Search

The Global Ensemble Forecast System (GEFS) is upgraded to version 12, in which the legacy Global Spectral Model (GSM) is replaced by a model with a new dynamical core—the Finite Volume Cubed-Sphere Dynamical Core (FV3). Extensive tests were performed to determine the optimal model and ensemble configuration. The new GEFS has cubed-sphere grids with a horizontal resolution of about 25 km and an increased ensemble size from 20 to 30. It extends the forecast length from 16 to 35 days to support subseasonal forecasts. The stochastic total tendency perturbation (STTP) scheme is replaced by two model uncertainty schemes: the stochastically perturbed physics tendencies (SPPT) scheme and stochastic kinetic energy backscatter (SKEB) scheme. Forecast verification is performed on a period of more than two years of retrospective runs. The results show that the upgraded GEFS outperforms the operational-at-the-time version by all measures included in the GEFS verification package. The new system has a better ensemble error–spread relationship, significantly improved skills in large-scale environment forecasts, precipitation probability forecasts over CONUS, tropical cyclone track and intensity forecasts, and significantly reduced 2-m temperature biases over North America. GEFSv12 was implemented on 23 September 2020.

For the newly implemented Global Ensemble Forecast System, version 12 (GEFSv12), a 31-yr (1989–2019) ensemble reforecast dataset has been generated at the National Centers for Environmental Prediction (NCEP). The reforecast system is based on NCEP’s Global Forecast System, version 15.1, and GEFSv12, which uses the Finite Volume 3 dynamical core. The resolution of the forecast system is ∼25 km with 64 vertical hybrid levels. The Climate Forecast System (CFS) reanalysis and GEFSv12 reanalysis serve as initial conditions for the Phase 1 (1989–99) and Phase 2 (2000–19) reforecasts, respectively. The perturbations were produced using breeding vectors and ensemble transforms with a rescaling technique for Phase 1 and ensemble Kalman filter 6-h forecasts for Phase 2. The reforecasts were initialized at 0000 (0300) UTC once per day out to 16 days with 5 ensemble members for Phase 1 (Phase 2), except on Wednesdays when the integrations were extended to 35 days with 11 members. The reforecast dataset was produced on NOAA’s Weather and Climate Operational Supercomputing System at NCEP. This study summarizes the configuration and dataset of the GEFSv12 reforecast and presents some preliminary evaluations of 500-hPa geopotential height, tropical storm track, precipitation, 2-m temperature, and MJO forecasts. The results were also compared with GEFSv10 or GEFS Subseasonal Experiment reforecasts. In addition to supporting calibration and validation for the National Water Center, NCEP Climate Prediction Center, and other National Weather Service stakeholders, this high-resolution subseasonal dataset also serves as a useful tool for the broader research community in different applications.

The National Centers for Environmental Prediction have generated an 18-yr (1999–2016) subseasonal (weeks 3 and 4) reforecast to support the Climate Prediction Center’s operational mission. To create this reforecast, the subseasonal experiment version of the GEFS was run every Wednesday, initialized at 0000 UTC with 11 members. The Climate Forecast System Reanalysis (CFSR) and Global Data Assimilation System (GDAS) served as the initial analyses for 1999–2010 and 2011–16, respectively. The analysis of 2-m temperature error demonstrates that the model has a strong warm bias over the Northern Hemisphere (NH) and North America (NA) during the warm season. During the boreal winter, the 2-m temperature errors over NA exhibit large interannual and intraseasonal variability. For NA and the NH, weeks 3 and 4 errors are mostly saturated, with initial conditions having a negligible impact. Week 2 errors (day 11) are ~88.6% and 86.6% of their saturated levels, respectively. The 1999–2015 reforecast biases were used to calibrate the 2-m temperature forecasts in 2016, which reduces (increases) the systematic error (forecast skill) for NA, the NH, the Southern Hemisphere, and the tropics, with a maximum benefit for NA during the warm season. Overall, analysis adjustment for the CFSR period makes bias characteristics more consistent with the GDAS period over the NH and tropics and substantially improves the corresponding forecast skill levels. The calibration of the forecast using week 2 bias provides similar skill to using weeks 3 and 4 bias, promising the feasibility of using week 2 bias to calibrate the weeks 3 and 4 forecast. Our results also demonstrate that 10-yr reforecasts are an optimal training period. This is particularly beneficial considering limited computing resources.

