1. Introduction
The planetary boundary layer (PBL) is the lowest layer of the atmosphere, which extends from as low as 5–10 m to about 3 km (Mahrt and Acevedo 2023; Stull 1988). The PBL is the part of the troposphere that is directly affected by the presence of Earth’s surface, dominated by the rapid exchange of heat, moisture, and momentum with a time scale of an hour or less (Stull 1988). The PBL height (PBLH) is one of the most critical parameters controlled among others by the surface fluxes of buoyancy and moisture that influence the interactions between the surface and the atmosphere (Edwards et al. 2020). Therefore, accurate measurements and modeling of PBLH are crucial to the studies of air quality, weather forecasting, and climate change (Allabakash and Lim 2020; Meng et al. 2023; Qu et al. 2017; Shi et al. 2021; Zhang and Li 2019). In addition, Johns and Doswell (1992) indicate that a proper representation of moisture, vertical lift, atmospheric stability, and wind shear, which mainly depend on the transport of moisture, momentum, and heat fluxes within the PBL, plays an important role in accurately simulating thunderstorms. The variations in mixed layer and moist-layer depths also have an impact on modeled storm structure and intensity (McCaul and Cohen 2002). Furthermore, the structure and evolution of thermodynamic and kinematic parameters in the PBL are largely affected by turbulence, which is not resolved directly by most of the atmospheric models. Instead, the net effect of the turbulence motions is estimated by PBL parameterizations (Stull 1988; Holton 2004). While great advances have been made with regard to boundary layer meteorology in the past 100 years, in terms of theory, observations, modeling, and computing (LeMone et al. 2019), numerous studies have still shown poor representation of PBL structure. Improving our understanding and characterization of the PBL has been recognized to be among the highest needs in many national and international documents including the recent National Academies of Sciences, Engineering, and Medicine (NASEM) 2017–27 decadal survey for Earth Science and Applications from Space (ESAS 2017; see NASEM 2018).
Having a robust model representation of PBL is still elusive even though PBL parameterizations in numerical weather forecast models have been widely studied across different regions of the world. Studies on these regions, which typically exhibit characteristic spatial and temporal heterogeneity and complexity in PBL properties and their interactions, are essentially distributed across North America, Europe, and Asia for the past decades (Jia and Zhang 2020), including studies focusing on the tropics (Hariprasad et al. 2014; Xie et al. 2012; Zhong et al. 2007) and midlatitude (Avolio et al. 2017; Borge et al. 2008; Cohen et al. 2015; García-Díez et al. 2013; Hu et al. 2010; LeMone et al. 2013; Milovac et al. 2016; Shin and Dudhia 2016; Shin and Hong 2011). Apart from a geographical focus, studies also pay attention to complex surfaces, like mountainous terrain (Banks et al. 2016) and urban areas (Lopez-Coto et al. 2020). Because the PBL turbulence and mean thermodynamic structure interact with each other and the PBL thermodynamics change throughout the day (Teixeira et al. 2021), model performance is strongly dependent on the stage of the PBL development, prevailing atmospheric conditions, specific weather events, geographical location, season, and complexity of the surface. As can be expected, these studies also have reported (albeit separately) that simulations of PBL are influenced by several factors including the choice of PBL schemes, land surface model, radiative transfer schemes, and model grid spacing. Each of these model configurations has unique strengths and weaknesses that depend on domain applications, precluding us at this point from having a unified model configuration that performs well in all cases. Moreover, studies have also reported that the performance of modeled PBL is sensitive to the vertical grid spacing. They noted that each PBL scheme has its own best-performing grid spacing (Shin and Dudhia 2016) and that for some PBL schemes, increasing the number of vertical levels improves the structures of PBL though it is compromised by model stability and computational cost (Smith et al. 2018).
The main goal of this study is to evaluate the performance of modeled PBLH and related key surface variables (2-m temperature, water vapor mixing ratio, heat fluxes, soil temperature, and moisture) and atmospheric profiles (temperature, water vapor, and wind) at southern Great Plains (SGP) during the Plains Elevated Convection at Night (PECAN) campaign by running a numerical weather prediction (NWP) model using multiple physics configurations (see section 2 for more details). The rationale for using multiphysics is that previous studies focusing on PBLH data assimilation (Dang et al. 2022; Tangborn et al. 2021) as well as the Data Assimilation Research Testbed (DART; Anderson et al. 2009) typically generate model ensembles by only perturbing the initial conditions (ICs). However, as we discussed earlier, the performance of a particular PBL scheme depends on application and regime. Here, we show that using different model physics (i.e., PBL schemes) can also generate sufficient spread in the model state variables aside from perturbations in initial and boundary conditions in accounting for model errors. More importantly, we evaluate the bias for each model simulation and quantify the bias due to the choice of schemes. We choose the field observational study of the PECAN (Geerts et al. 2017) campaign from June to July 2015 in the southern Great Plains, given the availability of measurements and unique scientific focus of the campaign, including new insights because of several data assimilation studies on these campaign measurements (Chipilski et al. 2020; Coniglio et al. 2019; Degelia et al. 2019; Johnson and Wang 2017; Johnson et al. 2017; Tangborn et al. 2021). LeMone et al. (2013) conducted a study focusing on modeled convective boundary layer (CBL) depths on five different fair weather days and evaluated them with observations from the 1997 Cooperative Atmosphere–Surface Exchange Study (CASES-97). Our study uses a similar approach for retrieving PBLH using atmospheric profiles across different modeled PBL schemes. In addition, our study 1) analyzed the nighttime stable boundary layer (SBL) height; 2) added Mellor–Yamada–Nakanishi–Niino eddy-diffusivity mass flux (MYNN-EDMF) (Olson et al. 2019), which is an updated PBL scheme; 3) conducted sensitivity analysis including vertical grid spacing and other model physics schemes (refer to sections 2a and 2b); 4) included cases with large fraction of cloud coverages; 5) included lidar observations that determine PBLH using vertical velocity variance (
