Modern Numerical Weather Prediction Research Articles

GEFSv12 High- and Low-Skill Day-10 Tornado Forecasts

Abstract On average, modern numerical weather prediction forecasts for daily tornado frequency exhibit no skill beyond day 10. However, in this extended-range lead window, there are particular model cycles that have exceptionally high forecast skill for tornadoes because of their ability to correctly simulate the future synoptic pattern. Here, model initial conditions that produced a more skillful forecast for tornadoes over the United States were exploited while also highlighting potential causes for low-skill cycles within the Global Ensemble Forecasting System, version 12 (GEFSv12). There were 88 high-skill and 91 low-skill forecasts in which the verifying day-10 synoptic pattern for tornado conditions revealed a western U.S. thermal trough and an eastern U.S. thermal ridge, a favorable configuration for tornadic storm occurrence. Initial conditions for high skill forecasts tended to exhibit warmer sea surface temperatures throughout the tropical Pacific Ocean and Gulf of Mexico, an active Madden–Julian oscillation, and significant modulation of Earth-relative atmospheric angular momentum. Low-skill forecasts were often initialized during La Niña and negative Pacific decadal oscillation conditions. Significant atmospheric blocking over eastern Russia—in which the GEFSv12 overforecast the duration and characteristics of the downstream flow—was a common physical process associated with low-skill forecasts. This work helps to increase our understanding of the common causes of high- or low-skill extended-range tornado forecasts and could serve as a helpful tool for operational forecasters. Significance Statement This research provides a framework for the anticipation of a more (or less) skillful 10-day tornado forecast in an operational numerical weather prediction system. High-skill forecasts were associated with substantial tropical convection and warm sea surface temperature throughout the Pacific Ocean and Gulf of Mexico, whereas the underlying cause of low-skill forecasts were typically associated with a blocking anticyclone over eastern Russia. These findings are important because they permit increased or decreased confidence in a long-range forecast of tornado occurrence based on a dynamical prediction system.

Early warning of low visibility using the ensembling of machine learning approaches for aviation services at Jay Prakash Narayan International (JPNI) Airport Patna

Extremely low visibility affects aviation services. Aviation services need accurate fog and low-visibility predictions for airport operations. Fog and low-visibility forecasting are difficult even with modern numerical weather prediction models and guiding systems. Limitations in comprehending the micro-scale processes that lead to fog formation, intensification, onset, and dissipation complicate fog prediction. This article predicts low visibility for Jay Prakash Narayan International Airport (JPNI), Patna, India, using a historical synoptic dataset. The proposed machine learning (ML) approaches optimize three meta-algorithm approaches: boosting (which reduces variances), bagging (which reduces bias), and stacking (which improves predictive forces). The ML approaches optimize the best prediction algorithms (at level 0) for fog (surface visibility ≤ 1000 m) and dense fog (surface visibility ≤ 200 m), and the suggested ensemble models at level 1 (an ensemble of level 0 ML approaches) deliver the highest performance and stability in prediction output. All time series perform well with the specified model (6-h to 1-h lead time for any combination of observed historical datasets). Airport management, planning, and decision-making rely on high reliability. Because it works well and is reliable, the proposed approaches can be used at other airports in India's Indo-Gangetic Plain.

Spectrum Synergy for Investigating Cloud Microphysics

Abstract Observations from spaceborne microwave (MW) and infrared (IR) passive sensors are the backbone of current satellite meteorology, essential for data assimilation into modern numerical weather prediction and for climate benchmarking. While MW and IR observations from space offer complementary features with respect to cloud properties, their synergy for cloud investigation is currently underexplored, despite the presence of both MW and IR sensors on operational meteorological satellites such as the EUMETSAT Polar System (EPS) MetOp series. As such, several key cloud microphysical properties are not part of the operational products available from EPS MetOp sensors. In addition, the EPS Second Generation (EPS-SG) series, scheduled for launch starting from 2024 onward, will carry sensors such as the Microwave Sounder (MWS) and IASI Next Generation (IASI-NG), enhancing spatial and spectral resolutions and thus capacity to retrieve cloud properties. This article presents the Combined MWS and IASI-NG Soundings for Cloud Properties (ComboCloud) project, funded by EUMETSAT with the overall objective to specify, prototype, and validate algorithms for the retrieval of cloud microphysical properties (e.g., water content and drop effective radius) from the synergy of passive MW and IR observations. The article presents the synergy rationale, the algorithm design, and the results obtained exploiting simulated observations from EPS and EPS-SG sensors, quantifying the benefits to be expected from the MW–IR synergy and the new generation sensors.