With the successful upgrade of its deterministic model GFS (v15) on June 12, 2019, NCEP has scheduled the implementation of its next Global Ensemble Forecast System (GEFS v12) in the summer of 2020. These two model upgrades on deterministic and ensemble forecast systems are substantially different from previous upgrades. A new dynamical core (FV3) is adopted for the first time in the NCEP operational models, replacing the previous spectral dynamical core. The previous 3-category Zhao-Carr microphysics scheme is also being replaced by a more advanced 6-category GFDL microphysics scheme. From an ensemble model perspective, the previous GEFS has already demonstrated great success in past decades for weather and week-2 prediction by providing reliable probabilistic forecasts. Recently, there has been a large demand for subseasonal prediction, and GEFS v12 forecasts will be extended to 35 days to cover this time range. To better represent large uncertainties associated with this time scale, SPPT (stochastic physics perturbed tendency) and SKEB (stochastic kinetic energy backscatter) stochastic schemes are taking the place of the original STTP (stochastic total tendency perturbation), and a prescribed SST generated from combination of NSST and bias corrected CFS forecasts is also applied to simulate the sub-seasonal variation of SST forcing. As a major system upgrade, a 2.5-year retrospective run of GEFS v12 is carried out to evaluate the model performance. A 30-year reforecast will be provided to stakeholders and the public to calibrate the forecast. The improvement of predictability and prediction skill will be studied through various measurements across tropical to extratropical areas in terms of deterministic (ensemble mean) and probabilistic (ensemble distribution) forecasts. The characteristics of model systematic error will be identified from comparing the major changes of the two state-of-art ensemble systems. As GEFS serves the most crucial model guidance for 5-7 day hurricane forecasts in support of the NHC (National Hurricane Center) and other customers, model capability in predicting tropical cyclone track and intensity is also examined from the retrospective runs. The results show there are significant improvements for tropical cyclone track forecast in North Atlantic and the western North Pacific, in particular, the intensity forecast is improved remarkably in all the basins.

NOAA is accelerating its efforts to improve the numerical guidance and prediction capability for extended range (weeks 3 and 4) prediction in its seamless forecast system. Madden Julian Oscillation (MJO) is the dominant mode of sub-seasonal variability in tropics and prediction skill of MJO is investigated in this paper. We used different configurations of the NCEP Global Ensemble Forecast System (GEFS) to perform the experiments. The configurations include (1) the operational version of the stochastic perturbation forced with operational Sea Surface Temperatures (SSTs); (2) an updated stochastic physics forced with operational SSTs; (3) an updated stochastic physics forced with bias-corrected SSTs that are from Climate Forecast System (Version 2); and (4) as in (3) but with the addition of a scale aware-convection scheme. We evaluated MJO prediction skill from the experiments using Wheeler–Hendon indices and also examined the performance of the forecast system on prediction of key MJO components. We found that using the updated stochastic scheme improved the MJO prediction lead-time by about 4 days. Further updating the underlying SSTs with the bias corrected CFSv2 forecast increased the MJO prediction lead time by another 1.7 days. The best configuration of the four experiments is the last configuration which extends forecast lead time by ~ 9 days. Further investigation shows that upper and lower level zonal wind over the tropics has larger improvement than the outgoing longwave radiation (OLR). The improvement of the MJO prediction skill appears to be related primarily to the improvement in the representation associated circulations and OLR over the tropical West Pacific.

The Global Ensemble Forecasting System (GEFS) is being extended from 16 to 35 days to cover the subseasonal period, bridging weather and seasonal forecasts. In this study, the impact of SST forcing on the extended-range land-only global 2-m temperature, continental United States (CONUS) accumulated precipitation, and MJO skill are explored with version 11 of the GEFS (GEFSv11) under various SST forcing configurations. The configurations consist of 1) the operational GEFS 90-day e-folding time of the observed real-time global SST (RTG-SST) anomaly relaxed to climatology, 2) an optimal AMIP configuration using the observed daily RTG-SST analysis, 3) a two-tier approach using the CFSv2-predicted daily SST, and 4) a two-tier approach using bias-corrected CFSv2-predicted SST, updated every 24 h. The experimental period covers the fall of 2013 and the winter of 2013/14. The results indicate that there are small differences in the ranked probability skill scores (RPSSs) between the various SST forcing experiments. The improvements in forecast skill of the Northern Hemisphere 2-m temperature and precipitation for weeks 3 and 4 are marginal, especially for North America. The bias-corrected CFSv2-predicted SST experiment generally delivers superior performance with statistically significant improvement in spatially and temporally aggregated 2-m temperature RPSSs over North America. Improved representation of the SST forcing (AMIP) increased the forecast skill for MJO indices up through week 2, but there is no significant improvement of the MJO forecast skill for weeks 3 and 4. These results are obtained over a short period with weak MJO activity and are also subject to internal model weaknesses in representing the MJO. Additional studies covering longer periods with upgraded model physics are warranted.