2. Materials and methods
a. WRF Model configurations
We use the Advanced Research version of WRF (ARW; Skamarock et al. 2019), to simulate the meteorological conditions and PBL characteristics in the SGP region during PECAN. The model setup contains three domains with a horizontal grid spacing of 10 km (d01: 200 × 160 grids), 3.33 km (d02: 301 × 250 grids), and 1.11 km (d03: 349 × 319 grids). The vertical grid spacings are set to 57 and 77 levels with 13 and 33 levels (∼230 and ∼90 m) below the lowest (i.e., nearest to the surface) 30% of the pressure column, respectively; 5 and 12 levels (∼100 and ∼40 m) below the lowest 5% of the column, respectively; and model top pressure (Ptop) set to 50 hPa. Note that we are not pursuing high vertical grids over the entire pressure column since our study mainly focuses on the PBL. Nevertheless, our 77-level runs have similar grid spacing for the lower part of the troposphere as the 137-level (32 below ∼700 mb; 1 mb = 1 hPa) European Centre for Medium-Range Weather Forecasts (ECMWF) model (Hersbach et al. 2020). The domain configuration and location of measurement sites used in this work, as well as the vertical pressure levels, are shown in Fig. 1. The initial and boundary conditions (ICBCs) are provided by National Centers for Environmental Prediction (NCEP) Global Forecast System (GFS; NCEP 2015) analysis with a spatial grid spacing of 0.25° × 0.25°, 26 pressure levels, and 3-hourly interval and fifth major global reanalysis produced by ECMWF (ERA5; ECMWF 2019; Hersbach et al. 2020) with spatial grid spacing of 31 km, 37 pressure levels, and hourly interval. The WRF Model simulations are conducted for the entire campaign period from 1 June to 15 July 2015. We run 3-day forecasts for each simulation and allocate a 6-h spinup time ahead of each run. We then stitch fifteen 3-day forecasts together to perform our evaluation. The forecasts are initialized by either ERA5 or GFS with WRF output files archived at hourly resolution. We used “cold” starts for each WRF run (instead of grid or observation nudging) since our analysis is intended for short-range forecasts (e.g., similar to Parish and Clark 2017; Smith et al. 2018, 2019), leading us not to pursue nudging. Studies have reported a negative impact on modeled PBL properties, including surface meteorological variables and the structure of inversion (Tran et al. 2018), when water vapor nudging is applied above PBL and when nudging is used below the model-calculated PBLH (Dzebre et al. 2019).
Map illustrating the one-way nested WRF domains and vertical grid spacing (below 700 mb) for this study. The innermost domain (d03) covers the location of FP measurement sites in Kansas and Oklahoma, while the largest domain (d01) extends to their neighboring states. The “57” and “77” denote two vertical grids used in this study, while “137” refers to the ECMWF with 137 pressure levels in total. Note that while the WRF Model uses terrain following pressure coordinates, here we present an example of its nominal pressure levels assuming surface pressure is 1013 mb.
Citation: Journal of Applied Meteorology and Climatology 63, 11; 10.1175/JAMC-D-24-0046.1
The physical parameterizations of the WRF Model are set to the “CONUS” suite for our reference run (See Table 1). The CONUS physical suite has been found to be skillful and robust in WRF forecasting applications in midlatitude and continental convection (Powers et al. 2017 and references therein) and has been used for the simulation over the SGP region (e.g., Johnson et al. 2016). We then create an ensemble of model configurations by selecting one model representation at a time. The ensemble includes configuration classes related to the following: 1) ICBC, 2) microphysics scheme, 3) PBL scheme, and 4) vertical grids (see Table 1). Note that we only turn on the cumulus scheme in domain 1 (10 km), while we turn it off for domains 2 and 3 (3.3 and 1.1 km, respectively) as the spatiotemporal scale of these domains allows for a convective-permitting simulation. The key difference between the selected two microphysics schemes [Thompson graupel microphysics scheme (Thompson et al. 2004, 2008) and Morrison double moment schemes (Morrison et al. 2009)] used in this study is mainly the parameterization of hydrometers (Bao et al. 2019), which can affect the cloud coverage and precipitation and indirectly affect surface radiation. More importantly, three PBL schemes of the WRF Model are explored in this study, including one nonlocal scheme [Yonsei University (YSU); Hong et al. 2006], one local closure scheme (MYJ; Janjić et al. 1994), and MYNN (Olson et al. 2019) scheme. These three PBL schemes are widely used in previous studies as we mentioned in section 1 (Avolio et al. 2017; Banks et al. 2016; Borge et al. 2008; Cohen et al. 2015; García-Díez et al. 2013; Hariprasad et al. 2014; Hu et al. 2010; LeMone et al. 2013; Lopez-Coto et al. 2020; Milovac et al. 2016; Shin and Dudhia 2016; Shin and Hong 2011; Smith et al. 2018; Tangborn et al. 2021; Xie et al. 2012; Zhong et al. 2007). In total, we run the WRF Model with 24 different combinations of schemes, ICBC, and vertical grids. We provide a brief overview of the selected PBL schemes with emphasis on how PBLH is determined in these schemes and their associated surface layer schemes in the following section.
Model configurations.
b. Brief description of WRF PBL schemes
The PBL schemes, along with the coupled surface scheme in the WRF Model, affect the heat, moisture, and momentum fluxes, as well as how these fluxes are transported within the PBL, and control the vertical diffusion and entrainment at the PBL top (Dudhia 2010; Skamarock et al. 2019). The role of land surface in land–atmosphere interactions has been discussed extensively in Santanello et al. (2018). The main difference between nonlocal (YSU) and local (MYJ) closure schemes is the extent of the region that can affect the PBL parameterizations at a given grid cell. In local closure schemes, only the neighbor levels at a given vertical level can affect the parameterizations directly at that point. However, for nonlocal schemes, multiple vertical layers can affect the variable at a given vertical level (Stensrud 2009).
The YSU scheme (Hong et al. 2006) is a revised PBL scheme based on Hong and Pan (1996) with the inclusion of an explicit treatment of the entrainment process at the top of PBL. The YSU scheme in the WRF Model can simulate the PBL structure and its development in a more realistic way. In the YSU scheme, the PBLH is defined by the first level with a bulk Richardson (Rib) number exceeding the threshold of 0. The YSU PBL scheme in this study is paired with the revised nonhydrostatic mesoscale meteorological model (MM5) Monin–Obukhov (Jiménez et al. 2012) surface layer scheme.
