Numerical Weather Prediction Centers Research Articles

The Assimilation of SSMIS Radiances in Numerical Weather Prediction Models

<para xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> The measurement uncertainty requirements imposed by numerical weather prediction (NWP) data assimilation applications for temperature sounding radiances are very demanding. For an ensemble of observations collected during an orbit, (postbias correction) measurement uncertainties of <formula formulatype="inline"> <tex>$\sim$</tex></formula>0.2 K (at 1 <formula formulatype="inline"><tex>$\sigma$</tex> </formula>) or better are required in tropospheric sounding channels to improve analyses, and hence forecasts, from current NWP models. A significant fraction of F-16 Special Sensor Microwave Imager/Sounder (SSMIS) observations are affected by calibration errors caused by solar intrusions into the warm calibration load and by thermal emission from the main reflector. The magnitude of these effects is as large as 1.5 K for the lower atmospheric temperature sounding channels. This paper describes the approach to correct for these effects, which involves data averaging, flagging solar intrusions, and modeling reflector emission. The resulting quality of the radiances is improved by a factor of three to four for mid-tropospheric temperature sounding channels. Observation minus background field differences are reduced from 0.5–0.8 K (at one standard deviation) for uncorrected data to 0.2 K for corrected data. Although localized biases remain in the corrected data, assimilation experiments using SSMIS data at four operational NWP centers (Met Office, ECMWF, NCEP, and NRL) show a neutral-to-positive impact on forecast quality in the Southern Hemisphere with, for example, mean sea-level pressure forecast errors at days 1–4 reduced by 0.5%–2.5%. Impacts in the Northern Hemisphere are neutral in most assimilation experiments. </para>

Simulations of Arctic mixed‐phase clouds in forecasts with CAM3 and AM2 for M‐PACE

Simulations of mixed‐phase clouds in forecasts with the NCAR Atmosphere Model version 3 (CAM3) and the GFDL Atmospheric Model version 2 (AM2) for the Mixed‐Phase Arctic Cloud Experiment (M‐PACE) are performed using analysis data from numerical weather prediction centers. CAM3 significantly underestimates the observed boundary layer mixed‐phase cloud fraction and cannot realistically simulate the variations of liquid water fraction with temperature and cloud height due to its oversimplified cloud microphysical scheme. In contrast, AM2 reasonably reproduces the observed boundary layer cloud fraction while its clouds contain much less cloud condensate than CAM3 and the observations. The simulation of the boundary layer mixed‐phase clouds and their microphysical properties is considerably improved in CAM3 when a new physically based cloud microphysical scheme is used (CAM3LIU). The new scheme also leads to an improved simulation of the surface and top of the atmosphere longwave radiative fluxes. Sensitivity tests show that these results are not sensitive to the analysis data used for model initialization. Increasing model horizontal resolution helps capture the subgrid‐scale features in Arctic frontal clouds but does not help improve the simulation of the single‐layer boundary layer clouds. AM2 simulated cloud fraction and LWP are sensitive to the change in cloud ice number concentrations used in the Wegener‐Bergeron‐Findeisen process while CAM3LIU only shows moderate sensitivity in its cloud fields to this change. This paper shows that the Wegener‐Bergeron‐Findeisen process is important for these models to correctly simulate the observed features of mixed‐phase clouds.

Toward a weak constraint operational 4D-Var system: application to the Burgers' equation

Strong constraint (perfect model assumption) 4D-Var algorithms are increasingly used for synoptic and global scale atmospheric data assimilation at operational numerical weather prediction centers around the world. The data assimilation windows currently range from 6 hours to 12 hours at different centers. It is preferable to have as many independent observations as possible in each data assimilation window under the variational framework. Longer data assimilation windows generally increase the information content from the observations, but also make the perfect model assumption more improper. It is clear that weak constraint (imperfect model assumption) 4D-Var algorithms will be required to properly combine the background forecast with high resolution observations in longer data assimilation windows in the not too distant future. A weak constraint observation space atmospheric 4D-Var system, NAVDAS-AR (NRL Atmospheric Variational Data Assimilation System - Accelerated Representer), is currently under development at the Naval Research Laboratory (NRL) in Monterey. The impact of the model errors is presented in a residual term resulting from a four-dimensional matrix/vector multiplication of the model error covariance with the adjoint field in the right hand side of the tangent linear model. The calculation of the residual term is generally very computationally intensive. However, it can be obtained accurately and efficiently when some special choices are made for the model error covariances. In this short contribution we present an approach to include the impact of the model errors in the NAVDAS-AR. The steps involved in this approach are demonstrated through an application to the Burgers' equation.

