Search

A climatology of significant tornadoes [SIGTOR, tornadoes rated (E)F2+ on the (enhanced) Fujita scale] within China and in three subregions, including northern, central, and southern China, is first presented for the period 1980–2016. In total, 129 SIGTOR are recorded in China, with an average of 3.5 per year. The tornado inflow environments of the south-central and southeast regions of the United States (USC and USSE) are compared with those of China and its subregions based on sounding-derived parameters including shear, storm-relative helicity, convective available potential energy (CAPE), lifting condensation level (LCL), etc. Soundings are extracted from the ERA5 reanalysis dataset. The results confirm that the SIGTOR in USSE are characterized by high shear, low CAPE, and low LCL whereas those in USC are characterized by moderate-to-high shear, high CAPE, and high LCL. The thermodynamic conditions of tornadic cases are favorable for China, with moderate-to-high CAPE and low-to-moderate LCL, but their kinematic conditions are much less favorable than in the United States, a fact that is believed to be primarily responsible for the lower tornado frequency and intensity in China. The high CAPE in China is due mostly to high humidity. For three subregions in China, the central China cases account for 60% of total samples, and its environmental features are similar to those of China. The average shear with northern China cases is stronger than that with the other two subregions, and the midlayer is relatively dry. The southern China SIGTOR have the most conducive humidity conditions, but the CAPE and shear there are the lowest. The northern, central, and southern China environments can be considered as representative of midlatitude, subtropical, and tropical regions. Significance Statement We document the climatological characteristics of significant tornadoes (SIGTOR) within China and compare the inflow environments of SIGTOR in China and its subregions with those in the U.S. central and southeastern regions. The availability of hourly high-resolution ERA5 data makes the environments based on extracted proximity soundings much more accurate than possible with earlier reanalyses. The environmental characteristics show systematic differences in the tornado environments of different regions of China and the United States and suggest different roles played by thermodynamic and kinematic conditions for tornado formation. Overall, the environmental differences are consistent with the resulting frequency and intensities of SIGTOR. The findings are helpful toward improving tornado forecasting and warning or even understanding of potential impacts of climate change on SIGTOR, especially in China, where such studies are rarer.

Parallel back-building convective lines are often observed extending to the southwest of some mesoscale convective systems (MCSs) embedded in the mei-yu front in China. The convective lines with echo training behavior can quickly develop into a stronger convective group of echoes, resulting in locally heavy rainfall within the mei-yu front rainband. The initiation mechanism of the back-building convective lines is still unclear and is studied based on high-resolution numerical simulation of a case that occurred during 27–28 June 2013. In the present case, the new convection along the convective lines was found to be forced by nonuniform interaction between the cold outflow associated with the mei-yu front MCSs and the warm southerly airflow on the south side of the mei-yu front, which both are modified by local terrain. The mei-yu front MCSs evolved from the western to the eastern side of a basin surrounded by several mesoscale mountains and induced cold outflow centered over the eastern part of the basin. The strong southwest airflow ahead of the mei-yu front passed the Nanling Mountains and impacted the cold outflow within the basin. The nonuniform interaction led to the first stage of parallel convective line formation, in which the low mountains along the boundary of the two airflows enhanced the heterogeneity of their interaction. Subsequently, the convective group quickly developed from the first stage convective lines resulted in apparent precipitation cooling that enhanced the cold outflow and made the cold outflow a sharp southward windshift. The enhanced cold outflow pushed the warm southerly airflow southward and impacted the mountains on the southeast side of the basin, where the roughly parallel mountain valleys or gaps play a controlling role in a second stage formation of parallel convective lines.

Based on the hourly gauge-satellite merged precipitation data with the spatial resolution of 0.1° × 0.1° during 2013 ~ 2018, we have evaluated the performance of the convection-permitting Weather Research and Forecasting model at Nanjing University (WRF_NJU) in forecasting the precipitation diurnal variation and the associated atmospheric circulation over the eastern Tibetan Plateau and its surrounding regions during summer. Results indicate that WRF_NJU model can well reproduce the diurnal cycle of the summer precipitation in terms of the diurnal peak time, duration and magnitude. In addition, the eastward propagation of rainfall systems with long duration along the eastern Tibetan Plateau (ETP) and its adjacent areas can also be properly captured. The WRF_NJU model can reasonably reproduce the relevant atmospheric circulation during summer as well. However, the model tends to underestimate the summer precipitation amount (PA) and precipitation frequency (PF) over most time of a day. Relatively larger biases in the occurring time and magnitude of PA and PF diurnal peaks can be noted over the ETP and Sichuan Basin. Further analysis suggests that the underestimation of PA over the ETP is attributed to the much lower moisture supply, weaker low-level southwesterly winds and less convective available potential energy (CAPE) in the WRF_NJU model than in ERA5. Over Sichuan Basin, the underestimated PA is related to the weaker upward motion, which is corresponding to the cold biases of surface air temperature in WRF_NJU. Findings of this study provide the basic model biases and may be helpful to further improve the model physical processes.