The MYJ scheme (Janjić 1994) is based on the Mellor–Yamada (M–Y) level 2.5 scheme (Mellor and Yamada 1982) with developments of robust, accurate, consistent, and affordable computational procedures for the application in atmospheric models. The MYJ scheme solves the turbulent kinetic energy (TKE) production and dissipation equations iteratively over a time step while keeping the master length scale constant. The master length scale refers to the mixing length scale, which is the harmonic mean of several length-scale contributions as discussed by Muñoz-Esparza et al. (2020). The PBLH is determined by the lowest model level at which the production of TKE is outbalanced by the dissipation of TKE or, in other words, at the level at which the TKE decreases to a preset value of 0.1 m2 s−2. The MYJ PBL scheme is used in the CONUS physical suite and paired with the Monin–Obukhov–Janjic (MOJ) surface layer scheme (Janjić et al. 1994) in our study.
In the WRF, version 4.4, the original MYNN scheme is updated to the MYNN-EDMF scheme (“MYNN” hereafter), which has been developed to handle both local (eddy diffusivity) and nonlocal (mass flux) parts and includes an option to turn on or off the mass-flux part (Olson et al. 2019). The MYNN-EDMF uses a hybrid diagnostic PBLH definition, in which the PBLH is defined by both virtual potential temperature profile-based (Ziθ) and TKE-based (ZiTKE) PBLH, and Ziθ dominates for neutral and unstable conditions, while ZiTKE dominates for stable condition (Olson et al. 2019). The MYNN-EDMF scheme is switched to the mass-flux scheme and paired with the MYNN surface layer scheme (Olson et al. 2021) in our study.
c. Comparison and analysis methods
1) Data processing
The model data are extracted from domain 3 (1.1 km) of the WRF Model outputs using the geographical coordinates of four PECAN sites with redundancy of ±0.02° to mitigate the drifting of radiosondes and the heterogeneity of the surface. The modeled outputs (as well as the measurements) are categorized by the cloud fractions and cloud location. The fair weather condition represents the time when cloud coverage is less than 6/8 of the sky using the cloud fractions measurement by total sky imager (TSI). The rationale for this is that when clouds become intense on a larger scale, the PBLH is not always well defined (LeMone et al. 2019; Sathyanadh et al. 2017; Zhong et al. 2020) by both model and observations. By filtering out the time with large cloud fractions, we can physically stratify the data and narrow down the time that the PBLH definition has ambiguities. During the period with a large cloud fraction, the first cloud base height measured by the ceilometer is compared with radiosonde-derived PBLH to determine whether the cloud is within the PBL. In total, 128 h out of the entire simulation period have cloud coverage higher than 6/8 of the sky, during which there are 32 radiosonde launches with 22 cases of clouds above the PBL and 10 cases of clouds within the PBL.
Given the clear differences in PBLH determination across PBL schemes and retrieval algorithms discussed in the previous section, comparisons become difficult if only the default PBLH definitions from the PBL schemes are used, as pointed out by Hu et al. (2010), LeMone et al. (2013), and Shin and Hong (2011). We implement an alternative PBLH estimate using the virtual potential temperature (θυ) and wind profiles from WRF ensembles (as well as each simulation) based on the same method of PBLH estimation for radiosonde profiles (Liu and Liang 2010). This method uses predetermined thresholds of θυ increment (δs = 1 K) near the surface to categorize PBL into the SBL, neutral boundary layer (NBL), and CBL separately. For NBL and CBL scenarios, a first guess height (k) is determined at which θk exceeds θυ at the lowest level (θ1) by a threshold of δu = 0.5 K (θk − θ1 > δu). Then, the search continues upward from the first guess height to find the CBL height at which the temperature gradient exceeds the threshold (
(a) An example of CBL height determination from radiosonde θυ. The lidar-derived PBLH is also included for comparison. (b) As in (a), but for SBL determination. (c) Profile of
Citation: Journal of Applied Meteorology and Climatology 63, 11; 10.1175/JAMC-D-24-0046.1
(a) Modeled ensemble-mean θυ profile overlayed with radiosonde-derived PBLH and cloud base height from the ceilometer for the entire simulation period at FP1. (b) Modeled PBLHpro, PBLHvar, and lidar- and radiosonde-derived PBLH. The model is from the runs with 77 vertical levels. The blank areas represent the time with a large fraction of cloud coverage discussed in section 2.
Citation: Journal of Applied Meteorology and Climatology 63, 11; 10.1175/JAMC-D-24-0046.1
We then calculate the diurnal cycle of PBLHpro and PBLHvar from each simulation output by time averaging the model simulations for each local time across the entire simulation period and confronted with lidar and radiosonde measurement to illustrate the general performance of each simulation under fair weather conditions (Fig. 4). Note that this step is to elucidate the systematic variation of PBLHs through statistics rather than focusing on local ranges in PBLH from specific events during the campaign. We also compare the corresponding radiosonde-derived PBLH to WRF PBLHvar and PBLHpro from the ensemble average at each launching time and report the median and interquartile of the bias along with the centered root-mean-square error (RMSE; appendix B; Fig. 5). Then, we report the centered RMSE and bias of lidar-derived PBLH relative to radiosonde-derived PBLH for those times where radiosonde and lidar both have measurements. The cases with a large fraction of cloud are analyzed separately in Fig. 6.
Median diurnal PBLH at FP1 site from WRF results. (a) PBLHvar at 57 levels and (b) PBLHvar at 77 levels for fair weather conditions. Derived PBLHpro for (c) 57 levels and (d) 77 levels at CST. The average diurnal profiles for each configuration of the WRF Model are labeled with different colors and line types. The red error bars denote the median and inner quartile (25th–75th) of PBLH retrieved from radiosondes. The black solid line and gray shades represent the median and inner quartile of lidar-derived PBLH, respectively. The legends for model configurations are described in the order of PBL scheme, ICBCs, and microphysics schemes, separated by “_”. MP8 and MP10 refer to the Thompson scheme that was used in the default CONUS setup and the Morrison double moment scheme, respectively.
Citation: Journal of Applied Meteorology and Climatology 63, 11; 10.1175/JAMC-D-24-0046.1
(a),(b) The bias and (c),(d) centered RMSE of WRF ensemble PBLHpro and PBLHvar relative to radiosonde-derived PBLH for fair weather conditions. (a),(c) The bias and centered RMSE at FP1 at 0000, 0600, 1200, 1500, and 1800 CST and (b),(d) bias and centered RMSE at FP2, FP3, and FP6 at 2100 CST. The MYJ series have been removed from PBLHvar-77 in this figure. The green “x” with error bar represents lidar PBLH relative to radiosondes at FP1 (lidar sonde). The marker represents the median of bias between the model and radiosonde at a given local time (model sonde), while the error bar denotes the interquartile. The n denotes the number of radiosonde launches at a given local time and site.
