Planetary Boundary Layer Schemes Research Articles

Evaluation of PBL parameterization schemes against direct observations during a land depression over Central India

Detailed observations from the Cloud Aerosol Interaction and Precipitation Enhancement Experiment (CAIPEEX) Phase III integrated observational campaign 2014 are used to validate six different planetary boundary layer (PBL) schemes in Weather Research and Forecasting (WRF) model during a land depression over central India in summer monsoon season. This work is a continuation of our previous study on a dry event associated with a heatwave over the same region (Sathyanadh et al. 2017). Four local and two nonlocal PBL schemes are evaluated with hourly observations of air temperature, eddy covariance fluxes, friction velocity, wind, and radiation components for the model performance of the complete energy budget and weather parameters. The models simulated a slightly warmer and drier surface layer with strong winds and overestimated sensible heat flux and ground heat flux during the wet event. The errors in the surface layer temperature are closely associated with errors in the sensible heat flux and incoming radiation, while the errors associated with moisture are linked to the excessive surface layer mixing. The interrelationships between various surface layer parameters and PBL heights are investigated to identify the key controlling factors of PBL evolution in the context of land depression.

Role of Advection on the Evolution of Near-Surface Temperature and Wind in Urban-Aware Simulations

AbstractThe role of advection of heat and momentum on the evolution of near-surface temperature and wind is evaluated in urban-aware simulations over Houston, Texas, under dry conditions on a light-wind day. Two sets of experiments, each consisting of four simulations using different planetary boundary layer (PBL) schemes, were conducted over 48 h using the default urban scheme (BULK) and the single-layer urban canopy model (SLUCM) available within the Weather Research and Forecasting Model. We focus on understanding and quantifying the role played by temperature and momentum advection, particularly on the windward and leeward sides of the city. Previous studies have largely ignored any quantitative analysis of impacts from the advection of momentum over an urban area. The horizontal advection of temperature was found to be more important in the BULK because of the larger surface temperature gradient caused by warmer surface temperatures over urban areas than in the SLUCM. An analysis of the momentum budget shows that horizontal advection of zonal and meridional momentum plays a prominent role during the period of peak near-surface winds and that this effect is more pronounced in the windward side of the city. The local tendency in peak winds in the leeward side lags that in the windward side by about 1–2 h, similar to the lag found in horizontal momentum advection. The sensitivity of the results to different urban and PBL schemes was explored. The results imply that representation and influence of land-use patterns via sophisticated urban parameterizations generate locally driven winds that best resemble observations.

Vertical Eddy Diffusivity Parameterization Based on a Large‐Eddy Simulation and Its Impact on Prediction of Hurricane Landfall

AbstractThe vertical eddy diffusivity (VED) in the planetary boundary layer (PBL) affects tropical cyclones' structure and evolution. Uncertainties of VED in the PBL parametrization scheme of a mesoscale model could be caused by the model's inability to represent large eddies (LEs). Using a large‐eddy simulation (LES) of Hurricane Harvey (2017) as a benchmark, the YSU PBL scheme is improved by adding effects of LEs derived from the LES. It is found that LEs' maximum intensity is linearly significantly related to the square of mean horizontal divergence below 400 m of height at the same air column with a slope coefficient of 1.97. The revised YSU scheme relates the VED of momentum (Km) to the square root of LEs' intensity with a slope coefficient of 0.08 in PBL and 0.20 in a free atmosphere. Compared to the original scheme, the modified YSU scheme leads to improved forecasts of landfalling hurricanes.

Parameterizing Subgrid Variations of Land Surface Heat Fluxes to the Atmosphere Improves Boreal Summer Land Precipitation Simulation With the NCAR CESM1.2

AbstractSubgrid horizontal variations of land surface heat fluxes to the atmosphere resulting from subgrid land cover heterogeneity are important in land‐atmosphere interaction in global climate models (GCMs). To incorporate it, a parameterization using stochastic sampling based on truncated normal distributions diagnosed from the land model and internal ensemble mean of multiple calls to the planetary boundary layer and deep convection schemes is developed. After its implementation in the National Center for Atmospheric Research (NCAR) Community Earth System Model version 1.2 (CESM1.2), large changes of simulated land precipitation are found to occur in regions where large subgrid variations of surface heat fluxes exist. The simulated precipitation is improved in boreal summer, especially over eastern China and the coastal areas of the Bay of Bengal. The improved precipitation is mainly a result of improved large‐scale moisture convergence and advection by altered vertical diffusion and convection.