In numerical simulations of deep convection at kilometer-scale horizontal resolutions, in-cloud subgrid-scale (SGS) turbulence plays an important role in the transport of heat, moisture, and other scalars. By coarse graining a 50 m high-resolution large-eddy simulation (LES) of an idealized supercell storm to kilometer-scale grid spacings ranging from 250 m to 4 km, the SGS fluxes of heat, moisture, cloud, and precipitating water contents are diagnosed a priori. The kilometer-scale simulations are shown to be within the “gray zone” as in-cloud SGS turbulent fluxes are comparable in magnitude to the resolved fluxes at 4 km spacing, and do not become negligible until ~500 m spacing. Vertical and horizontal SGS fluxes are of comparable magnitudes; both exhibit nonlocal characteristics associated with deep convection as opposed to local gradient-diffusion type of turbulent mixing. As such, they are poorly parameterized by eddy-diffusivity-based closures. To improve the SGS representation of turbulent fluxes in deep convective storms, a scale-similarity LES closure is adapted to kilometer-scale simulations. The model exhibits good correlations with LES-diagnosed SGS fluxes, and is capable of representing countergradient fluxes. In a posteriori tests, supercell storms simulated with the refined similarity closure model at kilometer-scale resolutions show better agreement with the LES benchmark in terms of SGS fluxes than those with a turbulent-kinetic-energy-based gradient-diffusion scheme. However, it underestimates the strength of updrafts, which is suggested to be a consequence of the model effective resolution being lower than the native grid resolution.

Based on a priori analysis of large-eddy simulations (LESs) of the convective atmospheric boundary layer, improved turbulent mixing and dissipation length scales are proposed for a turbulence kinetic energy (TKE)-based planetary boundary layer (PBL) scheme. The turbulent mixing length incorporates surface similarity and TKE constraints in the surface layer, and makes adjustments for lateral entrainment effects in the mixed layer. The dissipation length is constructed based on balanced TKE budgets accounting for shear, buoyancy, and turbulent mixing. A nongradient term is added to the TKE flux to correct for nonlocal turbulent mixing of TKE. The improved length scales are implemented into a PBL scheme, and are tested with idealized single-column convective boundary layer (CBL) cases. Results exhibit robust applicability across a broad CBL stability range, and are in good agreement with LES benchmark simulations. It is then implemented into a community atmospheric model and further evaluated with 3D real-case simulations. Results of the new scheme are of comparable quality to three other well-established PBL schemes. Comparisons between simulated and radiosonde-observed profiles show favorable performance of the new scheme on a clear day.

The impact of assimilating China's operational X‐band Phased‐Array radar's (X‐PAR) data on the analysis and warning forecast of the vortex structure and intensity of the June 8, 2018 Foshan, Guangdong province, tornadic storm was investigated for the first time using an Ensemble Kalman Filter (EnKF) data assimilation system. Both radar radial velocity (Vr) and reflectivity (Z) from two S‐band operational radars and one X‐PAR were assimilated. Deterministic forecasts were launched every 6 min from 05:42 UTC (20 min before the tornado touched down) to 06:00 UTC from the EnKF mean analysis field. Five experiments were conducted to examine the added capability of Z assimilation of the EnKF system, and to investigate the impact of assimilating X‐PAR data on the analysis and prediction of the tornadic storm. Compared to the experiment without Z assimilation, the assimilation of Z reduced the analysis error and greatly reduced the forecast error of Z. The assimilation of X‐PAR data greatly improved the vortex structure of the tornadic storm at low levels, and improved the intensity of the rear inflow of the tornadic storm, especially with a higher assimilation frequency. Compared to the experiments without X‐PAR data assimilation, assimilating X‐PAR data improved the predictability of tornadic storm.