Citation: Journal of Applied Meteorology and Climatology 63, 11; 10.1175/JAMC-D-24-0046.1
As in Fig. 5, but for cases with large cloud fraction that is (a) above PBL and (b) within PBL. Here, we show the bias as scatters rather than error bars due to the small number of data points (n ≤ 7) in each category.
Citation: Journal of Applied Meteorology and Climatology 63, 11; 10.1175/JAMC-D-24-0046.1
To elucidate further the physics behind the bias of the PBLHvar and PBLHpro, we also compare the modeled and observed surface variables including 2-m temperature (T2), water vapor mixing ratio (q2), 10 m wind speed (U10) as well as turbulence related variables like surface friction velocity (
Comparison of key surface variables and fluxes from each simulation (runs with 57 vertical levels) with observations at FP1 reported as daily average profiles for fair weather conditions. The Tsoil and qsoil represent the 10-cm soil temperature and volumetric water content, respectively. The label of each simulation is the same as that used in Fig. 4. The black solid line and gray shades represent the median and inner quartile of the surface measurements, respectively.
Citation: Journal of Applied Meteorology and Climatology 63, 11; 10.1175/JAMC-D-24-0046.1
2) Analysis of the ensemble spread
We estimate the net influence of each class of model configuration or factor (i.e., ICBC, PBL scheme, vertical grid, and microphysics) to ensuing range (here measured as spread) of campaign-period averaged PBLHvar, PBLHpro, T2, q2, U10, LH, HFX, Net Rad, and
Spread of surface state variables, fluxes, PBLHvar, and PBLHpro from each model factor reported at each local time for fair weather conditions. The σ denotes the spread which is averaged over the entire campaign for each local time at FP1. Details regarding the method are discussed in section 2c(2).
Citation: Journal of Applied Meteorology and Climatology 63, 11; 10.1175/JAMC-D-24-0046.1
3. Measurements during PECAN campaign
All the measurements in this study were collected during the PECAN field project in 2015. The PECAN campaign gathered a wide range of observations from the atmospheric sounding systems at the surface, mobile radars, and mesonets, as well as aircraft measurements during June–July 2015 with the purpose of better understanding the nocturnal mesoscale convective system and the relationship with LLJ, SBL, and atmospheric bore (Geerts et al. 2017). In total, there were 31 intensive operation periods (IOPs) and 12 unofficial filed operations (UFOs) during the entire PECAN campaign. The rationale for choosing the location of SGP during PECAN is the relatively flat terrain and the extensive observations of the atmospheric profiles in terms of both spatial and temporal resolutions.
The in situ measurement-derived PBLH from radiosonde thermodynamic profiles is the most common and established approach and has been extensively discussed in the literature (e.g., Clifford et al. 1994; Hennemuth and Lammert 2006; Holzworth 1964; Liu and Liang 2010; Seibert et al. 2000; Sivaraman et al. 2013). In addition, surface remote sensing techniques like lidar or ceilometer (Dang et al. 2019) and Doppler lidar (de Arruda Moreira et al. 2018; Pearson et al. 2010; Pichugina et al. 2008; Schween et al. 2014; Träumner et al. 2011; Tucker et al. 2009) have been used to retrieve PBLH at higher temporal resolution. A survey of these PBL observing system capabilities is described in Teixeira et al. (2021). Details regarding the capabilities and limitations of various types of ground-based remote sensing PBLH measurements can be found in Kotthaus et al. (2023). It should be recognized that while the retrieval algorithm differs for these measurement types and that there are more lidar-derived PBLH data across the day, a robust network of PBLH should include these measurement types given the paucity of PBLH measurements under spatially and temporally heterogeneous environments even for the southern Great Plains. Hence, we can leverage the collocation of these measurements to determine the range of PBLH across measurement types. Moreover, profiles of
Given the locations of the sites and the abundance of the observation, four fixed PECAN Integrated Sounding Array (PISA) sites (FP1, FP2, FP3, and FP6, as shown in Fig. 1) are chosen as the reference measurements when comparing with the WRF outputs. The measurements include TSI (Flynn and Morris 2023), cloud base height from the ceilometer (Zhang et al. 1997) at FP1, surface meteorological instrumentation (SMET; Clark and McAuliffe 2015; Delgado and Vermeesch 2016; Kyrouac et al. 2015; Turner 2016), radiosondes (Clark 2016; Holdridge et al. 2015; UCAR/NCAR 2015; Vermeesch 2015), and the PBLH derived from its potential temperature profile using the method of Liu and Liang (2010) as well as Doppler lidar–retrieved PBLH (Sivaraman and Zhang 2010). The Doppler lidar PBLH dataset is based on a
Site locations and measurements.
4. Results and discussion
a. Modeled PBLHvar and PBLHpro
The modeled ensemble-mean θυ profile superimposed with cloud base height and radiosonde-derived PBLH along with lidar-derived and modeled PBLH at FP1 across the entire simulation period are shown in Fig. 3. The average diurnal PBLHpro and PBLHvar profile from each WRF simulation and the radiosonde- and lidar-derived (at FP1) PBLH for fair weather conditions are shown in Fig. 4. Details of the centered RMSE and bias between model and radiosonde as well as lidar and radiosonde are also summarized in Fig. 5 (fair weather) and Fig. 6 (large fraction of cloud coverages).