Simulations of Indian summer monsoon using RegCM: a comparison with ERA and GFDL analysis

A comprehensive understanding of the variation of Indian summer monsoon (ISM) features is necessary for accurate prediction of seasonal and sub-seasonal scale meteorological parameters. In this study, the summer monsoon features over India are investigated using RegCM simulations over South Asia Coordinated Regional Downscaling Experiment (SA-CORDEX) domain with 45-km spatial resolution. The initial-boundary data are derived from the European Centre for Medium Range Weather Forecasts Re-Analysis (ERA) and Geophysical Fluid Dynamics Laboratory (GFDL) analysis for the period 1989 to 2005. For both model simulations, mixed convection scheme (Emmanuel over ocean and Grell over land), Holtslag planetary boundary layer (PBL) scheme and biosphere atmospheric transfer system (BATS) land surface scheme are used throughout the experiment. The simulation results are validated with Climatic Research Unit (CRU) observations of precipitation and temperature. The spatial distribution of precipitation and temperature is reasonably simulated with ERA and GFDL analysis as compared with CRU observation. However, the simulation of precipitation and temperature with ERA analysis iscloser to CRU observations as compared with GFDL analysis. The seasonal scale (JJAS) simulation of precipitation with ERA and GFDL illustrates wet bias (2 mm per day) as compared to CRU observation, whereas temperature simulated with ERA indicated warm bias (1 °C) and GFDL depicts cold bias (2 °C) against CRU observations. The wind magnitude and distribution of ERA simulations show considerably higher magnitude over the western Himalaya which is connected to low values of specific humidity over the same terrain as compared with GFDL.

Reanalysis Profile Downscaling with WRF Model and Sensitivity to PBL Parameterization Schemes Over a Subtropical Station

Atmospheric profiles are essential in weather prediction, atmospheric research, and satellite data calibration. In this study, we assess the use of the Weather Research and Forecasting (WRF) model to downscaling NCEP Climate Forecast System Version 2 (CFSv2) reanalysis atmospheric profiles. Moreover, a sensitivity analysis was conducted to Yonsei University (YSU) and Mellor–Yamada–Janjic (MYJ) Planetary Boundary Layer (PBL) parameterization schemes. WRF simulations were carried out using two nested grids with horizontal resolutions of 12 and 3 km. Simulated profiles were evaluated against radiosonde observations in Southern Brazil. Results indicate that WRF can skillfully simulate vertical profiles of water vapor mixing ratio ( q ) and potential temperature ( θ ). The entire profile comparison yielded high correlation coefficients (mostly higher than 0.9), low moist and cold average biases, and overall RMSE (MAE) of 0.84 (0.44) g/Kg and 2.30 (1.61) K for q and θ , respectively. In the lower troposphere (0-3 km), errors are larger for q and lower for θ . Statistically, there is no significant difference in the retrieved profiles on increasing the horizontal resolution. The setting choice is a compromise between spatial detail needs and computational cost. Furthermore, the overall results did not indicate any preferred PBL scheme that outperforms in all case days. The performance depends on the meteorological conditions that prevail in the different cases. The WRF model and even data directly from NCEP CFSv2 reanalysis are useful to represent the vertical structure of the atmosphere when local radiosondes are not available.

