Explore the vast range of titles available.

In this paper, we performed an analysis of the 500 most relevant scientific articles published since 2018, concerning machine learning methods in the field of climate and numerical weather prediction using the Google Scholar search engine. The most common topics of interest in the abstracts were identified, and some of them examined in detail: in numerical weather prediction research—photovoltaic and wind energy, atmospheric physics and processes; in climate research—parametrizations, extreme events, and climate change. With the created database, it was also possible to extract the most commonly examined meteorological fields (wind, precipitation, temperature, pressure, and radiation), methods (Deep Learning, Random Forest, Artificial Neural Networks, Support Vector Machine, and XGBoost), and countries (China, USA, Australia, India, and Germany) in these topics. Performing critical reviews of the literature, authors are trying to predict the future research direction of these fields, with the main conclusion being that machine learning methods will be a key feature in future weather forecasting.

NOAA’s National Centers for Environmental Prediction (NCEP) Production Suite (NPS) currently includes over 20 operational weather forecast systems, providing forecasts from the mesoscale to global seasonal outlooks. In an effort to optimize resources, the NPS is being simplified to far fewer systems within the Unified Forecast System (UFS) framework that nevertheless span NOAA’s weather prediction mission: short-range regional and atmospheric composition (RRFS, WoF), medium-range subseasonal (GEFS) to seasonal (SFS), marine and coastal (GFS, GEFS, NWPS, GLWU), hurricanes (HAFS), on-demand atmospheric dispersion (HySPLIT), hydrology (NWM), and space weather (WAM/IPE) (see appendix for a full list of abbreviation definitions). An international constellation of low-Earth-orbit and geosynchronous-equatorial-orbit satellites are expanding our Earth intelligence; for example, polar-orbiting satellites now provide 85% of the data used in global weather models. In addition to the advancement of DA research, the Joint Effort for Data assimilation Integration (JEDI), operated by the UCAR Joint Center for Satellite Data Assimilation (JCSDA), provides a common software infrastructure for full community engagement in the testing, research, and development of new observations and DA methods.

Promising new opportunities to apply artificial intelligence (AI) to the Earth and environmental sciences are identified, informed by an overview of current efforts in the community. Community input was collected at the first National Oceanic and Atmospheric Administration (NOAA) workshop on “Leveraging AI in the Exploitation of Satellite Earth Observations and Numerical Weather Prediction” held in April 2019. This workshop brought together over 400 scientists, program managers, and leaders from the public, academic, and private sectors in order to enable experts involved in the development and adaptation of AI tools and applications to meet and exchange experiences with NOAA experts. Paths are described to actualize the potential of AI to better exploit the massive volumes of environmental data from satellite and in situ sources that are critical for numerical weather prediction (NWP) and other Earth and environmental science applications. The main lessons communicated from community input via active workshop discussions and polling are reported. Finally, recommendations are presented for both scientists and decision-makers to address some of the challenges facing the adoption of AI across all Earth science.

This study explores an analog-based method to generate an ensemble [analog ensemble (AnEn)] in which the probability distribution of the future state of the atmosphere is estimated with a set of past observations that correspond to the best analogs of a deterministic numerical weather prediction (NWP). An analog for a given location and forecast lead time is defined as a past prediction, from the same model, that has similar values for selected features of the current model forecast. The AnEn is evaluated for 0–48-h probabilistic predictions of 10-m wind speed and 2-m temperature over the contiguous United States and against observations provided by 550 surface stations, over the 23 April–31 July 2011 period. The AnEn is generated from the Environment Canada (EC) deterministic Global Environmental Multiscale (GEM) model and a 12–15-month-long training period of forecasts and observations. The skill and value of AnEn predictions are compared with forecasts from a state-of-the-science NWP ensemble system, the 21-member Regional Ensemble Prediction System (REPS). The AnEn exhibits high statistical consistency and reliability and the ability to capture the flow-dependent behavior of errors, and it has equal or superior skill and value compared to forecasts generated via logistic regression (LR) applied to both the deterministic GEM (as in AnEn) and REPS [ensemble model output statistics (EMOS)]. The real-time computational cost of AnEn and LR is lower than EMOS.

Africa is poised for a revolution in the quality and relevance of weather predictions, with potential for great benefits in terms of human and economic security. This revolution will be driven by recent international progress in nowcasting, numerical weather prediction, theoretical tropical dynamics, and forecast communication, but will depend on suitable scientific investment being made. The commercial sector has recognized this opportunity and new forecast products are being made available to African stakeholders. At this time, it is vital that robust scientific methods are used to develop and evaluate the new generation of forecasts. The Global Challenges Research Fund (GCRF) African Science for Weather Information and Forecasting Techniques (SWIFT) project represents an international effort to advance scientific solutions across the fields of nowcasting, synoptic and short-range severe weather prediction, subseasonal-to-seasonal (S2S) prediction, user engagement, and forecast evaluation. This paper describes the opportunities facing African meteorology and the ways in which SWIFT is meeting those opportunities and identifying priority next steps. Delivery and maintenance of weather forecasting systems exploiting these new solutions requires a trained body of scientists with skills in research and training, modeling and operational prediction, and communications and leadership. By supporting partnerships between academia and operational agencies in four African partner countries, the SWIFT project is helping to build capacity and capability in African forecasting science. A highlight of SWIFT is the coordination of three weather forecasting “Testbeds”—the first of their kind in Africa—which have been used to bring new evaluation tools, research insights, user perspectives, and communications pathways into a semioperational forecasting environment.