A regional ensemble Kalman filter (EnKF) data assimilation (DA) and forecast system was recently established based on the Gridpoint Statistical Interpolation (GSI) analysis system. The EnKF DA system was tested with continuous three-hourly updated cycles followed by 18-h deterministic forecasts from every three-hourly ensemble mean analysis. Initial tests showed negative to neutral impacts of assimilating satellite radiance data due to the improper bias correction procedure. In this study, two bias correction schemes within the established EnKF DA system are investigated and the impact of assimilating additional polar-orbiting satellite radiance is also investigated. Two group experiments are conducted. The purpose of the first group is to evaluate the bias correction procedure. Two online bias correction methods based on GSI 3DVar and EnKF algorithms are used to assimilate AMSU-A radiance data. Results show that both variational and EnKF-based bias correction procedures effectively reduce the observation and background radiance differences, achieving positive impacts on forecasts. With proper bias correction, we assimilate full radiance observations including AMSU-A, AMSU-B, AIRS, HIRS3/4, and MHS in the second group. The relative percentage improvements (RPIs) for all forecast variables compared to those without radiance data assimilation are mostly positive, with the RPI of upper-air relative humidity being the largest. Additionally, precipitation forecasts on a downscaled 13-km grid from 40-km EnKF analyses are also improved by radiance assimilation for almost all forecast hours.

An EF4 supercell tornado that occurred on 23 June 2016 in Funing, China is simulated using the WRF model. Five nested forecast experiments are run, with the horizontal grid spacing of the inner-most nest being 4000, 1333, 444, 148, and 49 m, respectively. All experiments reproduce the general characteristics of the observed supercell storm, but tornado-like vortex does not develop until the grid spacing is 444 m or less. The tornadoes on the 444, 148 and 49 m grids reach EF1, EF2 and EF3 intensities, respectively. The tornado vortex simulated on the 444 m grid tends to maintain a one-cell structure, while a ring of high-vorticity develops during the mature stage of the tornado vortex on the 148 m grid leading to a two-cell structure. On the 49 m grid, microscale vortices along the outflow boundary merge into and help organize the main tornado vortex while multiple ‘suction vortices’ develop later along the high-vorticity ring leading to a multi-vortex tornado. These sub-vortices create localized regions of intense winds due to the super-positioning of the sub-vortex and main vortex circulations and the system translation speed. Sub-vortices also create irregular strong surface wind patterns with localized ‘damage cores’, broadening the tornadic wind swath; these are consistent with the Funing tornado damage survey. Based on the results, we recommend a horizontal grid spacing of at least 500 m for real-time warn-on-forecast applications in order to capture tornado-like vortices while for tornado dynamics, prediction and predictability studies 50 m or less is recommended.

The vertical turbulent velocity variance normalized by the convective velocity squared as a function of the boundary layer depth–normalized height [i.e., ] in the convective boundary layer (CBL) over a homogeneous surface exhibits a near-universal profile, as demonstrated by field observations, laboratory experiments, and numerical simulations. The profile holds over a wide CBL stability range set by the friction velocity, CBL depth, and surface heating. In contrast, the normalized horizontal turbulent velocity variance increases monotonically with decreasing stability. This study investigates the independence of the profile to changes in CBL stability, or more precisely, wind shear. Large-eddy simulations of several convective and neutral cases are performed by varying surface heating and geostrophic winds. Analysis of the turbulent kinetic energy budgets reveals that the conversion term between and depends almost entirely on buoyancy. This explains why does not vary with shear, which is a source to only. Further analysis through rotational and divergent decomposition suggests that the near-universal profile of is fundamentally related to the dynamics and interactions of local and nonlocal CBL turbulence. Specifically, the preferential interactions between local wavenumbers and the downscale energy cascade of CBL turbulence offer plausible explanations to the universal profile of .