Impact of WRF Parameterization Schemes on Track and Intensity of Extremely Severe Cyclonic Storm “Fani”

Weather Research and Forecasting (WRF) model version 4.1.3 is configured in the present study to understand the impact of microphysical parameterization schemes on the track and intensity of extremely severe cyclonic storm (ESCS) “Fani” that occurred during 26 April to 4 May 2019 in the Bay of Bengal. The model is customized with six different microphysical parameterization schemes along with Kain–Fritsch cumulus for convection and the Yonsei University planetary boundary-layer scheme for computation of heat and moisture transport. The simulations are conducted at 27-km horizontal resolution with 41 levels in the vertical for 150 h (00UTC 28 April to 06UTC 4 May 2019). Model-simulated features such as the track, track error, minimum central sea-level pressure (MSLP), maximum sustained surface wind (MSW), and precipitation are validated against India Meteorological Department (IMD) and Tropical Rainfall Measuring Mission (TRMM) observation data. Although all the schemes predict track patterns similar to observation, the MP3 scheme shows the least track error even after landfall, and its track length is maximum in comparison with other schemes, albeit less than the observed track. The MP3 scheme exhibits a better track for “Fani,” followed by MP14. The lowest mean direct position error (along-track error) of 85 km (68.5 km) is noted for the MP3 scheme, followed by 112.5 km (82.3 km) for the MP8 scheme. Analysis of the MSLP and MSW using the t-test suggests that the MP2 scheme has better forecast ability, followed by MP6. Analysis of the quantitative/spatial matching of the rainfall forecasts and TRMM observations using the Method for Object-Based Diagnostic Evaluation (MODE) tool suggests that the MP8 (MP14) scheme is capable of simulating the rainfall beyond (up to) 48 h. Analysis of the zonal cross-section of horizontal and vertical winds shows that intense convection is well simulated by the MP2, MP3, and MP4 schemes during the change of intensity from cyclonic storm to severe cyclonic storm (SCS), by MP6 and MP14 from SCS to very severe cyclonic storm (VSCS) and from VSCS to ESCS, and by MP8 at ESCS level. All the schemes can capture the eyewall and simulate the westerly that allowed the system to move north-northeastwards. The synoptic flow at different pressure levels is reproduced by all the schemes. However, assessing the impact of the microphysics parameterizations on the synoptic flow of “Fani” at low resolution was difficult. Analysis of Qc and Qr indicates definite impacts of the microphysical parameterizations on the vertical structure of “Fani” in producing rainfall. The MP3 scheme exhibits a better cloud pattern up to upper troposphere level, followed by the MP2 scheme. Analysis of the latent heat flux predicted using the different microphysics schemes suggests that the latent heat flux release at VSCS intensity level or above contributes to further intensification of the TC, and this feature is very well simulated by the MP2 scheme.

High-resolution Simulation of an Extreme Heavy Rainfall Event in Shanghai Using the Weather Research and Forecasting Model: Sensitivity to Planetary Boundary Layer Parameterization

In this study, an extreme rainfall event that occurred on 25 May 2018 over Shanghai and its nearby area was simulated using the Weather Research and Forecasting model, with a focus on the effects of planetary boundary layer (PBL) physics using double nesting with large grid ratios (15:1 and 9:1). The sensitivity of the precipitation forecast was examined through three PBL schemes: the Yonsei University Scheme, the Mellor–Yamada–Nakanishi Niino Level 2.5 (MYNN) scheme, and the Mellor–Yamada–Janjic scheme. The PBL effects on boundary layer structures, convective thermodynamic and large-scale forcings were investigated to explain the model differences in extreme rainfall distributions and hourly variations. The results indicated that in single coarser grids (15 km and 9 km), the extreme rainfall amount was largely underestimated with all three PBL schemes. In the inner 1-km grid, the underestimated intensity was improved; however, using the MYNN scheme for the 1-km grid domain with explicitly resolved convection and nested within the 9-km grid using the Kain–Fritsch cumulus scheme, significant advantages over the other PBL schemes are revealed in predicting the extreme rainfall distribution and the time of primary peak rainfall. MYNN, with the weakest vertical mixing, produced the shallowest and most humid inversion layer with the lowest lifting condensation level, but stronger wind fields and upward motions from the top of the boundary layer to upper levels. These factors all facilitate the development of deep convection and moisture transport for intense precipitation, and result in its most realistic prediction of the primary rainfall peak.