On average, the radiosonde-derived PBLH is about 200–300 m (15%–20%) higher than the Doppler lidar during the daytime, and the spread can go up to 700 m at the sunset period (Fig. 5), with the lidar tending to underestimate PBLH. The discrepancies are similar to Hicks et al. (2019), in which they report a root-mean-square of 421 m for PBLH between lidar and radiosonde at temperate sites. The uncertainties between radiosonde- and lidar-derived PBLH give us information on how much discrepancy one can expect from measurements with different retrieval mechanisms. For the model, the YSU scheme yields the closest PBLH result compared with radiosonde, followed by the MYNN scheme during the daytime, while the MYJ scheme tends to underestimate especially for PBLHvar from 77 levels (up to 700 m). However, the PBLHvar from YSU and MYJ tend to overestimate almost twice as high as the MYNN series (Figs. 4a,b) at night. It is worth noting that the use of a consistent methodology (i.e., use of thermodynamic profiles) in determining PBLH (PBLHpro) can improve the fit of the model to observations when compared to PBLHvar at the same vertical level (Fig. 4a vs Fig. 4c and Fig. 4b vs Fig. 4d) by ∼100 m (∼6%) during most of the daytime. Another feature is that the PBLHpro shows some improvement at 0000 central standard time (CST) (FP1) and 2100 CST (FP2 and FP6) in terms of the bias against radiosonde (Figs. 5a,b). In terms of centered RMSE (Figs. 5c,d), PBLHpro show ∼80 m less relative to PBLHvar at 0000 and 0600 CST (FP1) and 2100 CST (FP6) during the night. Additionally, the nighttime spread of PBLHpro is much less (∼100) compared with PBLHvar (Fig. 4a vs Fig. 4c and Fig. 4b vs Fig. 4d). The decrease in bias and spread clearly shows the advantage of using a consistent method to retrieve PBLH. Moreover, the improvement of PBLHpro reaches up to ∼200 m (10%) during the sunset transition period when the daytime maximum model bias occurs. Regarding the maximum bias at sunset, we point to LeMone et al. (2019) where they conclude that the temperature-gradient-defined PBLH stays roughly the same or slowly decreases, while turbulence-defined PBLH decreases faster during the evening transition period. Thus, the different definitions (θυ, TKE, or Rib) from different PBL schemes, as well as different measurement types (i.e., lidar vs radiosonde), will exhibit the separation of turbulent- and temperature-defined PBLH and large differences in derived PBLH at this transition period. The separation between turbulence and θυ defined PBLH can also be seen in an NBL case in Fig. 12c.
The increased vertical grids have some benefit in improving (100–150 m) daytime PBLHvar and PBLHpro at 1500 and 1800 CST (Fig. 5). On the other hand, the most obvious change from increased vertical grid spacing (Fig. 4b) comes from the set of model runs using the MYJ PBL scheme which yields a much lower PBLH of ∼400 m (∼25%) on average during daytime. Hu et al. (2010) and LeMone et al. (2013) also found that the original PBLH (PBLHvar in our study) read from the MYJ series output files yields lower PBLH than θυ-derived PBLH (PBLHpro in our study). Given that the PBLHpro for MYJ (Fig. 4d) series does not have this magnitude of underestimation, the definition of PBLHvar from the MYJ scheme claims most of the cause. Another possible reason for this is that the 77 vertical levels have steered away from the optimum grid spacing for the MYJ scheme reported in Shin and Dudhia (2016). Given this underestimation, we removed the MYJ series with 77 levels (PBLHvar) from the ensemble in Figs. 5 and 6. For cloud-intense cases (Fig. 6), high variances are seen during daytime despite the bias being centered around zero. The PBLHpro also shows larger variations compared to PBLHvar at 1500 and 1800 CST. Here, we show the bias from each case due to a small number of data points in each category (n ≤ 7). The θυ profiles of some of the cloud-intense cases are shown in Fig. 12 and are further discussed in section 4c.
b. Modeled key surface variables
In Fig. 7, we show the diurnal cycle of the surface state variables (T2, q2, and U10), heat fluxes (LH and HFX), Net Rad, and friction velocity
The relatively dryer conditions modeled by the YSU series in this study have also been reported by Hu et al. (2010). A similar magnitude of the bias between modeled and observations of surface state variables is reported in these studies as well: Avolio et al. (2017), Banks et al. (2016), Hu et al. (2010), and Xie et al. (2012). Steeneveld et al. (2011) evaluated the WRF Model skill for LH, HFX, and
In Fig. 8, we present our analysis of the spread for each class of model configuration (or factor) and illustrate the influence of each class of model configuration on the spread shown in Fig. 7. Please see section 2c for a description of our analysis. We find that the errors of the model are mainly related to the stage of the diurnal cycle rather than the drift of errors due to forecast time. In terms of the spread in key surface variables and fluxes, we find that the ICBCs have the strongest influence on the spread of HFX and LH (Figs. 7a,b). The order is then followed by PBL schemes microphysics and vertical grid spacing. The ICBC, PBL scheme, and microphysics have similar impact on T2, q2, U10, and
c. Modeled atmospheric profiles
Profiles of θυ, q, and U from each WRF simulation are reported against radiosonde observations in Fig. 9 (57 level) and Fig. 10 (77 level). The bias for the ensemble (12 simulations) θυ, q, and U profiles from each model vertical grid spacing (55 and 77) is also shown in Fig. 13. Cases with a deep CBL (≥3 km) and those with a large fraction of clouds are included in Figs. 11 and 12, respectively. Note that the normalization process in Figs. 9 and 10 smooths out the variations of the atmospheric profiles, especially for the height of θυ inversion. The analyses of PBLH based on individual atmospheric profiles are already discussed in section 4a. The purpose of the composite is to evaluate the collective behavior and the overall performance of each model simulation in reproducing the thermodynamic structure under fair weather conditions. In general, the model is overpredicting θυ by ∼0.5 K during nighttime and 1–2 K during daytime within the PBL. The YSU series tends to estimate a much dryer q profile than MYJ and MYNN series by 1–1.5 g kg−1 and 0.5–1.0 K warmer θυ profile throughout the day, consistent with our comparison with T2 and q2 in Fig. 7. Most model simulations tend to overpredict U by 2–3 m s−1, and the bias and the spread of bias (Fig. 13) are larger during nighttime than daytime. Modeled θυ and q also tend to have a larger bias among different simulations below the top of PBL than above during the daytime. The inversion of θυ tends to occur lower when compared with observations and can be a result of the underprediction of HFX during the morning growth period mentioned in the previous section and the mixing and transport of that heating by the parameterization of the PBL scheme. As for the nighttime, the model in general captures well θυ near the surface and the curvature change between the transition of SBL to residual layer (RL) at 0000 CST. However, at 0600 CST, the slightly higher spread of bias for PBLHpro in Fig. 5 could be a sign of too much vertical mixing in stable conditions in NWP models as indicated by Holtslag et al. (2013), and the overprediction of
Profiles of median θυ, q, and U from each WRF simulation (57 levels) confront with radiosonde measurements at 0000, 0600, 1500, and 1800 CST for fair weather conditions. The label for each simulation is the same as Figs. 4 and 7. The radiosonde profiles are shown in median (solid black) and interquartile (gray shade). Note that we use the 0–1-km range for nighttime and the 0–2.5-km range for daytime, given the typical PBLH at each time. The number of radiosonde launches is shown in Fig. 5.
Citation: Journal of Applied Meteorology and Climatology 63, 11; 10.1175/JAMC-D-24-0046.1
As in Fig. 9, but for a model with 77 levels.