Persistent open ridges and blocking highs (maxima) of 500-hPa geopotential height (Z500; PMZ) adjacent in space and time are identified and tracked as one event with a Lagrangian objective approach to derive their climatological statistics with some dynamical reasoning. A PMZ starts with a core that contains a local eddy maximum of Z500 and its neighboring grid points whose eddy values decrease radially to about 20 geopotential meters (GPMs) smaller than the maximum. It connects two consecutive cores that share at least one grid point and are within 10° of longitude of each other using an intensity-weighted location. The PMZ ends at the core without a successor. On each day, the PMZ impacts an area of grid points contiguous to the core and with eddy values decreasing radially to 100 GPMs. The PMZs identified and tracked consist of persistent ridges, omega blockings and blocked anticyclones either connected or as individual events. For example, the PMZ during 2–13 August 2003 corresponds to persistent open ridges that caused the extreme heatwave in Western Europe. Climatological statistics based on the PMZs longer than 3 days generally agree with those of blockings. In the Northern Hemisphere, more PMZs occur in DJF season than in JJA and their duration both exhibit a log-linear distribution. Because more omega-shape blocking highs and open ridges are counted, the PMZs occur more frequently over Northeast Pacific than over Atlantic-Europe during cool seasons. Similar results are obtained using the 200-hPa geopotential height (in place of Z500), indicating the quasi-barotropic nature of the PMZ.

A new version of the Global Ensemble Forecast System (GEFS, v11) is tested and compared with the operational version (v10) in a 2-yr parallel run. The breeding-based scheme with ensemble transformation and rescaling (ETR) used in the operational GEFS is replaced by the ensemble Kalman filter (EnKF) to generate initial ensemble perturbations. The global medium-range forecast model and the Global Forecast System (GFS) analysis used as the initial conditions are upgraded to the GFS 2015 implementation version. The horizontal resolution of GEFS increases from Eulerian T254 (~52 km) for the first 8 days of the forecast and T190 (~70 km) for the second 8 days to semi-Lagrangian T574 (~34 km) and T382 (~52 km), respectively. The sigma pressure hybrid vertical layers increase from 42 to 64 levels. The verification of geopotential height, temperature, and wind fields at selected levels shows that the new GEFS significantly outperforms the operational GEFS up to days 8–10 except for an increased warm bias over land in the extratropics. It is also found that the parallel system has better reliability in the short-range probability forecasts of precipitation during warm seasons, but no clear improvement in cold seasons. There is a significant degradation of TC track forecasts at days 6–7 during the 2012–14 TC seasons over the Atlantic and eastern Pacific. This degradation is most likely a sampling issue from a low number of TCs during these three TC seasons. The results for an extended verification period (2011–14) and the recent two hurricane seasons (2015 and 2016) are generally positive. The new GEFS became operational at NCEP on 2 December 2015.

Two perturbation generation schemes, the ensemble transformation with rescaling (ETR) and the ensemble Kalman filter (EnKF), are compared for the NCEP operational environment for the Global Ensemble Forecast System (GEFS). Experiments that utilize each of the two schemes are carried out and evaluated for two boreal summer seasons. It is found that these two schemes generally have comparable performance. Experiments utilizing both perturbation methods fail to generate sufficient spread at medium-range lead times beyond day 8. In general, the EnKF-based experiment outperforms the ETR in terms of the continuous ranked probability skill score (CRPSS) in the Northern Hemisphere (NH) for the first week. In the SH, the ensemble mean forecast is more skillful from the ETR perturbations. Additional experiments are performed with the stochastic total tendency perturbation (STTP) scheme, in which the total tendencies of all model variables are perturbed to represent the uncertainty in the forecast model. An improved spread–error relationship is found for the ETR-based experiments, but the STTP increases the ensemble spread for the EnKF-based experiment that is already overdispersive at early lead times, especially in the SH. With STTP employed, an increase in the EnKF-based CRPSS in the NH is reduced with a larger degradation in both the probability and ensemble-mean forecast skills in the SH. The results indicate that a rescaling of the EnKF initial perturbations and/or tuning of the STTP scheme is required when STTP is applied using the EnKF-based perturbations. This study provided guidance for the replacement of ETR with EnKF perturbations as part of the 2015 GEFS implementation.