Sensitivity Study of Planetary Boundary Layer Parameterization Schemes for the Simulation of Tropical Cyclone ‘Fani’ Over the Bay of Bengal Using High Resolution Wrf-Arw Model

Comprehensive sensitivity analyses on physical parameterization schemes of WRF-ARW (V3.8.1) model have been carried out for the simulation of Tropical Cyclone (TC) Fani that formed in the Bay of Bengal (BoB) and crossed the Bangladesh and Odisha coast of India on 3rd May 2019. Global Forecasting System (GFS) data 1⁰ and 0.25⁰ from National Centre for Environment Prediction (NCEP) is used as initial and lateral boundary conditions. The six different Planetary Boundary Layer (PBL) schemes used in this research are YSU, BouLac, TEMF, Shin-Hong, GBM and MRF. The meteorological parameters, which have been studied to identify the effect of PBL during the propagation and movement of TC Fani are Estimated Central Pressure (ECP), Maximum Wind Speed (MWS) at 10m height, average relative humidity (%), temperature (⁰C) and potential temperature (0K) at 2m height, Modified Convective Available Potential Energy (MCAP), average PBL height and average high clouds (%). The area considered for these averages are 82-92 ⁰E and 7-22 ⁰N inside the model domain. The simulated ECP by TEMF scheme are 930, 932, 937, 929, 944 and 932 hPa for the Initial Conditions (ICs) at 0000 UTC of 27, 28, 29, 30 April, and 01 May and 02 May, respectively and observed ECP was 932 hPa. The intensity of pressure fall by the TEMF scheme is similar to that observed up to the time of crossing the land of TC Fani. The MWS simulated by the TEMF scheme is almost similar to that of the observed MWS at 10m height and all other schemes have simulated much lower MWS. The temperature at 2m height is positively correlated with the ECP and MWS at 10m height. The TEMF scheme has simulated maximum high clouds for all ICs and for all through the simulation time. The error was systematic for all PBL schemes for the ICs at 0000 UTC of 30 April at 0.25⁰ and 1⁰ GFS data but the track error was much less for 1⁰ GFS data than that of 0.25⁰ GFS data. The TEMF scheme has simulated the most deviated track and MRF scheme has simulated less deviated track for all through the simulation. The study has shown large variations of track and intensity among the different PBL schemes. The PBL schemes have a major impact on the track and intensity of TC Fani. The intensity simulated by the TEMF scheme is better but the track error is higher than that of other schemes.
 Journal of Engineering Science 11(2), 2020, 1-18

Predictability of Orographic Rainfall Features Over Sylhet Using Nwp Model

Topography and orography are two physical factors which produce high impact rainfall over the North-eastern part of Bangladesh. To predict the orographic rainfall of 29 March 2017 over Sylhet, Bangladesh an attempt has been performed using Weather Research and Forecasting (WRF) model. The model has been run in a single domain of 10 km horizontal resolution for 48-h and 72-h using six hourly global final datasets from 0000 UTC of each initial day of the event as initial and lateral boundary conditions with NSSL 2-moment microphysics scheme, Kain–Fritsch cumulous scheme and Yonsei University Planetary Boundary Layer (PBL) scheme. The model outputs such as sea level pressure, wind flow, vorticity, wind shear, humidity, Convective Available Potential Energy (CAPE), Convective Inhibition, Lifted Index, K-index, Total Total Index and rainfall have been analyzed. The model predicted weather parameters were visualized by Grid Analysis and Display System (GrADS) and Geographic Information System (GIS) software and validated with observed data of Bangladesh Meteorological Department (BMD), Tropical Rainfall Measuring Mission (TRMM) and European Centre for Medium-Range Weather Forecasts (ECMRWF) data. The analysis determines that the CAPE of magnitude 800- 1000 JKg-1, positive vorticity of (6-10) ×10-5s-1 and relative humidity of 80-100% up to 500-400 hPa levels are accountable for the happening of the orographic extreme rainfall and other parameters are compatible with the observed or theoretical values. This study indicates that the model with an appropriate model set up is capable to predict the orographic precipitation realistically well and can be used for upcoming events.
 Journal of Engineering Science 11(2), 2020, 61-73