Citation: Journal of Applied Meteorology and Climatology 63, 11; 10.1175/JAMC-D-24-0046.1
(a)–(f) The θυ profiles from each model simulation (77 levels) and radiosonde measurements for cases of deep CBL [≥3 km, except for (e)]. Cloud base height, radiosonde, lidar-derived PBLH, and PBLHpro from model simulations are also shown in the figure. The label for each simulation is the same as Figs. 4 and 7.
Citation: Journal of Applied Meteorology and Climatology 63, 11; 10.1175/JAMC-D-24-0046.1
As in Fig. 11, but for cases of a large fraction of cloud coverage. (a)–(c) Cloud base higher than PBLH and (d)–(f) cloud within PBL.
Citation: Journal of Applied Meteorology and Climatology 63, 11; 10.1175/JAMC-D-24-0046.1
Profiles of the bias for θυ, q, and U between model ensemble (model minus sonde) from two vertical levels and radiosonde measurements at 0000, 0600, 1500, and 1800 CST at FP1 for fair weather conditions. The results are reported as a median bias for all matched (in time and vertical levels) radiosonde launches during the entire campaign at each local time. The solid lines represent the median of bias, and the shaded area represents the interquartile.
Citation: Journal of Applied Meteorology and Climatology 63, 11; 10.1175/JAMC-D-24-0046.1
When the number of model vertical levels increased, the most noticeable accuracy improvement in the θυ profile is observed at 0600 and 1800 CST by ∼0.4 and 0.2 K, respectively, below 0.5 km (Fig. 13). However, during other times, the improvement in terms of their bias to radiosonde profiles is limited. Several studies focusing on the LLJ modeling at SGP regions find that increasing the vertical grids of the model can improve the details in the simulation of LLJ, but the benefits are limited (Parish and Clark 2017; Smith et al. 2018, 2019). We also want to point out that the determination of PBLHpro is more related to the shape (i.e., the height of θυ gradient or curvature change) of the profiles rather than the bias relative to the measurements (see section 2c on PBLH determination). The improvement of PBLHpro with higher vertical grids mentioned in the previous section can be seen in Fig. 11 (77 levels at 1500 and 1800 CST) where the modeled θυ inversion is more clearly defined with the height closer to the observed inversion when compared with Fig. 10 (57 levels at 1500 and 1800 CST), even if most of the model simulations are overestimating θυ.
In addition, the MYJ_ERA5_MP10_77 simulation predicts the closest θυ profile (and second best for 57 levels) compared with radiosonde at 1500 CST (Fig. 10) in terms of absolute error although, as previously mentioned, the PBLHvar from MYJ series at 77-level has the largest underestimation in daytime PBLH (Fig. 4b). Although some MYJ simulations (e.g., MYJ_GFS) do exhibit lower predicted inversion height than other PBL schemes, the magnitude of PBLHpro underestimation (Fig. 4d) is smaller than the PBLHvar (Fig. 4b). Many studies (Hu et al. 2010; Holtslag et al. 1995; Hong and Pan 1996; Srinivas et al. 2007) indicate that the local scheme (MYJ in our study) produces too weak vertical mixing. The literatures are consistent with our results where the TKE-defined PBLH is much lower than θυ-based PBLH. The MYJ case is a perfect example that highlights the benefit of retrieving and comparing modeled PBLH with observation in a physically consistent way. It is clear that the correct shape of θυ (and therefore q) profile is the key to estimating PBLHpro accurately, and by using the same PBLH retrieval method for observations, we can have a fair evaluation of the model without worrying about the varying definition of PBLH across PBL schemes. The MYJ case suggests that the transport and mixing of surface heat by the PBL scheme are as important as the sensible heat flux (HFX) itself in shaping the overall thermodynamic structure of the CBL. The importance of mixing is evident from the fact that the HFX from the MYJ series is not lower than that from other schemes (see Fig. 7b).
Recall from Fig. 7 that all simulations tend to overpredict q2, while the model tends to underpredict q profile (as shown in Figs. 9 and 10) above the surface. Note that the errors in q2 can also be due to erroneous qsoil although the model tends to predict drier soil moisture (Fig. 7h). The lack of moisture above the surface further suggests a lack of mixing and transport of properties between the surface and the CBL. Several studies (Avolio et al. 2017; Banks et al. 2016; Hu et al. 2010; Steeneveld et al. 2011; Xie et al. 2012) reported in their evaluation of WRF simulations over different locations of the world that WRF shows a clear inconsistency between the fit of potential temperature and HFX to observations. The WRF needs to adjust surface heat fluxes as well as the transport and mixing process of surface heating and moisture to achieve a better fit on the thermodynamic profiles, suggesting nonlinear feedback on thermodynamics, turbulence, and surface energy budget.
Apart from the full campaign-period average, the lack of mixing can also be seen in individual cases (Fig. 11). The modeled CBL is only 2.0–2.5 km for cases in Figs. 11b and 11e when the observed depth of well-mixed CBL is over 3 km. The spread of θυ in Fig. 11 (excluding panel f) has a similar magnitude when compared with the composite from Figs. 9 and 10. In addition, the ICBC also plays a crucial role in cases in Figs. 11a and 11f when the difference of CBL depth between GFS (solid line) and ERA5 (dotted line) is more than 1 km. The cases of clouds within the PBL characterized by two temperature inversions are shown in Figs. 11c and 11f. In these cases, the Doppler lidar shows a PBLH close to the first inversion and the cloud base, while radiosonde retrieves the PBLH near the second inversion. The difference is due to the radiosonde method where the first guess height is constrained by 1) the first-level temperature plus an increment (section 2c) and 2) a strong superadiabatic profile near the surface. In the case in Fig. 11f, the model runs with GFS show higher CBL depth and overall better fit of θυ profile than ERA5 runs when compared with the radiosonde.