Validating Basic Surface Variables in the Australian Bureau of Meteorology Model with CEOP EOP3 In-situ Data

The Coordinated Enhanced Observing Period (CEOP) Project has initiated the collection of model output location time series (MOLTS) data from numerical weather prediction and assimilation centers, including the Australian Bureau of Meteorology. These were designed to complement the collection of in situ observational data sets at the same 41 locations. This study is a preliminary attempt to assess the differences and similarities between the MOLTS and time series of the in-situ data. The MOLTS from the Bureau’s global model for a number of atmospheric surface variables are compared with in-situ data for 6 locations for the entire year of EOP3 (October 2002 to September 2003). Using surface pressure as a prototype a number of preliminary comparisons of different properties of the two types of data are performed. The model fields from the 6 hour assimilation forecast series and from the concatenated 12 to 36 hour forecasts are first compared with hourly in-situ observations in simple gross terms; simple statistics and scatter plots have been obtained to identify areas for further investigation. Global wavelet power spectra are compared to show the behavior of the dominant modes of variability and the average annual diurnal variations of the variables are derived. Some case studies with other variables with increasing sensitivity to sub-grid scales and increasing dependency on surface properties are then documented. The model and in-situ time series show similar behavior for surface pressure with greater differences for the other variables. Precipitation shows the greatest differences.

High Clouds over Oceans in the ECMWF 15- and 45-Yr Reanalyses

Abstract The reanalysis programs of numerical weather prediction (NWP) centers provide global, comprehensive descriptions of the atmosphere and of the earth’s surface over long periods of time. The high realism of their representation of key NWP parameters, like temperature and winds, implies some realism for less emblematic parameters, such as cloud cover, but the degree of this realism needs to be documented. This study aims to evaluate the high clouds over open oceans in the ECMWF 15- and 45-yr reanalyses. The assessment is based on a new 23-yr climatology of monthly frequencies of high-cloud occurrence retrieved from the infrared radiances measured by operational polar satellites. It is complemented by data from the International Satellite Cloud Climatology Project. It is shown that the 45-yr ECMWF reanalysis dramatically improves on the previous 15-yr reanalysis for the realism of seasonal and interannual variations in high clouds, despite remaining systematic errors. More than 60% of the observed anomalies during the January 1979–February 2002 period over large oceanic basins are captured by the latest reanalysis. However the realism of the analyses in the areas and in the years with sparse observations appears to be poor. Consequently, the interannual variations may not be reliable before January 1979 in most parts of the world. Possible improvements of the handling of assimilated satellite observations before and after this date are suggested.

Test of a Poor Man’s Ensemble Prediction System for Short-Range Probability Forecasting

Abstract Current operational ensemble prediction systems (EPSs) are designed specifically for medium-range forecasting, but there is also considerable interest in predictability in the short range, particularly for potential severe-weather developments. A possible option is to use a poor man’s ensemble prediction system (PEPS) comprising output from different numerical weather prediction (NWP) centers. By making use of a range of different models and independent analyses, a PEPS provides essentially a random sampling of both the initial condition and model evolution errors. In this paper the authors investigate the ability of a PEPS using up to 14 models from nine operational NWP centers. The ensemble forecasts are verified for a 101-day period and five variables: mean sea level pressure, 500-hPa geopotential height, temperature at 850 hPa, 2-m temperature, and 10-m wind speed. Results are compared with the operational ECMWF EPS, using the ECMWF analysis as the verifying “truth.” It is shown that, despite its smaller size, PEPS is an efficient way of producing ensemble forecasts and can provide competitive performance in the short range. The best relative performance is found to come from hybrid configurations combining output from a small subset of the ECMWF EPS with other different NWP models.

Short‐ to medium‐range superensemble precipitation forecasts using satellite products: 1. Deterministic forecasting

As a comprehensive extension of the previous multimodel/multianalysis superensemble (SE) studies of rainfall forecasts, the benefits and prospects of the SE precipitation forecasts are explored using satellite products. Three different precipitation ensemble configurations are first established from a great number of numerical experiments. These configurations are multianalysis (MA), multicumulus‐scheme (MC), and multimodel (MM) configurations. A set of MA ensemble comes from the use of several different satellite‐derived rain rates through the physical initialization procedure within the Florida State University Global Spectral Model (FSUGSM) system. Six different state‐of‐the art cumulus parameterization schemes are incorporated into the FSUGSM in order to introduce the MC ensemble configuration. The MM configuration is composed of an FSU control forecast and those provided by five operational numerical weather prediction centers. In addition to the original technique, a possible deterministic SE enhancement technique (regression dynamic linear model) is then proposed and applied to the above three configurations of ensemble members as well as all of them together. The impact of a higher resolution family of models on the performance of SE forecasts is extensively investigated by repeating the above procedure with T170 resolution precipitation forecasts. Results show that short‐ to medium‐range SE forecasts are generally superior in skill to various conventional forecasts. A notably improved (∼20%) quantitative precipitation forecast is exhibited by the newly proposed SE technique. The MM configuration proved to be the most effective ensemble prediction system. Although a higher‐resolution SE forecast requires a large amount of computing time, it turns out that the impact is significant not only in skill scores but also in resolving mesoscale‐based convective disturbances.