Modeling the Shallow Cumulus-Topped Boundary Layer at Gray Zone Resolutions

Abstract This study investigates simulated fair-weather shallow cumulus-topped boundary layer (SCTBL) on kilometer- and subkilometer-scale horizontal resolutions, also known as the numerical gray zone of boundary layer turbulence. Based on a priori analysis of a simulated classic SCTBL with large-eddy simulation, its gray zone scale is determined. The dominant length scale of the cloud layer (CL) is found to be the effective cloud diameter, while that of the underlying mixed layer (ML) is the size of organized convection. The two scales are linked by a simple geometric argument based on vertically coherent updrafts, and are quantified through spectral analysis. Comparison to a simulated dry convective boundary layer (CBL) further reveals that the ML gray zone scale does not differentiate between clear and cloudy conditions with the same bulk stability. A posteriori simulations are then performed over a range of resolutions to evaluate the performance of a recently developed scale-adaptive planetary boundary layer (PBL) scheme. Simulation results suggest indifferences to the scale-adaptive capability. Detailed analyses of flux partition reveal that, in the absence of a shallow cumulus scheme, overly energetic resolved fluxes develop in the CL at gray zone and coarse resolutions, and are responsible for overpredicted resolved convection in the ML. These results suggest that modifications are needed for scale-adaptive PBL schemes under shallow cumulus-topped conditions. Significance Statement Shallow cumulus (ShCu) clouds play an important role in the dynamical and radiative processes of the atmospheric boundary layer. As the grid resolution of modern numerical weather prediction models approach kilometer and subkilometer scales, also known as the gray zone, accurate modeling of ShCu clouds becomes challenging due to difficulties in their parameterization. This study identifies the spatial scale that sets the gray zone of ShCu clouds, providing the key to building better parameterizations. Performance of existing parameterizations developed for clear-sky conditions is evaluated for cloudy conditions, exposing deficiencies and motivating further development.

Deep Learning Augmented Data Assimilation: Reconstructing Missing Information with Convolutional Autoencoders

Abstract Remote sensing data play a critical role in improving numerical weather prediction (NWP). However, the physical principles of radiation dictate that data voids frequently exist in physical space (e.g., subcloud area for satellite infrared radiance or no-precipitation region for radar reflectivity). Such data gaps impair the accuracy of initial conditions derived from data assimilation (DA), which has a negative impact on NWP. We use the barotropic vorticity equation to demonstrate the potential of deep learning augmented data assimilation (DDA), which involves reconstructing spatially complete pseudo-observation fields from incomplete observations and using them for DA. By training a convolutional autoencoder (CAE) with a long simulation at a coarse “forecast” resolution (T63), we obtained a deep learning approximation of the “reconstruction operator,” which maps spatially incomplete observations to a model state with full spatial coverage and resolution. The CAE was applied to an incomplete streamfunction observation (∼30% missing) from a high-resolution benchmark simulation and demonstrated satisfactory reconstruction performance, even when only very sparse (1/16 of T63 grid density) observations were used as input. When only spatially incomplete observations are used, the analysis fields obtained from ensemble square root filter (EnSRF) assimilation exhibit significant error. However, in DDA, when EnSRF takes in the combined data from the incomplete observations and CAE reconstruction, analysis error reduces significantly. Such gains are more pronounced with sparse observation and small ensemble size because the DDA analysis is much less sensitive to observation density and ensemble size than the conventional DA analysis, which is based solely on incomplete observations. Significance Statement Data assimilation plays a critical role in improving the skills of modern numerical weather prediction by establishing accurate initial conditions. However, unobservable regions are common in observation data, particularly those derived from remote sensing. The nonlinear relationship between data from observable regions and the physical state of unobservable regions may impede DA efficiency. As a result, we propose that deep learning be used to improve data assimilation in such cases by reconstructing a spatially complete first guess of the physical state with deep learning and then applying data assimilation to the reconstructed field. Such deep learning augmentation is found effective in improving the accuracy of data assimilation, especially for sparse observation and small ensemble size.