The θυ profiles, in general, exhibit higher variability across different model simulations for cases with a large fraction of cloud coverage (Fig. 12) when compared with fair weather conditions (Figs. 9 and 10, Fig. 11, excluding panel f). The variability in terms of θυ near the surface is as large as 7 K for the cases with clouds within or near the top of PBL (Figs. 12a,e,f), with the impact of different microphysics (as indicated by lines with and without asterisks) being prominent. The large spread of θυ profile poses challenges in defining the PBLH for cloud-topped PBLs, as each model simulation illustrates a very different inversion height. The high spread of the θυ profiles also explains the higher variation in PBLHpro discussed in section 4a (Fig. 6). Simulating cloudy PBLs is known to be problematic (Teixeira et al. 2021), and the need for a better representation of cloud-topped PBL structure by PBL parameterization has been highlighted, as pointed out by Teixeira et al. (2008).
d. Impacts on PBLHvar/PBLHpro and atmospheric profiles
Since we have already discussed the various PBLH definitions in different PBL schemes earlier, it would be intuitive to only use PBLHpro to conduct a sensitivity analysis of model performance to PBLH. However, the default PBLH output (i.e., PBLHvar) is currently used directly in WRF coupled with chemistry (WRF-Chem) (Fast et al. 2006; Grell et al. 2005), and studies have reported that the modeled PBLH have impacts on surface pollutant concentrations (Du et al. 2020; Luo et al. 2022; Lv et al. 2020). Additionally, the PBLHvar from MYJ defines PBLH from TKE (partially for MYNN due to its hybrid definition), and this is the closest definition to turbulence-defined PBLH. Thus, we will briefly discuss the impacts on PBLHvar in this section. Our results show that the PBL schemes have the largest impact on the range of PBLHvar (Fig. 6i) with a spread of ∼150 m (∼75%) during nighttime and up to 400 m (25%) during daytime. The ICBC has the second largest impact, with a spread of 75–100 m (38%–57%) during nighttime and increases to 200 m (13%) during daytime. The microphysics schemes have a smaller impact with 20–50 m (10%–25%) during nighttime and up to 120–200 m (8%–13%) during daytime, while the impact from vertical grid spacing is about ∼100 m throughout the day. In addition, regardless of the factor, the time of peak spread occurs during the sunset period (i.e., 1700–1900 LT), which is the same time that the modeled PBLHvar/PBLHpro has the largest centered RMSE and bias with observations.
The impact of each factor on atmospheric profiles (Fig. 14) is essentially the impact on PBLHpro (Fig. 8h). The statistics show that the ICBCs and PBL scheme have comparable impacts of ∼0.5 K on θυ and 0.5–0.8 g kg−1 on q. However, the choice of ICBCs has the largest impact (1–2 m s−1) on U profile, followed by the PBL scheme (1–1.5 m s−1). The impact of vertical grid spacing is about half the magnitude of ICBCs and the PBL scheme at the lower part of the CBL and becomes important near the top of the CBL (1–1.5 km). The higher impact near CBL top points to the better predicted θυ inversion during daytime with increased vertical grids. The impact from microphysics is about half of the magnitude when compared with the impact of ICBCs and PBL scheme. Apart from the radiative effect of clouds, the latent heat release and evaporative cooling can also affect the atmospheric thermodynamic structure. Note that we have removed cases with a large fraction of cloud coverage and undefined θυ inversion; thus, the impact from microphysics is expected to be small for fair weather conditions. However, the choice of microphysics becomes important when a large fraction of clouds exist within or near the top of PBL (Fig. 12). The sensitivity of PBL variables to the PBL scheme in Greece (Banks et al. 2016) shows an even larger HFX and PBLH spread at some sites compared to our study. Avolio et al. (2017) tested the performance of different PBL schemes in Italy and reported that the MYNN scheme has a mean high bias of ∼700 m in PBLH and ∼100 W m−2 high bias in HFX during daytime when compared with other PBL schemes (including MYJ and YSU). Our study shows a smaller bias (<200 m and <40 W m−2) for MYNN. The comparison between the literature and our results highlights the impact of the PBL scheme on 1) PBLH calculations and 2) key variables that control the development of PBL. Like the results in Figs. 7 and 8, the modeled atmospheric profiles tend to group themselves by the choice of PBL scheme regardless of the choice of ICBCs for at least near the surface.
As in Fig. 8, but for the spread (σ) of θυ, q, and U profiles at FP1 due to each factor at 0000, 0600, 1500, and 1800 CST for fair weather conditions. The σ here has the same definition as Fig. 8 and in section 2c.
Citation: Journal of Applied Meteorology and Climatology 63, 11; 10.1175/JAMC-D-24-0046.1
Our results on the importance of PBL schemes and ICBCs on surface heat fluxes (Fig. 8) and atmospheric profiles (Fig. 14) suggest that the surface heating and net effect of turbulent mixing are largely influenced by the PBL scheme (parameterizations) along with the initial and boundary conditions. These are the key components to better simulate the vertical structure of thermodynamic/dynamic variables within the PBL. Given the fact that the HFX is predicted well in general by the model (Fig. 7) and the underprediction of the θυ inversion height (Figs. 9–11), the PBL schemes are not mixing enough of that surface HFX throughout the PBL and cannot predict enough entrainment at PBL top during the daytime. Note, however, that the overpredicted mixing during nighttime has influences on the curvature change of θυ at the transition height of SBL to RL. Given the magnitude of the spread due to the PBL scheme, we suggest that using multiple model physics schemes is as important (if not more important) as initial and boundary conditions in accounting for the model errors within the ensemble framework. The multiphysics is especially important in obtaining more physically and dynamically consistent error covariances for data assimilation (Adam et al. 2016; Chipilski et al. 2020; Coniglio et al. 2019; Degelia et al. 2019; Dang et al. 2022; Johnson and Wang 2017; Johnson et al. 2017; Tangborn et al. 2021), as the resulting sensitivities mainly drive the analysis increments (Anderson et al. 2009).
5. Summary and conclusions
We conduct 24 WRF simulations for the entire PECAN field campaign period using different combinations of PBL schemes, microphysics schemes, initial and boundary condition datasets, and vertical grid spacing. We evaluate each simulation of PBLH from WRF outputs (PBLHvar) as well as using modeled atmospheric profiles (PBLHpro) with measurements during the campaign. The ensembles of PBLHvar and PBLHpro (from the ensemble of profiles) are also evaluated against radiosonde-derived PBLH. Based on the amount and location of the cloud, we stratified the data into fair weather conditions, a large fraction of cloud coverage above the PBL and within the PBL. On average, the radiosonde-derived PBLH is about 200–300 m (10%–20%) higher than Doppler lidar–derived PBLH during the daytime, and the spread can be as large as 700 m during the sunset transition period for fair weather conditions. Overall, the modeled PBLHvar underpredicts by 300–500 m (20%–30%) during daytime and overpredicts by ∼200 m (∼55%) during nighttime when compared with radiosonde retrievals. Note that the largest spread between modeled and observed PBLH (600–700 m) occurs during the sunset transition period, which is the same time as the maximum observed bias (∼700 m), and when modeled PBL schemes exert the strongest impact on the results of PBLHvar. When the same profile method is used (i.e., PBLHpro), the median bias of daytime and nighttime modeled PBLHpro decreased by ∼100 m (∼6%) and ∼200 m (40%), respectively, while the spread during the sunset transition period is also decreased by up to 200 m (∼10%). The PBLHpro also has less spread across different simulations at nighttime. For cases with a large fraction of cloud, especially for cloud-topped PBLs, the spread between model simulations becomes larger, and the model has difficulties reproducing the correct thermodynamic profile.