AIRS near-real-time products and algorithms in support of operational numerical weather prediction

The assimilation of Atmospheric InfraRed Sounder, Advanced Microwave Sounding Unit-A, and Humidity Sounder for Brazil (AIRS/AMSU/HSB) data by Numerical Weather Prediction (NWP) centers is expected to result in improved forecasts. Specially tailored radiance and retrieval products derived from AIRS/AMSU/HSB data are being prepared for NWP centers. There are two types of products - thinned radiance data and full-resolution retrieval products of atmospheric and surface parameters. The radiances are thinned because of limitations in communication bandwidth and computational resources at NWP centers. There are two types of thinning: (1) spatial and spectral thinning and (2) data compression using principal component analysis (PCA). PCA is also used for quality control and for deriving the retrieval first guess used in the AIRS processing software. Results show that PCA is effective in estimating and filtering instrument noise. The PCA regression retrievals show layer mean temperature (1 km in troposphere, 3 km in stratosphere) accuracies of better than 1 K in most atmospheric regions from simulated AIRS data. Moisture errors are generally less than 15% in 2-km layers, and ozone errors are near 10% over approximately 5-km layers from simulation. The PCA and regression methodologies are described. The radiance products also include clear field-of-view (FOV) indicators. The residual cloud amount, based on simulated data, for FOVs estimated to be clear (free of clouds) is about 0.5% over ocean and 2.5% over land.

Potential improvement to forecasts of two severe storms using targeted observations

AbstractThe potential to improve short‐range forecasts of two extratropical storms by using supplementary observations in regions lacking accurate observations is investigated. In the idealized framework used here, a control and a truth experiment are selected from a set of forecasts initialized with analyses from different numerical weatherprediction centres. Synthetic soundings of wind and temperature are created from the truth experiment and are assimilated with four‐dimensional variational analysis using the operational observation‐error estimates for radiosondes and the initial condition of the control experiment as background. Through multiple analysis/forecast experiments we obtain a nonlinear estimate of the optimal zone for observing (OZO); that is the zone in which the use of a given number of supplementary observations leads to the largest reduction in forecast error. We evaluate targeting techniques based on either total‐energy singular vectors (TESVs) or on Hessian singular vectors (HSVs) by comparison with the OZO and by comparison with experiments in which the same amount of supplementary observations are distributed in an untargeted manner, namely with a random distribution scheme (RDS).Overall, the HSV targeting is superior to the TESV targeting in the two cases. In one case there is a significant difference between the target regions determined with TESVs and HSVs. The HSV‐based observing strategy resembles the OZO in terms of the observing region and the achieved forecast‐error reduction. With the RDS, the forecast error is variable and likely to be larger than the forecast error obtained with singular‐vector targeting. Experiments with target regions of different sizes show that supplementary observations in an area of about 3×106 km2 are required to achieve a significant forecast improvement. A two‐dimensional sampling pattern with soundings spaced at a distance of about 1–2 times the horizontal correlation length‐scale of the backgrounderror estimate appears very efficient. In additional impact experiments for one case, observations were perturbed with noise to represent observational error. The perturbations are almost as likely to improve the forecast as to worsen it compared with the forecast using unperturbed observations. Copyright © 2002 Royal Meteorological Society.

Radiance and Jacobian intercomparison of radiative transfer models applied to HIRS and AMSU channels

The goals of this study are the evaluation of current fast radiative transfer models (RTMs) and line‐by‐line (LBL) models. The intercomparison focuses on the modeling of 11 representative sounding channels routinely used at numerical weather prediction centers: 7 HIRS (High‐resolution Infrared Sounder) and 4 AMSU (advanced microwave sounding unit) channels. Interest in this topic was evident by the participation of 24 scientists from 16 institutions. An ensemble of 42 diverse atmospheres was used and results compiled for 19 infrared models and 10 microwave models, including several LBL RTMs. For the first time, not only radiances but also Jacobians (of temperature, water vapor, and ozone) were compared to various LBL models for many channels. In the infrared, LBL models typically agree to within 0.05–0.15 K (standard deviation) in terms of top‐of‐the‐atmosphere brightness temperature (BT). Individual differences up to 0.5 K still exist, systematic in some channels, and linked to the type of atmosphere in others. The best fast models emulate LBL BTs to within 0.25 K, but no model achieves this desirable level of success for all channels. The ozone modeling is particularly challenging. In the microwave, fast models generally do quite well against the LBL model to which they were tuned. However, significant differences were noted among LBL models. Extending the intercomparison to the Jacobians proved very useful in detecting subtle or more obvious modeling errors. In addition, total and single gas optical depths were calculated, which provided additional insight on the nature of differences.