The diurnal variation of precipitation over the Dabie Mountains (DBM) in eastern China during the 2013 mei-yu season is investigated with forecasts of a regional convection-permitting model. Simulated precipitation is verified against surface rain-gauge observations. The observed morning precipitation peak on the windward (relative to the prevailing synoptic-scale wind) side of the DBM is reproduced with good spatial and temporal accuracy. The interaction between the DBM and a nocturnal boundary layer low-level jet (BLJ) due to the inertial oscillation mechanism is shown to be responsible for this precipitation peak. The BLJ is aligned with the lower-level southwesterly synoptic-scale flow that carries abundant moisture. The BLJ core is established at around 0200 LST upwind of the mountains. It moves towards the DBM and reaches maximum intensity at about 70 km ahead of the mountains. When the BLJ impinges upon the windward side of the DBM in the early morning, mechanical lifting of moist air leads to condensation and subsequent precipitation.

During the afternoon of 28 April 2015, a multicellular convective system swept southward through much of Jiangsu Province, China, over about 7 h, producing egg-sized hailstones on the ground. The hailstorm event is simulated using the Advanced Regional Prediction System (ARPS) at 1-km grid spacing. Different configurations of the Milbrandt–Yau microphysics scheme are used, predicting one, two, and three moments of the hydrometeor particle size distributions (PSDs). Simulated reflectivity and maximum estimated size of hail (MESH) derived from the simulations are verified against reflectivity observed by operational S-band Doppler radars and radar-derived MESH, respectively. Comparisons suggest that the general evolution of the hailstorm is better predicted by the three-moment scheme, and neighborhood-based MESH evaluation further confirms the advantage of the three-moment scheme in hail size prediction. Surface accumulated hail mass, number, and hail distribution characteristics within simulated storms are examined across sensitivity experiments. Results suggest that multimoment schemes produce more realistic hail distribution characteristics, with the three-moment scheme performing the best. Size sorting is found to play a significant role in determining hail distribution within the storms. Detailed microphysical budget analyses are conducted for each experiment, and results indicate that the differences in hail growth processes among the experiments can be mainly ascribed to the different treatments of the shape parameter within different microphysics schemes. Both the differences in size sorting and hail growth processes contribute to the simulated hail distribution differences within storms and at the surface.

Turbulent mixing in the daytime convective boundary layer (CBL) is carried out by organized nonlocal updrafts and smaller local eddies. In the upper mixed layer of the CBL, heat fluxes associated with nonlocal updrafts are directed up the local potential temperature gradient. To reproduce such countergradient behavior in parameterizations, a class of planetary boundary layer schemes adopts a countergradient correction term in addition to the classic downgradient eddy-diffusion term. Such schemes are popular because of their simple formulation and effective performance. This study reexamines those schemes to investigate the physical representations of the gradient and countergradient (GCG) terms, and to rebut the often-implied association of the GCG terms with heat fluxes due to local and nonlocal (LNL) eddies. To do so, large-eddy simulations (LESs) of six idealized CBL cases are performed. The GCG fluxes are computed a priori with horizontally averaged LES data, while the LNL fluxes are diagnosed through conditional sampling and Fourier decomposition of the LES flow field. It is found that in the upper mixed layer, the gradient term predicts downward fluxes in the presence of positive mean potential temperature gradient but is compensated by the upward countergradient correction flux, which is larger than the total heat flux. However, neither downward local fluxes nor larger-than-total nonlocal fluxes are diagnosed from LES. The difference reflects reduced turbulence efficiency for GCG fluxes and, in terms of physics, conceptual deficiencies in the GCG representation of CBL heat fluxes.

A grid-refinement-based method is implemented in a community atmospheric model to improve the representation of convective boundary layer (CBL) turbulence on gray-zone [i.e., ~O(1) km] grids. At this resolution, CBL convection is partially resolved and partially subgrid scale (SGS), such that neither traditional mesoscale planetary boundary layer (PBL) schemes nor SGS closures for large-eddy simulations (LESs) are appropriate. The proposed method utilizes two-way interactive nesting to refine the horizontal resolution of the unstable surface layer of the daytime CBL. SGS turbulent mixing in the fine nest and coarse parent grids are parameterized by an LES turbulence closure and a PBL scheme, respectively. The method does not rely on predetermined empirical functions to introduce grid (scale) dependency and in theory works with any PBL scheme. Compared to the stand-alone gray-zone simulation, the proposed approach shows improvements in terms of higher-order statistics, the timing of the onset of resolved convection, and the convective structures. A deficiency of the method exists when the nest domain is limited to the surface layer; the convective structures become gradually contaminated by spurious convection on the parent gray-zone grid. A deeper nest domain alleviates the issue at increased computational costs.