Simulation analysis of local land atmosphere coupling in rainy season over a typical underlying surface in the Tibetan Plateau

Abstract. The local land–atmosphere coupling (LoCo) investigates the interactions between soil conditions, surface fluxes, planetary boundary layer (PBL) growth, and the formations of convective clouds and precipitation. Studying LoCo over the Tibetan Plateau (TP) is of great significance for understanding the TP's role in the Asian water tower. A series of real-case simulations, using the Weather Research and Forecasting (WRF) model with different combinations of land surface model (LSM) schemes and PBL schemes, has been carried out to investigate the LoCo characteristics over a typical underlying surface in the central TP in the rainy season. The LoCo characteristics in the study area are analyzed by applying a mixing diagram to the simulation results. The analysis indicates that the WRF simulations, using the Noah with BouLac, Mellor-Yamada Nakanishi and Niino Level-2.5 PBL (MYNN), and Yonsei University (YSU) produce closer results to the observation in terms of curves of Cp⋅θ and Lv⋅q, surface fluxes (Hsfc and LEsfc), entrainment fluxes (Hent, and LEent) at site BJ of Nagqu Station (BJ/Nagqu) than those using the Community Land Model (CLM) with BouLac, MYNN, and YSU. The frequency distributions of Hsfc, LEsfc, Hent, and LEent in the study area confirm this result. The spatial distributions of simulated Hsfc, LEsfc, Hent, and LEent, using WRF with Noah and BouLac, suggest that the spatial distributions of Hsfc and LEsfc in the study area are consistent with that of soil moisture, but the spatial distributions of Hent and LEent are quite different from that of soil moisture. A close examination of the relationship between entrainment fluxes and cloud water content (QCloud) reveals that the grids with small Hent and large LEent tend to have high QCloud and Hsfc, suggesting that high Hsfc is conducive to convective cloud formation, which leads to small Hent and large LEent. A sensitivity analysis of LoCo to the soil moisture at site BJ/Nagqu indicates that, on a sunny day, an increase in soil moisture leads to an increase in LEsfc but decreases in Hsfc, Hent, and LEent. The sensitivity of the relationship between simulated maximum daytime PBL height (PBLH) and mean daytime evapotranspiration (ET) in the study area to soil moisture indicates the rate at which the maximum daytime PBLH decreases with the mean ET increase as the initial soil moisture goes up. The analysis of simulated Hsfc, LEsfc, Hent, and LEent under different soil moisture conditions reveals that the frequency of Hent ranging from 80 to 240 W m−2 and the frequency of LEent ranging from −240 to −90 W m−2 both increase as the initial soil moisture increases. Coupled with the changes in QCloud, the changes in Hent and LEent as the initial soil moisture increases indicate that the rise in soil moisture leads to an increase in the cloud amount but a decrease in QCloud.

Comparison of Urban Heat Island Intensity Estimation Methods Using Urbanized WRF in Berlin, Germany

In this study, we present a meso-scale simulation of the urban microclimate in Berlin, Germany, using the Weather Research and Forecasting (WRF) numerical weather prediction platform. The objective of the study is to derive an accurate estimate of the near-surface urban heat island (UHI) intensity. The simulation is conducted over a two-week summer period. We compare different physical schemes, different urban canopy schemes and different methods for estimating the UHI intensity. The urban fraction of each urban category is derived using the Copernicus Impervious Density data and the Corine Land Cover data. High-resolution City Geography Markup Language (CityGML) data is used to estimate the building height densities required by the multi-layer urban canopy model (UCM). Within the single-layer UCM, we implement an anthropogenic heat profile based on the large scale urban consumption of energy (LUCY) model. The optimal model configuration combines the WRF Single Moment Five-Class (WSM5) microphysics scheme, the Bougeault–Lacarrère planetary boundary layer scheme, the eta similarity (Mellor–Yamada–Janjic) surface layer scheme, the Noah Multi-Parameterization land surface model, the Dudhia and Rapid Radiative Transfer Model (RRTM) radiation schemes, and the multi-layer UCM (including the building energy model). Our simulated UHI intensity results agree well with measurements with a root mean squared error of 0.86K and a mean bias error of 0.20K. After model validation, we proceed to compare several UHI intensity calculation methods, including the ‘ring rural reference’ (RRR) method and the ‘virtual rural reference’ (VRR) method. The VRR mthod is also known as the ‘urban increment’ method. We suggest and argument that the VRR approach is superior.