On the Use of NCEP–NCAR Reanalysis Surface Marine Wind Fields for a Long-Term North Atlantic Wave Hindcast

Abstract This paper uses a state-of-the-art, third-generation wave model to evaluate the marine surface wind fields produced in the National Centers for Environmental Protection–National Center for Atmospheric Research (NCEP–NCAR) Reanalysis (NRA) project. Three alternative NRA wind fields were initially considered by assessing the resulting wave hindcasts against wave measurements in the North Atlantic Ocean. The surface 10-m wind field was found to be the most skillful and was selected for further analysis. While the wind fields from the NRA were found to be at least as skillful as the best of the analyses produced by operational Numerical Weather Prediction centers, they had significant deficiencies when compared to kinematically analyzed wind fields carried out in detailed hindcast studies. Storm peak wave heights in extratropical storms were systematically underestimated at higher sea states due to underestimation of peak wind speeds in major jet streak features propagating about intense extratropica...

A Scalable Version of the Navy Operational Global Atmospheric Prediction System Spectral Forecast Model

The Navy Operational Global Atmospheric Prediction System (NOGAPS) includes a state-of-the-art spectral forecast model similar to models run at several major operational numerical weather prediction (NWP) centers around the world. The model, developed by the Naval Research Laboratory (NRL) in Monterey, California, has run operational at the Fleet Numerical Meteorological and Oceanographic Center (FNMOC) since 1982, and most recently is being run on a Cray C90 in a multi-tasked configuration. Typically the multi-tasked code runs on 10 to 15 processors with overall parallel efficiency of about 90%. resolution is T159L30, but other operational and research applications run at significantly lower resolutions. A scalable NOGAPS forecast model has been developed by NRL in anticipation of a FNMOC C90 replacement in about 2001, as well as for current NOGAPS research requirements to run on DOD High-Performance Computing (HPC) scalable systems. The model is designed to run with message passing (MPI). Model design criteria include bit reproducibility for different processor numbers and reasonably efficient performance on fully shared memory, distributed memory, and distributed shared memory systems for a wide range of model resolutions. Results for a wide range of processor numbers, model resolutions, and different vendor architectures are presented. Single node performance has been disappointing on RISC based systems, at least compared to vector processor performance. This is a common complaint, and will require careful re-examination of traditional numerical weather prediction (NWP) model software design and data organization to fully exploit future scalable architectures.

Cloud Cover in the ECMWF Reanalysis

Abstract Data from reanalyses recently carried out by several climate and numerical weather prediction centers will find a variety of applications in different branches of atmospheric science. A careful evaluation of the many aspects of these datasets is a prerequisite for their successful use. This paper describes the implementation of a fully prognostic cloud scheme into the ECMWF reanalysis system and provides a first assessment of the simulation of cloud cover by comparing it with monthly mean cloud cover derived from satellite observations in the context of the International Satellite Cloud Climatology Project for the years 1983–90. Special emphasis is put on the major cloud regimes and their intra- and interannual variation. The main deficiencies identified are an underestimation of extratropical cloud cover over the oceans by 10%–15%, an overestimation of trade wind cumulus cover by about 10%–15%, an underestimation of stratocumulus off the west coasts of the subtropical continents by 15%, and an u...

The Second-Generation Global Forecast System at the Central Weather Bureau in Taiwan