Three-moment bulk microphysics schemes are attractive since they can determine the shape parameter α of the gamma particle size distributions (PSDs) from the predicted moments, but they are more computationally expensive compared with two-moment schemes. In this study, new diagnostic relations between α and the mean-mass diameter are developed for rain, graupel and hail hydrometeor categories based on the simulations for a severe hailstorm over eastern China using the Milbrandt and Yau (MY) three-moment scheme. The new relations are introduced into the MY two-moment scheme and applied to the simulation of the same hailstorm at 1 km grid spacing. Different configurations of the MY schemes, including one-moment, two-moment with fixed α, with the original and new α diagnostic relations, and with the new α relation applied to hail only, and three-moment schemes, are also used to simulate the hailstorm. The new fully diagnostic-α two-moment scheme is found to improve the simulation of the general storm structure compared with the other two-moment schemes. It also reproduces the maximum estimated size of hail closer to that from the three-moment scheme while still retaining computational efficiency. Explicit prediction of hail by the new diagnostic-α two-moment scheme, including the surface accumulated hail mass, number and distribution is the most consistent with that of the three-moment scheme. The experiment with the new α relation applied to hail only predicts a larger number of smaller hail and higher radar reflectivity, consistent with rain and graupel size over-estimation from excessive size sorting. Detailed microphysical budget analyses for hail, graupel and rain indicate that the new diagnostic relations yield substantial improvements in hydrometeor mass growth processes and size sorting representation.

The impact of horizontal propagation of mountain waves on the orographic gravity wave drag (OGWD) in the stratosphere and lower mesosphere of the Northern Hemisphere is evaluated for the first time. Using a fine-resolution (1 arc min) terrain and 2.5°×2.5° European Centre for Medium-Range Weather Forecasts ERA-Interim reanalysis data during 2011–2016, two sets of OGWD are calculated offline according to a traditional parameterization scheme (without horizontal propagation) and a newly proposed scheme (with horizontal propagation). In both cases, the zonal mean OGWDs show similar spatial patterns and undergo a notable seasonal variation. In winter, the OGWD is mainly distributed in the upper stratosphere and lower mesosphere of middle to high latitudes, whereas the summertime OGWD is confined in the lower stratosphere. Comparison between the two sets of OGWD reveal that the horizontal propagation of mountain waves tends to decrease (increase) the OGWD in the lower stratosphere (middle to upper stratosphere and lower mesosphere). Consequently, including the horizontal propagation of mountain waves in the parameterization of OGWD can reduce the excessive OGWD in the lower stratosphere and strengthen the insufficient gravity wave forcing in the mesosphere, which are the known problems of traditional OGWD schemes. The impact of horizontal propagation is more prominent in winter than in summer, with the OGWD in western Tibetan Plateau, Rocky Mountains, and Greenland notably affected.

The model gray zone refers to the range of grid spacings comparable to the dominant length scale of the flow. In the gray zone, the flow is partially resolved and partially subgrid scale (SGS). Neither ensemble-averaging-based parameterizations nor turbulence closures are appropriate for parameterizing the effects of SGS motions on the resolved flow. The gray zone of the convective boundary layer (CBL) is in the range of CBL depth, typically O(1) km. A new approach that seeks explicit resolution of the unstable surface layer through a nest layer of fine grid spacing is proposed to improve CBL parameterization in the gray zone. To provide the theoretical basis for the approach, a linear analytic model is presented, and one-way nested simulations are performed to investigate the dynamical coupling between the surface layer and the mixed layer. The analytic model shows that at the onset of thermal instability, the vertical and horizontal structures of the mixed layer are set by surface-layer forcings. The nested 3D simulations extend the findings from the analytic model and further reveal potential improvements in high-order statistics and resolved convective structures both including and extending above the nest region compared to the stand-alone gray-zone simulations. This study suggests that when the most energetic scales of CBL convection are resolved in the surface layer, the overall simulation of the CBL improves at gray-zone resolutions.