Large eddy simulations of the Martian convective boundary layer: Towards developing a new planetary boundary layer scheme

The grid resolutions typically employed in atmospheric global circulation and mesoscale models are not sufficient to explicitly resolve the turbulence processes within the planetary boundary layer (PBL). turbulent fluxes are, therefore, fully parameterized in those models, based on empirical relationships for the mixing length scale using PBL schemes. However, microscale models use the large-eddy simulation (LES) technique to resolve turbulence processes. Here, we perform LES computations for the daytime Martian planetary boundary layer, ranging from weakly to strong convective conditions, using the Mars implementation of planetWRF, the MarsWRF model. In this study, our main focus is to investigate the turbulence statistics and turbulent spectrum utilizing our LES results. Then, using the computed turbulence kinetic energy and its dissipation rate, a generic formulation for the mixing length scale variation in the Martian convective boundary layer is proposed. This mixing length formulation is used to derive a Mars-specific PBL scheme and its performance is compared to the PBL scheme currently in use by the MarsWRF model, the MRF scheme. The proposed scheme is tested both in global and mesoscale simulations, which are used to evaluate the convective boundary layer height and near-surface meteorology conditions at the InSight landing site. The presently proposed PBL scheme results in an improved prediction of convective boundary layer height that agree better with the observational estimations acquired by radio occultations of Mars Express in comparison to the MRF scheme. Also, the prediction of near-surface winds at the InSight landing site is slightly improved.

Gray Zone Partitioning Functions and Parameterization of Turbulence Fluxes in the Convective Atmospheric Boundary Layer

AbstractHere, we present the first attempt to fully represent three‐dimensional turbulence fluxes in the “Terra Incognita” or the gray zone in other words. In order to derive partitioning functions, representing the partitioning between subgrid and total fluxes, we make use of high‐resolution large‐eddy simulations (LES), which are performed with the Weather Research and Forecasting (WRF) model. LES computations are performed for various levels of convective instability, ranging from pure buoyant to strongly sheared convection. Then, the resulting reference‐LES fields are successively coarse grained from its original microscale grid spacing ( m) up to typical mesoscale grid spacings ( km). The given process is applied by means of an advanced filter, that is, the Butterworth filter. It enables a clear scale‐specific filtering that results in a more controlled energy transition from lower to higher wavenumbers, unlike the drawbacks of current filters in use. Finally, we parameterize the subgrid scale (SGS) partitioning functions of 10 SGS turbulence quantities: momentum fluxes (τij, six terms), heat fluxes (qj, three terms), and turbulence kinetic energy (k). Turbulence partitioning relations are parameterized in a scale‐aware, stability‐dependent, and height‐dependent form, using the sigmoidal Gompertz function. Thus, the new gray zone model provides a framework that bridges the mesoscale and microscale limits and that is suitable for the development of next generation three‐dimensional, multiscale turbulence parameterization methods or planetary boundary layer schemes.

The role of atmospheric forcings and WRF physical set-up on convective initiation over Córdoba, Argentina