The National Centers for Environmental Prediction (NCEP) Global Ensemble Forecast System (GEFS) has been in daily operation to provide probabilistic guidance for public since December 1992. Since July 2017, the GEFS was extended from 16 days to 35 days forecast to support NCEP Climate Prediction Center (CPC)’s sub-seasonal forecast. The latest GEFS version was upgraded in three areas to improve sub-seasonal forecast: (1) introducing a new set of stochastic physical perturbations to improve model uncertainty representation for the tropics; (2) a 2-tiered SST approach to consider ocean impact; and (3) a new scale-aware convection scheme to improve model physics for tropical convection and MJO forecasts. The new set of stochastic physical perturbations include stochastic kinetic energy backscatter to make up subscale energy lost during model integration; stochastic physics perturbation tendency with five different spatial and temporal scales to perturb physical tendency; and stochastic perturbed humidity on the model lower level. After upgraded to new set of stochastic physical perturbations, the MJO forecast skill has been improved from 12.5 days of a 25-month period to nearly 22 days by combining all three modifications include stochastic physics. In the extratropics, the 500-hPa geopotential height; surface temperature and precipitation are improved for sub-seasonal timescale as well. However, the raw forecast skills of surface temperature and precipitation are extremely low, and the results imply that calibration may be important and necessary for surface temperature and precipitation forecast for the sub-seasonal timescale due to the large systematic model errors.

Two widely used precipitation analyses are the Climate Prediction Center (CPC) unified global daily gauge analysis and Stage IV analysis based on quantitative precipitation estimate with multisensor observations. The former is based on gauge records with a uniform quality control across the entire domain and thus bears more confidence, but provides only 24-h accumulation at ⅛° resolution. The Stage IV dataset, on the other hand, has higher spatial and temporal resolution, but is subject to different methods of quality control and adjustments by different River Forecasting Centers. This article describes a methodology used to generate a new dataset by adjusting the Stage IV 6-h accumulations based on available joint samples of the two analyses to take advantage of both datasets. A simple linear regression model is applied to the archived historical Stage IV and the CPC datasets after the former is aggregated to the CPC grid and daily accumulation. The aggregated Stage IV analysis is then adjusted based on this linear model and then downscaled back to its original resolution. The new dataset, named Climatology-Calibrated Precipitation Analysis (CCPA), retains the spatial and temporal patterns of the Stage IV analysis while having its long-term average and climate probability distribution closer to that of the CPC analysis. The limitation of the methodology at some locations is mainly associated with heavy to extreme precipitation events, which the Stage IV dataset tends to underestimate. CCPA cannot effectively correct this because of the linear regression model and the relative scarcity of heavy precipitation in the training data sample.

The ensemble transform with rescaling (ETR) method has been used to produce fast-growing components of analysis error in the NCEP Global Ensemble Forecast System (GEFS). The rescaling mask contained in the ETR method constrains the amplitude of perturbations to reflect regional variations of analysis error. However, because of a lack of suitable three-dimensional (3D) analysis error estimation, in the operational GEFS the mask is based on the estimated analysis error at 500 hPa and is not flow dependent but changes monthly. With the availability of an ensemble-based data assimilation system at NCEP, a 3D mask can be computed. This study generates initial perturbations by the ensemble transform with 3D rescaling (ET_3DR) and compares the performance with the ETR. Meanwhile, the ET_3DR is also applied within the ensemble Kalman filter (EnKF) method (hereafter EnKF_3DR). Results from a set of experiments indicate that the 3D mask suppresses perturbations less in unstable regions. Relative to the ETR, the large amplitudes of the ET_3DR initial perturbations at 500 hPa better reflect areas of baroclinic instability over the extratropics and deep convection over the tropics. Furthermore, the maxima of the vertical distribution for the ET_3DR initial perturbations correspond to the heights of the subtropical westerly and tropical easterly jet regions. Such perturbations produce faster spread growths. Results with EnKF_3DR also show benefits from an orthonormalization by the ensemble transform algorithm and amplitude constraint by the 3D mask rescaling. Thus, the EnKF_3DR forecasts outperform the EnKF.