Compared to the representation of vertical turbulent mixing through various planetary boundary layer (PBL) schemes, the treatment of horizontal turbulent mixing in the boundary layer has received much less attention. In mesoscale and convective-scale models, subgrid-scale horizontal turbulent mixing has traditionally been associated with mesoscale circulations or eddies. Its parameterization most often adopts the gradient-diffusion model, where the horizontal mixing coefficients are usually set constant, or through the 2D Smagorinsky formulation, or in some cases based on the 1.5-order turbulence kinetic energy (TKE) closure. For horizontal turbulent mixing associated with boundary layer eddies, the traditional schemes are shown to perform poorly. This work investigates the characteristic turbulence velocity and length scales based on analysis of a well-resolved, wide-domain large-eddy simulation of a convective boundary layer (CBL). To improve the representation of horizontal turbulent mixing by CBL eddies, a class of schemes is proposed with different levels of sophistication. The first two schemes can be used together with first-order PBL schemes, while the third uses TKE to define its characteristic velocity scale and can be used together with TKE-based higher-order PBL schemes. The proposed parameterizations are tested a posteriori in idealized simulations of turbulent dispersion of a passive scalar. Comparisons show improved horizontal dispersion by the proposed schemes and further demonstrate the weakness of the existing schemes.

In the late afternoon of 19 March 2014, a severe hailstorm swept through eastern-central Zhejiang province, China. The storm produced golf-ball sized hail, strong winds, and lighting, lasting approximately one hour over the coastal city of Taizhou. The Advanced Regional Prediction System (ARPS) is used to simulate the hailstorm using different configurations of the Milbrandt-Yau microphysics scheme that predict one, two, or three moments of the hydrometeor particle size distribution. Simulated fields, including accumulated precipitation and maximum estimated hail size (MESH), are verified against rain gauge observations and radar-derived MESH, respectively. For the case of the 19 March 2014 storms, the general evolution is better predicted with multi-moment microphysics schemes than with the one-moment scheme; the three-moment scheme produces the best forecast. Predictions from the three-moment scheme qualitatively agree with observations in terms of size and amount of hail reaching the surface. The life cycle of the hailstorm is analyzed, using the most skillful, three-moment forecast. Based upon the tendency of surface hail mass flux, the hailstorm life cycle can be divided into three stages: developing, mature, and dissipating. Microphysical budget analyses are used to examine microphysical processes and characteristics during these three stages. The vertical structures within the storm and their link to environmental shear conditions are discussed; together with the rapid fall of hailstones, these structures and conditions appear to dictate this pulse storm's short life span. Finally, a conceptual model for the life cycle of pulse hailstorms is proposed.

Wavelet-decomposition-based intensity-scale skill (ISS) score is a verification metric which decomposes the forecast fields into different scales and then calculates verification scores. However, due to the “double-penalty” issues, ISS at small scales are often very low even when the forecasts appear subjectively skillful. The displacement error is an important reason for the low ISS. To address this problem, verification methods based on a combination of neighborhood and scale-separation verification approaches are explored. Instead of calculating ISS, a neighborhood-based fractions skill score at different spatial scales, which we call IS_FSS is proposed. Additionally, to reduce the impact of intensity bias, percentile-based instead of fixed thresholds are used in IS_FSS, leading to ISP_FSS. Those two newly developed verification scores are then used to assess WRF forecasts at 4 km and 12 km grid spacings (WRF-4 and WRF-12, respectively) for a case and the entire Meiyu season of 2016. Compared to ISS scores, both IS_FSS and ISP_FSS show more positive verification scores of both WRF-4 and WRF-12 at small scales. Moreover, IS_FSS and ISP_FSS are able to differentiate WRF-12 and WRF-4 forecasts at smaller scales when ISS cannot. Both scores indicate that WRF-4 outperforms WRF-12 for spatial scales from 12 km through 96 km. The scores of WRF-12 improve more when using ISP_FSS than WRF-4, because the former has higher intensity bias.

On 21 July 2012, an extreme rainfall event that recorded a maximum rainfall amount over 24 hours of 460 mm, occurred in Beijing, China. Most operational models failed to predict such an extreme amount. In this study, a convective-permitting ensemble forecast system (CEFS), at 4-km grid spacing, covering the entire mainland of China, is applied to this extreme rainfall case. CEFS consists of 22 members and uses multiple physics parameterizations. For the event, the predicted maximum is 415 mm d-1 in the probability-matched ensemble mean. The predicted high-probability heavy rain region is located in southwest Beijing, as was observed. Ensemble-based verification scores are then investigated. For a small verification domain covering Beijing and its surrounding areas, the precipitation rank histogram of CEFS is much flatter than that of a reference global ensemble. CEFS has a lower (higher) Brier score and a higher resolution than the global ensemble for precipitation, indicating more reliable probabilistic forecasting by CEFS. Additionally, forecasts of different ensemble members are compared and discussed. Most of the extreme rainfall comes from convection in the warm sector east of an approaching cold front. A few members of CEFS successfully reproduce such precipitation, and orographic lift of highly moist low-level flows with a significantly southeasterly component is suggested to have played important roles in producing the initial convection. Comparisons between good and bad forecast members indicate a strong sensitivity of the extreme rainfall to the mesoscale environmental conditions, and, to less of an extent, the model physics.