The accurate determination of the location and timing of convective initiation (CI) is a crucial aspect of forecasting and high impact weather warning systems associated with deep convection. Central Argentina, particularly the Sierras de Córdoba (SDC), is characterized by the regular occurrence of deep convective cells that reach supercell characteristics producing large hail and occasionally tornadoes.In order to analyze the sensitivity of the CI over SDC, an ensemble of 12 simulations with the WRF model at convection-permitting resolution (3 km.) was carried out to study a marginal supercell initiated on October 17, 2017 at the SDC foothills. The ensemble was built using 3 different microphysics (MP) parameterizations (WSM6, Thompson and Morrison) and 2 planetary boundary layer (PBL) parameterizations (Yonsei University and Mellor Yamada Janjic). Also, the sensitivity of initial and boundary conditions is considered by forcing the model with 2 different global analyses: the European Centre for Medium-Range Weather Forecast (ECMWF) model, and the Global Data Assimilation System (GDAS).The two different forcing data show low-level moisture differences, which later modulate the life-cycle of the simulated cells. A tracking algorithm was used to capture their life-cycle, and CI was detected at both the hills and the plains. Simulations forced with ECMWF analyses tend to show better results with characteristics closer to the ones observed including: initiation over the eastward slope of SDC, stronger cells and slow propagation. Among all simulations analyzed, the best performance was obtained with the combination of Thompson MP and Yonsei University PBL schemes.

The making of the New European Wind Atlas – Part 1: Model sensitivity

Abstract. This is the first of two papers that document the creation of the New European Wind Atlas (NEWA). It describes the sensitivity analysis and evaluation procedures that formed the basis for choosing the final setup of the mesoscale model simulations of the wind atlas. The suitable combination of model setup and parameterizations, bound by practical constraints, was found for simulating the climatology of the wind field at turbine-relevant heights with the Weather Research and Forecasting (WRF) model. Initial WRF model sensitivity experiments compared the wind climate generated by using two commonly used planetary boundary layer schemes and were carried out over several regions in Europe. They confirmed that the most significant differences in annual mean wind speed at 100 m a.g.l. (above ground level) mostly coincide with areas of high surface roughness length and not with the location of the domains or maximum wind speed. Then an ensemble of more than 50 simulations with different setups for a single year was carried out for one domain covering northern Europe for which tall mast observations were available. We varied many different parameters across the simulations, e.g. model version, forcing data, various physical parameterizations, and the size of the model domain. These simulations showed that although virtually every parameter change affects the results in some way, significant changes in the wind climate in the boundary layer are mostly due to using different physical parameterizations, especially the planetary boundary layer scheme, the representation of the land surface, and the prescribed surface roughness length. Also, the setup of the simulations, such as the integration length and the domain size, can considerably influence the results. We assessed the degree of similarity between winds simulated by the WRF ensemble members and the observations using a suite of metrics, including the Earth Mover's Distance (EMD), a statistic that measures the distance between two probability distributions. The EMD was used to diagnose the performance of each ensemble member using the full wind speed and direction distribution, which is essential for wind resource assessment. We identified the most realistic ensemble members to determine the most suitable configuration to be used in the final production run, which is fully described and evaluated in the second part of this study (Dörenkämper et al., 2020).

Evaluation and improvement study of the Planetary Boundary-Layer schemes during a high PM2.5 episode in a core city of BTH region, China

Accurate depictions of planetary boundary layer (PBL) processes are important for both meteorological and air quality simulations. This study examines the sensitivity of the model performance of the Weather Research Forecasting model coupled with Chemistry (WRF-Chem) to five different PBL schemes and further to different turbulence parameters for the simulation of a winter haze episode in Tianjin, a core city of the Beijing-Tianjin-Hebei (BTH) region in China. To provide a direct and comprehensive evaluation of the PBL schemes, measurements from multiple instruments are employed, including both meteorological and air quality quantities from near-surface observations, vertical sounding measurements and ceilometer data. Moreover, the vertical distribution of the turbulent exchange coefficient is derived from sounding measurements and is utilized to evaluate the PBL schemes. The results suggest that the Mellor-Yamada-Janjic (MYJ) scheme is generally statistically superior to the other schemes when comparing observations. However, considerable model discrepancies still exist during certain stages of this haze episode, which are found to be predominantly due to the deficiency of MYJ in distinguishing the intensity of turbulent mixing between different pollution stages. To improve the model performance, this study further tests the impact of different closure parameters on the simulation of winter haze episode. In the MYJ scheme, the closure parameters play a key role in the turbulent mixing within the PBL and therefore in haze simulations. Sensitivity experiments with different MYJ parameters confirm this diagnosis and suggest that a larger Prandtl number (Pr), rather than the default value in the MYJ formulation, may be more applicable for haze simulations under stable atmospheric conditions.