The global forecast system (GFS), which started its operation in 1988 at the Central Weather Bureau in Taiwan, has been upgraded to incorporate better numerical methods and more complete parameterization schemes. The second-generation GFS uses multivariate optimum interpolation analysis and incremental nonlinear normal-mode initialization to initialize the forecast model. The forecast model is a global primitive equation model with a resolution of 18 sigma levels in the vertical and 79 waves of triangular truncation in the horizontal. The forecast model includes a 1.5-order eddy mixing parameterization, a gravity wave drag parameterization, a shallow convection parameterization, a relaxed version of Arakawa–Schubert cumulus parameterization, grid-scale condensation calculation, and longwave and shortwave radiative transfer calculations with consideration of fractional clouds. The performance of the second-generation GFS is significantly better than the first-generation GFS. For two 3-month periods in winter 1995/96 and summer 1996, the second-generation GFS provided forecasters with 5-day forecasts where the averaged 500-mb height anomaly correlation coefficients for the Northern Hemisphere were greater than 0.6. Observational data available to the GFS are much less than those at other numerical weather prediction centers, especially in the Tropics and Southern Hemisphere. The GRID messages of 5° resolution, ECMWF 24-h forecast 500-mb height and 850- and 200-mb wind fields available once a day on the Global Telecommunications System are used as supplemental observations to increase the data coverage for the GFS data assimilation. The supplemental data improve the GFS performance both in the analysis and forecast.

The Second–Generation Global Forecast System at the Central Weather Bureau in Taiwan

Abstract The global forecast system (GFS), which started its operation in 1988 at the Central Weather Bureau in Taiwan, has been upgraded to incorporate better numerical methods and more complete parameterization schemes. The second-generation GFS uses multivariate optimum interpolation analysis and incremental nonlinear normal-mode initialization to initialize the forecast model. The forecast model is a global primitive equation model with a resolution of 18 sigma levels in the vertical and 79 waves of triangular truncation in the horizontal. The forecast model includes a 1.5-order eddy mixing parameterization, a gravity wave drag parameterization, a shallow convection parameterization, a relaxed version of Arakawa–Schubert cumulus parameterization, grid-scale condensation calculation, and longwave and shortwave radiative transfer calculations with consideration of fractional clouds. The performance of the second-generation GFS is significantly better than the first-generation GFS. For two 3-month periods in winter 1995/96 and summer 1996, the second-generation GFS provided forecasters with 5-day forecasts where the averaged 500-mb height anomaly correlation coefficients for the Northern Hemisphere were greater than 0.6. Observational data available to the GFS are much less than those at other numerical weather prediction centers, especially in the Tropics and Southern Hemisphere. The GRID messages of 5° resolution, ECMWF 24-h forecast 500-mb height and 850- and 200-mb wind fields available once a day on the Global Telecommunications System are used as supplemental observations to increase the data coverage for the GFS data assimilation. The supplemental data improve the GFS performance both in the analysis and forecast.

Deterministic methods for prediction of tropical cyclone tracks

ABSTRACT. The paper presents a state-of-art review of different objective techniques available for tropical cyclone track prediction. A brief description of current theories of tropical cyclone motion is given. Deterministic models with statistical and dynamical methods have been discussed. Recent advances in the understanding of cyclone structure and motion aspects have led to improved prediction of tropical cyclones. There has been considerable progress in the field of prediction by dynamical methods. High resolution Limited Area Models (LAM) as well as Global Circulation Models (GCM) are now being used extensively by most of the leading operational numerical weather prediction (NWP) centres in the world The major achievements towards improvement of such models have come from improved horizontal resolution of the models, inclusion of physical processes, use of synthetic and other non-conventional data in the data assimilation schemes and nudging method for initial matching of analysed cyclone centres with corresponding observations.&#x0D; A brief description of further improvement in deterministic approach for prediction of tropical cyclone tracks is outlined.&#x0D;

Accounting for Surface Wave Distortion of the Marine Wind Profile in Low-Level Ocean Storms Wind Measurements

Abstract Marine wind measurements at three heights (3.0,4.5, and 5.0 m) from both moored and drifting buoys during the Ocean Storms Experiment are described. These winds are compared with each other, with winds from ships, from subsurface ambient acoustic noise, and from the analyses of three numerical weather prediction centers. In the mean, wind directions generally differ by only a small constant offset of a few degrees. No wave influence on the wind direction is evident, because the differences are not systematic and, with few exceptions, they are less than the expected error. After correcting for some apparent calibration and instrument bias, the Ocean Storms wind speeds display similar behavior when compared to the analyzed wind products. There is excellent agreement up to a transition wind speed between 7 and 10 m s−1, above which all the measured winds tend to be relatively low. The transition speed is found to increase with anemometer height, so this behavior is interpreted as being due to the di...

On the bogussing of tropical cyclones in numerical models: A comparison of vortex profiles

At the resolutions currently in use, and with the sparse oceanic data coverage, numerical analyses cannot adequately represent tropical cyclone circulations for use in numerical weather prediction models. In many cases there is no circulation present at all. Most numerical weather prediction centers therefore employ a “bogussing” scheme to force a tropical cyclone vortex into the numerical analysis. The standard procedure is to define a synthetic data distribution based on an analytically prescribed vortex, which is passed to the analysis scheme as a set of high quality observations.