The Thematic Mapper onboard Landsat 4, 5, and Enhanced Thematic Mapper Plus ( $\text{TM/ETM}+$ ) onboard Landsat 7 have frequency bands (green and SWIR) to effectively measure water body extents and their changes via the Modified Normalized Difference Water Index (MNDWI). Here, we developed a technique, called the thematic imagery-altimetry system (TIAS), to infer the vertical water changes from MNDWI horizontal water extent changes by integrating long-term $\text{TM/ETM}+$ imageries with available digital elevation models (DEMs). The result is a technique to quantify water level changes of natural or artificial water bodies over two decades. Several DEMs were used to compute intersects with $\text{TM/ETM}+$ water extent time series to evaluate the robustness of the technique. These DEMs include: the Advanced Spaceborne Thermal Emission and Reflection Radiometer Global Digital Elevation Map version 2 (ASTER-GDEM2, at 1 arcsec resolution), the Shuttle Radar Topography Mission version 2 (SRTM C-band at 1 arcsec), and the Global Multiresolution Terrain Elevation Data (GMTED2010 at 7.5 arcsec). We demonstrated our technique near Hoover Dam (HD) in Lake Mead to quantify its respective decadal water level changes. The dammed water had experienced extraordinary level variation in the past 20 years due to natural decline from intake or artificial impoundments. The discrepancy of the HD water level changes from an analysis of 32-year (1984–2015) time series, including 584 Landsat scenes, using the GMTED2010 DEM, has a RMSE reached $0.85 \pm 0.63\; \text{m}$ (91% of data) as compared with in situ stage record.

The traditional threat score based on fixed thresholds for precipitation verification is sensitive to intensity forecast bias. In this study, the neighborhood precipitation threat score is modified by defining the thresholds in terms of the percentiles of overall precipitation instead of fixed threshold values. The impact of intensity forecast bias on the calculated threat score is reduced. The method is tested with the forecasts of a tropical storm that re-intensified after making landfall and caused heavy flooding. The forecasts are produced with and without radar data assimilation. The forecast with assimilation of both radial velocity and reflectivity produce precipitation patterns that better match observations but have large positive intensity bias. When using fixed thresholds, the neighborhood threat scores fail to yield high scores for forecasts that have good pattern match with observations, due to large intensity bias. In contrast, the percentile-based neighborhood method yields the highest score for the forecast with the best pattern match and the smallest position error. The percentile-based method also yields scores that are more consistent with object-based verifications, which are less sensitive to intensity bias, demonstrating the potential value of percentile-based verification.

A coupled ensemble square root filter–three-dimensional ensemble-variational hybrid (EnSRF–En3DVar) data assimilation (DA) system is developed for the operational Rapid Refresh (RAP) forecasting system. The En3DVar hybrid system employs the extended control variable method, and is built on the NCEP operational gridpoint statistical interpolation (GSI) three-dimensional variational data assimilation (3DVar) framework. It is coupled with an EnSRF system for RAP, which provides ensemble perturbations. Recursive filters (RF) are used to localize ensemble covariance in both horizontal and vertical within the En3DVar. The coupled En3DVar hybrid system is evaluated with 3-h cycles over a 9-day period with active convection. All conventional observations used by operational RAP are included. The En3DVar hybrid system is run at ⅓ of the operational RAP horizontal resolution or about 40-km grid spacing, and its performance is compared to parallel GSI 3DVar and EnSRF runs using the same datasets and resolution. Short-term forecasts initialized from the 3-hourly analyses are verified against sounding and surface observations. When using equally weighted static and ensemble background error covariances and 40 ensemble members, the En3DVar hybrid system outperforms the corresponding GSI 3DVar and EnSRF. When the recursive filter coefficients are tuned to achieve a similar height-dependent localization as in the EnSRF, the En3DVar results using pure ensemble covariance are close to EnSRF. Two-way coupling between EnSRF and En3DVar did not produce noticeable improvement over one-way coupling. Downscaled precipitation forecast skill on the 13-km RAP grid from the En3DVar hybrid is better than those from GSI 3DVar analyses.