A Review and Evaluation of Planetary Boundary Layer Parameterizations in Hurricane Weather Research and Forecasting Model Using Idealized Simulations and Observations

This paper reviews the evolution of planetary boundary layer (PBL) parameterization schemes that have been used in the operational version of the Hurricane Weather Research and Forecasting (HWRF) model since 2011. Idealized simulations are then used to evaluate the effects of different PBL schemes on hurricane structure and intensity. The original Global Forecast System (GFS) PBL scheme in the 2011 version of HWRF produces the weakest storm, while a modified GFS scheme using a wind-speed dependent parameterization of vertical eddy diffusivity (Km) produces the strongest storm. The subsequent version of the hybrid eddy diffusivity and mass flux scheme (EDMF) used in HWRF also produces a strong storm, similar to the version using the wind-speed dependent Km. Both the intensity change rate and maximum intensity of the simulated storms vary with different PBL schemes, mainly due to differences in the parameterization of Km. The smaller the Km in the PBL scheme, the faster a storm tends to intensify. Differences in hurricane PBL height, convergence, inflow angle, warm-core structure, distribution of deep convection, and agradient force in these simulations are also examined. Compared to dropsonde and Doppler radar composites, improvements in the kinematic structure are found in simulations using the wind-speed dependent Km and modified EDMF schemes relative to those with earlier versions of the PBL schemes in HWRF. However, the upper boundary layer in all simulations is much cooler and drier than that in dropsonde observations. This model deficiency needs to be considered and corrected in future model physics upgrades.

Bay of Bengal cyclones Mora and Maarutha in regional atmospheric modeling system

The coastal region of India is exposed to nearly 7% of tropical cyclones occurring in the world. The large basin of Bay of Bengal incites cyclones with varying intensity that is capable enough to cause huge damage to life and money. To understand the underlying mechanism behind these TCs and enhance their forecast accuracy, it is essential to simulate them precisely using different numerical models by applying appropriate micro-physics, Cumulus Convection (CC) and Planetary Boundary Layer (PBL) schemes. The present study uses the Regional Atmospheric Modeling System (RAMS) to simulate TCs Mora and Maarutha in the Bay of Bengal region and associated landfall in Northeast India and Bangladesh. The study is an approach to investigate and analyze winds, precipitation and other dynamical variables associated with the formation of these cyclones by providing initial and boundary conditions from the NCEP-FNL (Final) operational global analysis data (6 hourly temporal and 1.0° spatial resolution). The model output is further nudged towards the observational soundings which are obtained from NCAR-ADP upper air observational dataset. Along with this, an attempt has been made to obtain the optimum configuration of distribution and intensity of a precipitation field to match observations.

Assimilation of the Rain Gauge Measurements Using Particle Filter

AbstractThe well‐recognized constraint of nonlinear and non‐Gaussian distribution of rainfall observation limits its assimilation in the high‐dimensional numerical weather prediction (NWP) model. In this study, rainfall observed from Indian Meteorological Department (IMD) rain gauges over Indian landmass is assimilated in the Weather Research and Forecasting (WRF) model using particle filter. In the framework of imperfect weather models, particles (or ensembles) for rainfall predictions are created with various combinations of model physics (viz., cumulus parameterization, microphysics and planetary boundary layer schemes). The multiple hypotheses are used to determine the weights for different particles, and this is the step where IMD rainfall data are used for assimilation. Further, a resampling step is performed to generate new particles from high weight particles using stochastic kinetic‐energy backscatter scheme (SKEBS) method in which dynamical variables are perturbed into the model physics. Results, based on rainfall verification scores, suggest that the assimilation of the rainfall using particle filter could improve the prediction of rainfall over CNT runs (unweighted particles; without assimilation). Moreover, surface and vertical profile of temperature, water vapor mixing ratio (WVMR), and wind speed are also improved in 24‐hr forecasts.