within fully coupled oceanic and atmospheric modeling systems and foundational weather forecast models. WFIP3 conducted a comprehensive 18-month observational study over the Northeast U.S. outer continental shelf, a high-use coastal zone, using a 3D multiscale sensor array to highly resolve the temporal, vertical, and horizontal structure of the coupled atmospheric and oceanic boundary layers. Multiple land-based study sites adjacent to the coastal ocean observed surface meteorology and vertical profiles of atmospheric properties via passive infrared and microwave radiometers, active lidars and radars, and radiosondes. At sea, an array of surface flux buoys and two vertical profiling lidar buoys observed both atmospheric and oceanic properties, augmented by land-based oceanographic radar systems and routine ship-based surveys. Intensive observations of the marine atmospheric boundary layer over the ocean was done from an air-sea interaction flux tower and extended deployments of a large autonomous barge platform. Numerous critical forecasting phenomena were observed that are being evaluated within regional coupled and uncoupled modeling systems, including the National Oceanic and Atmospheric Administration’s foundational High-Resolution Rapid Refresh forecast model.