The main task of this study is to introduce a statistical postprocessing algorithm to reduce the bias in the National Centers for Environmental Prediction (NCEP) and Meteorological Service of Canada (MSC) ensemble forecasts before they are merged to form a joint ensemble within the North American Ensemble Forecast System (NAEFS). This statistical postprocessing method applies a Kalman filter type algorithm to accumulate the decaying averaging bias and produces bias-corrected ensembles for 35 variables. NCEP implemented this bias-correction technique in 2006. NAEFS is a joint operational multimodel ensemble forecast system that combines NCEP and MSC ensemble forecasts after bias correction. According to operational statistical verification, both the NCEP and MSC bias-corrected ensemble forecast products are enhanced significantly. In addition to the operational calibration technique, three other experiments were designed to assess and mitigate ensemble biases on the model grid: a decaying averaging bias calibration method with short samples, a climate mean bias calibration method, and a bias calibration method using dependent data. Preliminary results show that the decaying averaging method works well for the first few days. After removing the decaying averaging bias, the calibrated NCEP operational ensemble has improved probabilistic performance for all measures until day 5. The reforecast ensembles from the Earth System Research Laboratory’s Physical Sciences Division with and without the climate mean bias correction were also examined. A comparison between the operational and the bias-corrected reforecast ensembles shows that the climate mean bias correction can add value, especially for week-2 probability forecasts.

Hydrological processes are strongly coupled with atmospheric processes related, for example, to precipitation and temperature, and a coupled atmosphere–land surface system is required for a meaningful hydrological forecast. Since the atmosphere is a chaotic system with limited predictability, ensemble forecasts offer a practical tool to predict the future state of the coupled system in a probabilistic fashion, potentially leading to a more complete and informative hydrologic prediction. As ensemble forecasts with coupled meteorological–hydrological models are operationally running at major numerical weather prediction centers, it is currently possible to produce a gridded streamflow prognosis in the form of a probabilistic forecast based on ensembles. Evaluation and improvement of such products require a comprehensive assessment of both components of the coupled system. In this article, the atmospheric component of a coupled ensemble forecasting system is evaluated in terms of its ability to provide reasonable forcing to the hydrological component and the effect of the uncertainty represented in the atmospheric ensemble system on the predictability of streamflow as a hydrological variable. The Global Ensemble Forecast System (GEFS) of NCEP is evaluated following a “perfect hydrology” approach, in which its hydrological component, including the Noah land surface model and attached river routing model, is considered free of errors and the initial conditions in the hydrological variables are assumed accurate. The evaluation is performed over the continental United States (CONUS) domain for various sizes of river basins. The results from the experiment suggest that the coupled system is capable of generating useful gridded streamflow forecast when the land surface model and the river routing model can successfully simulate the hydrological processes, and the ensemble strategy significantly improves the forecast. The expected forecast skill increases with increasing size of the river basin. With the current GEFS system, positive skill in short-range (one to three days) predictions can be expected for all significant river basins; for the major rivers with mean streamflow more than 500 m3 s−1, significant skill can be expected from extended-range (the second week) predictions. Possible causes for the loss of skills, including the existence of systematic error and insufficient ensemble spread, are discussed and possible approaches for the improvement of the atmospheric ensemble forecast system are also proposed.

During May 1998, the Center for Analysis and Prediction of Storms (CAPS) at the University of Oklahoma coordinated a multi-institution numerical forecast project known as the Storm and Mesoscale Ensemble Experiment (SAMEX). SAMEX involved, for the first time, the real-time operation of four different ensembles of mesoscale models over the same region of the United States. The main purpose of this paper is the evaluation of the ensemble forecasts, performed at a relatively coarse resolution of 30 km. An additional SAMEX goal not discussed here is to compare the value of the ensemble forecasts against single forecasts made over smaller subregions of the Great Plains at both intermediate (10 km) and high (3 km) resolution. The SAMEX ’98 ensembles consisted of a single 36-h control forecast from the ARPS (at CAPS), the Penn State‐NCAR fifth-generation Mesoscale Model (at NSSL), and the Eta Model and Regional Spectral Model (at NCEP), all with horizontal resolutions of approximately 30 km, and perturbed runs, resulting in a grand ensemble of 25 members. The forecasts of geopotential heights, temperatures, and moisture were verified against the Eta operational analyses, rather than observations. Unlike global ensembles, which tend to be useful in the medium range, the mesoscale SAMEX ensembles provided useful information in the short range. A major result is that the performance of the ensemble of multiple forecast systems is much better than that of each individual ensemble system, probably because it represents more realistically the current uncertainties in both models and initial conditions. A similar advantage from the use of multimodel, multianalysis systems has been observed with global ensembles. The SAMEX results also show that perturbations to model physics parameterizations, as well as the use of consistent perturbations in the boundary conditions, are important for regional ensemble forecasting. Efforts are now under way to compare the ensemble forecasts against those made using higher spatial resolution, and follow-on SAMEX experiments are anticipated in other geographical areas and weather regimes. Although the main results of this paper appear to be very robust, they were based on a small number of cases, and similar experiments carried out during other periods will help to test their significance.