Ensemble prediction using a new dataset of ECMWF initial states – OpenEnsemble 1.0

Abstract. Ensemble prediction is an indispensable tool in modern numerical weather prediction (NWP). Due to its complex data flow, global medium-range ensemble prediction has almost exclusively been carried out by operational weather agencies to date. Thus, it has been very hard for academia to contribute to this important branch of NWP research using realistic weather models. In order to open ensemble prediction research up to the wider research community, we have recreated all 50+1 operational IFS ensemble initial states for OpenIFS CY43R3. The dataset (OpenEnsemble 1.0) is available for use under a Creative Commons licence and is downloadable from an https server. The dataset covers 1 year (December 2016 to November 2017) twice daily. Downloads in three model resolutions (TL159, TL399, and TL639) are available to cover different research needs. An open-source workflow manager, called OpenEPS, is presented here and used to launch ensemble forecast experiments from the perturbed initial conditions. The deterministic and probabilistic forecast skill of OpenIFS (cycle 40R1) using this new set of initial states is comprehensively evaluated. In addition, we present a case study of Typhoon Damrey from year 2017 to illustrate the new potential of being able to run ensemble forecasts outside of major global weather forecasting centres.

Validating the Past: Simulating the Weather in Space Mission Earth Views

AbstractAs we celebrate the fiftieth anniversary of NASA’s Apollo missions, images of Earth simulated with the ECMWF Integrated Forecasting System (IFS) are visually compared with pictures collected during space missions of the past five decades, in particular from the Apollo missions (1968–72). The numerical weather reforecasts use the latest version of the IFS and are initialized from (re)analysis data, which provide our current best representation of the atmospheric state for any given date back to the 1950s. Visible images of our planet are produced from the IFS with a simple simulator whose main inputs are the solar fluxes at the top of the atmosphere. First, a comparison to recent imagery from deep space illustrates the high level of performance of the IFS on recent dates. Then, the validation of the IFS against photographs taken by Apollo 11 and 17 both in-flight and from the lunar surface exhibits a significant level of agreement, despite the absence or very limited number of satellite observations available. This short study confirms that the combination of high-quality initial conditions with a modern numerical weather prediction model can yield reasonably accurate reforecasts of global meteorological conditions, especially cloud systems, for dates as far back as the late 1960s.

Numerical experiments in calculating snow water equivalent and estimating their impact on the formation of spring floods

Seasonal snow cover has a significant impact on forming spring floods. Sparse snow course-measuring network does not meet the requirements of modern tasks related to the technologies of numerical weather prediction (NWP) systems and runoff formation models. Moreover, insufficient volume of hydrometeorological data creates a need to improve spring floods forecasting methods by means of available modern hydrometeorological information related to snow cover. To work out an efficient solution to the issue of initial snow data preparation we need a complex approach including the use of data from satellite, atmospheric models, physical-mathematical models of snow cover and insitu information. This approach will provide modern NWP and hydrological models with reliable initial data on snow cover (snow water equivalent – SWE, snow density – SD). The main purpose of our investigation is related to approbation of satellite data and development of snow cover calculation methods for NWP and hydrological models. Numerous SWE and SD experiments have been performed in order to achieve this aim. A regional snow data assimilation system for COSMORu was implemented during the research. Moreover, a new method of hydrological modelling of spring floods based on ECOMAG model with initial information from COSMO-Ru, SnoWE and in-situ data has been proposed and tested.

Historical Reanalysis: What, How, and Why?

AbstractHistorical reanalyses combine observations of past weather with simulations from modern numerical weather prediction models to provide a consistent, global history of the weather. Recently, a new reanalysis was released that allows observations of the ocean to impact the atmosphere and vice versa. “CERA‐20C: A Coupled Reanalysis of the Twentieth Century” by P. Laloyaux et al. (2018, https://doi.org/10.1029/2018MS001273) describes the first coupled centennial reanalysis, thereby providing more balanced estimates of the ocean‐atmosphere system and allowing for a broader range of studies of the entire Earth system. Results suggest that similar methods could also be leveraged to improve modern weather forecasts.

Ensemble Kalman filter assimilation of near-surface observations over complex terrain: comparison with 3DVAR for short-range forecasts