Twice-daily 48-h tropical cyclone (TC) forecasts were produced for the fall 2010 Atlantic hurricane season using the Advanced Research core of the Weather Research and Forecasting (WRF-ARW) model on a large 4-km grid covering much of the northern Atlantic. WRF forecasts initialized from operational Global Forecast System (GFS) analyses based on the gridpoint statistical interpolation (GSI) three-dimensional variational data assimilation (3DVAR) system and from experimental global ensemble Kalman filter (EnKF) analyses, and corresponding global GFS forecasts were intercompared. For the track, WRF forecasts show improvement over GFS forecasts using either set of initial conditions (ICs). The EnKF-initialized GFS and WRF are also better than the corresponding GSI-initialized forecasts, but the difference is not always statistically significant. At all lead times, the WRF track errors are comparable to or smaller than the National Hurricane Center (NHC) official track forecast error, with those of the EnKF WRF being smallest. For weaker TCs, more improvement comes from the model (resolution) than from the ICs. For hurricane intensity TCs, EnKF ICs produce better track forecasts than GSI ICs, with the best forecast coming from WRF at most lead times. For intensity, EnKF ICs consistently outperform GSI ICs in both models for weaker TCs. For hurricane-strength TCs, EnKF ICs produce forecasts statistically indistinguishable from GSI ICs in either model. For all TCs combined, WRF produces about half the error of the corresponding GFS simulation beyond 24 h, and at 36 and 48 h, the errors are smaller than those from NHC official forecasts. The improvement is even greater for hurricane-strength TCs. Overall, the WRF forecasts initialized with EnKF ICs have the smallest intensity error, and the difference is statistically significant compared to the GFS forecasts.

A regional ensemble Kalman filter (EnKF) system is established for potential Rapid Refresh (RAP) operational application. The system borrows data processing and observation operators from the gridpoint statistical interpolation (GSI), and precalculates observation priors using the GSI. The ensemble square root Kalman filter (EnSRF) algorithm is used, which updates both the state vector and observation priors. All conventional observations that are used in the operational RAP GSI are assimilated. To minimize computational costs, the EnKF is run at ⅓ of the operational RAP resolution or about 40-km grid spacing, and its performance is compared to the GSI using the same datasets and resolution. Short-range (up to 18 h, the RAP forecast length) forecasts are verified against soundings, surface observations, and precipitation data. Experiments are run with 3-hourly assimilation cycles over a 9-day convectively active retrospective period from spring 2010. The EnKF performance was improved by extensive tuning, including the use of height-dependent covariance localization scales and adaptive covariance inflation. When multiple physics parameterization schemes are employed by the EnKF, forecast errors are further reduced, especially for relative humidity and temperature at the upper levels and for surface variables. The best EnKF configuration produces lower forecast errors than the parallel GSI run. Gilbert skill scores of precipitation forecasts on the 13-km RAP grid initialized from the 3-hourly EnKF analyses are consistently better than those from GSI analyses.

In recent years, traceability systems have been developed as effective tools for improving the transparency of supply chains, thereby guaranteeing the quality and safety of food products. In this study, we proposed a cattle/beef supply chain traceability model and a traceability system based on radio frequency identification (RFID) technology and the EPCglobal network. First of all, the transformations of traceability units were defined and analyzed throughout the cattle/beef chain. Secondly, we described the internal and external traceability information acquisition, transformation, and transmission processes throughout the beef supply chain in detail, and explained a methodology for modeling traceability information using the electronic product code information service (EPCIS) framework. Then, the traceability system was implemented based on Fosstrak and FreePastry software packages, and animal ear tag code and electronic product code (EPC) were employed to identify traceability units. Finally, a cattle/beef supply chain included breeding business, slaughter and processing business, distribution business and sales outlet was used as a case study to evaluate the beef supply chain traceability system. The results demonstrated that the major advantages of the traceability system are the effective sharing of information among business and the gapless traceability of the cattle/beef supply chain.