Despite having the largest associated research community and a rapidly growing private sector, the lack of a well-coordinated national research and development effort for U.S. numerical weather prediction continues to impede our ability to utilize more of the scientific and technical capacity of the nation more efficiently. Over the last few years, considerable progress has been made toward developing a community-friendly Unified Forecast System (UFS) by embracing an open innovation approach that is mutually beneficial to the public, private, and academic sectors. Once fully implemented, the UFS has the potential to catalyze a significant increase in the efficacy of our nation’s weather, water, and climate science and prediction.

Generating accurate weather forecasts of planetary boundary layer (PBL) properties is challenging in many geographical regions, oftentimes due to complex topography or horizontal variability in, for example, land characteristics. While recent advances in high-performance computing platforms have led to an increase in the spatial resolution of numerical weather prediction (NWP) models, the horizontal gridcell spacing (Δ x ) of many regional-scale NWP models currently fall within or are beginning to approach the gray zone (i.e., Δ x ≈ 100–1000 m). At these gridcell spacings, three-dimensional (3D) effects are important, as the most energetic turbulent eddies are neither fully parameterized (as in traditional mesoscale simulations) nor fully resolved [as in traditional large-eddy simulations (LES)]. In light of this modeling challenge, we have implemented a 3D PBL parameterization for high-resolution mesoscale simulations using the Weather Research and Forecasting Model. The PBL scheme, which is based on the algebraic model developed by Mellor and Yamada, accounts for the 3D effects of turbulence by calculating explicitly the momentum, heat, and moisture flux divergences in addition to the turbulent kinetic energy. In this study, we present results from idealized simulations in the gray zone that illustrate the benefit of using a fully consistent turbulence closure framework under convective conditions. While the 3D PBL scheme reproduces the evolution of convective features more appropriately than the traditional 1D PBL scheme, we highlight the need to improve the turbulent length scale formulation.

The Weather Research and Forecasting (WRF) Model is a numerical weather prediction model supported by the National Center for Atmospheric Research (NCAR) to a worldwide community of users. In recognition of the growing use of cloud computing, NCAR is now supporting the model in cloud environments. Specifically, NCAR has established WRF setups with select cloud service providers and produced documentation and tutorials on running WRF in the cloud. Described here are considerations in WRF cloud use and the supported resources, which include cloud setups for the WRF system and a cloud-based tool for model code testing.

Historical observations of severe weather and simulated severe weather environments (i.e., features) from the Global Ensemble Forecast System v12 (GEFSv12) Reforecast Dataset (GEFS/R) are used in conjunction to train and test random forest (RF) machine learning (ML) models to probabilistically forecast severe weather out to days 4–8. RFs are trained with ∼9 years of the GEFS/R and severe weather reports to establish statistical relationships. Feature engineering is briefly explored to examine alternative methods for gathering features around observed events, including simplifying features using spatial averaging and increasing the GEFS/R ensemble size with time lagging. Validated RF models are tested with ∼1.5 years of real-time forecast output from the operational GEFSv12 ensemble and are evaluated alongside expert human-generated outlooks from the Storm Prediction Center (SPC). Both RF-based forecasts and SPC outlooks are skillful with respect to climatology at days 4 and 5 with diminishing skill thereafter. The RF-based forecasts exhibit tendencies to slightly underforecast severe weather events, but they tend to be well-calibrated at lower probability thresholds. Spatially averaging predictors during RF training allows for prior-day thermodynamic and kinematic environments to generate skillful forecasts, while time lagging acts to expand the forecast areas, increasing resolution but decreasing overall skill. The results highlight the utility of ML-generated products to aid SPC forecast operations into the medium range.

To forecast tropical cyclone (TC) intensity and structure changes with fidelity, numerical weather prediction models must be “high definition,” i.e., horizontal grid spacing ≤ 3 km, so that they permit clouds and convection and resolve sharp gradients of momentum and moisture in the eyewall and rainbands. Storm-following nests are computationally efficient at fine resolutions, providing a practical approach to improve TC intensity forecasts. Under the Hurricane Forecast Improvement Project, the operational Hurricane Weather Research and Forecasting (HWRF) system was developed to include telescopic, storm-following nests for a single TC per model integration. Subsequently, HWRF evolved into a state-of-the-art tool for TC predictions around the globe, although its single-storm nesting approach does not adequately simulate TC–TC interactions as they are observed. Basin-scale HWRF (HWRF-B) was developed later with a multistorm nesting approach to improve the simulation of TC–TC interactions by producing high-resolution forecasts for multiple TCs simultaneously. In this study, the multistorm nesting approach in HWRF-B was compared with a single-storm nesting approach using an otherwise identical model configuration. The multistorm approach demonstrated TC intensity forecast improvements, including more realistic TC–TC interactions. Storm-following nests developed in HWRF and HWRF-B will be foundational to NOAA’s next-generation hurricane application in the Unified Forecast System.