ABSTRACTSurface observations are the main conventional observations for weather forecasts. However, in modern numerical weather prediction, the use of surface observations, especially those data over complex terrain, remains a unique challenge. There are fundamental difficulties in assimilating surface observations with three-dimensional variational data assimilation (3DVAR). In this study, a series of observing system simulation experiments is performed with the ensemble Kalman filter (EnKF), an advanced data assimilation method to compare its ability to assimilate surface observations with 3DVAR. Using the advanced research version of the Weather Research and Forecasting (WRF) model, results from the assimilation of observations at a single observation station demonstrate that the EnKF can overcome some fundamental limitations that 3DVAR has in assimilating surface observations over complex terrain. Specifically, through its flow-dependent background error term, the EnKF produces more realistic analysis increments over complex terrain in general. More comprehensive comparisons are conducted in a short-range weather forecast using a synoptic case with two severe weather systems: a frontal system over complex terrain in the western US and a low-level jet system over the Great Plains. The EnKF is better than 3DVAR for the analysis and forecast of the low-level jet system over flat terrain. However, over complex terrain, the EnKF clearly performs better than 3DVAR, because it is more capable of handling surface data in the presence of terrain misrepresentation. In addition, results also suggest that caution is needed when dealing with errors due to model terrain representation. Data rejection may cause degraded forecasts because data are sparse over complex terrain. Owing to the use of limited ensemble sizes, the EnKF analysis is sensitive to the choice of horizontal and vertical localisation scales.

Beating the Uncertainties: Ensemble Forecasting and Ensemble‐Based Data Assimilation in Modern Numerical Weather Prediction

Accurate numerical weather forecasting is of great importance. Due to inadequate observations, our limited understanding of the physical processes of the atmosphere, and the chaotic nature of atmospheric flow, uncertainties always exist in modern numerical weather prediction (NWP). Recent developments in ensemble forecasting and ensemble‐based data assimilation have proved that there are promising ways to beat the forecast uncertainties in NWP. This paper gives a brief overview of fundamental problems and recent progress associated with ensemble forecasting and ensemble‐based data assimilation. The usefulness of these methods in improving high‐impact weather forecasting is also discussed.

Numerical Methods in Atmospheric and Oceanic Modelling: The Andre J. Robert Memorial Volume

Most people, even including some in the scientific community, do not realize how much the weather forecasts they use to guide the activities of their daily lives depend on very complex mathematics and numerical methods that are the basis of modern numerical weather prediction (NWP). André Robert (1929–1993), to whom Numerical Methods in Atmospheric and Oceanic Modelling is dedicated, had a career that contributed greatly to the growth of NWP and the role that the atmospheric computer models of NWP play in our society. There are probably no NWP models running anywhere in the world today that do not use numerical methods introduced by Robert, and those of us who work with and use these models everyday are indebted to him.The first two chapters of the volume are chronicles of Robert's life and career. The first is a 1987 interview by Harold Ritchie, one of Robert's many proteges and colleagues at the Canadian Atmospheric Environment Service. The interview traces Robert's life from his birth in New York to French Canadian parents, to his emigration to Quebec at an early age, his education and early employment, and his rise in stature as one of the preeminent research meteorologists of our time. An amusing anecdote he relates is his impression of weather forecasts while he was considering his first job as a meteorologist in the early 1950s. A newspaper of the time placed the weather forecast and daily horoscope side by side, and Robert regarded each to have a similar scientific basis. Thankfully he soon realized there was a difference between the two, and his subsequent career certainly confirmed the distinction.

History of Numerical Weather Prediction at the National Meteorological Center

The first modern numerical weather prediction (NWP) models were developed for the computer that was announced in 1932 at the Institute for Advanced Study, in Princeton, New Jersey. Within 3 yr three agencies of the United States Government jointly created a numerical weather prediction service, but it was quickly discovered that current models had very serious defects. After considerable research, the first operationally effective model was achieved in 1958—a barotropic model covering most of the Northern Hemisphere. Over the years, models have evolved through multilevel filtered equation models and several primitive equation models. Analysis and data assimilation systems necessary for timeliness were also developed, and have likewise evolved. The result has been a revolution in weather forecasting.

Meteorological fields from remote infrared sensing

Modern numerical weather prediction techniques require global observations of the atmospheric state and structure parameters. The current meteorological observing system, which is based on radiosonde balloon observations, has extensive gaps. Remote sensing of the Earth atmosphere emission spectrum from satellites can fill these gaps. The physical basis for extracting information on meteorological fields from such remote observations is explained. The problem reduces to that of solving a linear Fredholm equation of the first kind in the presence of noisy data. There is no unique solution to such a problem. The mathematical techniques-inversion techniques-that are currently used to solve the problem are reviewed. Examples are given of meteorological fields obtained from remote infrared sensing from satellites. Results indicate that meteorological parameters such as temperature and geopotential height of constant pressure surfaces can be measured-in conditions of clear skies-to accuracies approaching that of the radiosonde system. Other meterological variables, e.g., water vapor and ozone, can be determined to a lesser degree of accuracy. Applications of the remotely sensed fields are described. Problem areas and suggested solutions are discussed.