Our statistical analyses indicate that the largest spread (as measured by absolute differences across model configurations) of modeled PBLH, key surface variables, and atmospheric profiles comes from the choice of PBL scheme, which can result in different interpretations of PBLH and subsequent parameterizations of turbulence mixing. The spread is then followed by ICBC, which mainly alters the initial condition of the state variables and prescribes the lateral and surface boundary conditions, thereby affecting modeled surface fluxes and atmospheric profiles. The bias of modeled LH, HFX, and
Our findings also point directly to the need to consider these observational and model variances, as well as the contrast of these variances during daytime, nighttime, and sunset period when evaluating model performance and improving model prediction via data assimilation (Dang et al. 2022; Tangborn et al. 2021), notwithstanding other important considerations such as the treatment of land–atmosphere interaction in these models. Better measurements and understanding of modeled PBL processes are essential for improving severe weather prediction (winter storms: Cohen et al. 2015; Johns and Doswell 1992, and convective initiation: McCaul and Cohen 2002) at time scales ranging from hours to days. Considering the extent of variability resulting from changing the PBL scheme, as illustrated in Figs. 8 and 14, conducting WRF simulations with multiple physics configurations and perturbing the initial and boundary conditions for ensemble forecasting, particularly emphasizing the PBL, are advised to more accurately capture the range of variability in the state variables. The variability in state variables is crucial for better capturing the error covariances in ensemble-based data assimilation systems (Anderson et al. 2009). More importantly, the systematic bias associated with given PBL schemes cannot be addressed by just using different ICBCs. Thus, model simulations with lower performance can degrade the overall performance of the ensemble, especially for cloud-topped PBLs, as demonstrated in sections 4a and 4c. Additionally, physics schemes that excel in one aspect of the surface or atmospheric properties may not perform well in others, as discussed in sections 4b and 4c, as well as pointed out by Shin and Dudhia (2016). Accounting for these variations using multiphysics can encapsulate a more realistic range in the error covariances with implications as well on the design of observation operators in assimilating PBLH data from various measurement (retrieval) methods. Last, this is important as well for air quality prediction, where its performance depends on the accurate representation of the PBL process, given reports of the impact of modeled PBLH on surface pollutant concentrations (Du et al. 2020; Luo et al. 2022; Lv et al. 2020). A better understanding of the modeled PBLH variations provides insight into the transport, dispersion, and prediction of surface pollutants. These considerations have important implications for future activities directed toward designing future observing systems for the PBL and corresponding model–data fusion systems assimilating these derived measurements and retrievals.
Acknowledgments.
This work is supported by NASA Decadal Survey Incubation Grant 80NSSC22K1104 (NNH21ZDA001N-DSI) (Co-Is: A. F. Arellano/UA, B. Demoz/UMBC, J. Santanello/NASA GSFC, and W. Blumberg/NASA GSFC, and collaborators: J. Anderson/NCAR and A. Tangborn/NOAA). We especially thank Dr. Tangborn for his insightful thoughts regarding his proposed approach to PBLH data assimilation and Dr. Santanello for his advice on data processing.
Data availability statement.
The ERA5 data used for the ICBCs of the WRF Model in the study are available at the Research Data Archive at the National Center for Atmospheric Research via https://doi.org/10.5065/BH6N-5N20 with open access. The GFS data used for the ICBCs of the WRF Model in the study are available at the Research Data Archive at the National Center for Atmospheric Research via https://doi.org/10.5065/D65D8PWK with open access. The lidar-retrieved PBLH data in the study are available at the Atmospheric Radiation Measurement (ARM) user facility via https://doi.org/10.5439/1095386 with open access. The ceilometer retrieved cloud base height data in the study are available at the ARM user facility via https://doi.org/10.5439/1181954 with open access. The radiosonde data used for PBLH retrieval in the study are available at the ARM user facility via https://doi.org/10.5439/1786358, https://doi.org/10.5065/D6SJ1HSG, https://doi.org/10.5065/D6CV4FTM, and https://doi.org/10.5065/D6RR1WN0 with open access. The surface friction velocity measurements in the study are available at the ARM user facility via https://doi.org/10.5439/1025039 with open access. The Surface Meteorological Instrumentation (SMET) data in the study are available at the ARM user facility via https://doi.org/10.26023/35HZ-2W5G-Z20H, https://doi.org/10.5065/D6FQ9TPH, https://doi.org/10.5065/D6GM85DZ, and https://doi.org/10.5065/D6765CD0 with open access. The surface net radiation measurements in the study are available at the ARM user facility via https://doi.org/10.5439/1027268 with open access. The surface sensible and latent heat flux measurements in the study are available at the ARM user facility via https://doi.org/10.5439/1989601 with open access. The soil temperature and moisture measurements in the study are available at the ARM user facility via https://doi.org/10.5439/1350681 with open access. The Total Sky Imager (TSI) data in the study are available at the ARM user facility via https://doi.org/10.5439/1992207 with open access. Version 4.4 of the WRF Model used for weather forecast is preserved at https://github.com/wrf-model/WRF, available via open access with registration, and developed openly at https://www2.mmm.ucar.edu/wrf/users/.
APPENDIX A
PBLH Determination from Radiosonde
The PBLH determination from radiosonde profiles is from the study of Liu and Liang (2010). Our study interpolates the radiosonde profiles to a 50-m unified grid.
APPENDIX B
Centered RMSE Definition
The centered RMSE definition is as follows:
where
REFERENCES
Adam, S., A. Behrendt, T. Schwitalla, E. Hammann, and V. Wulfmeyer, 2016: First assimilation of temperature lidar data into an NWP model: Impact on the simulation of the temperature field, inversion strength and PBL depth. Quart. J. Roy. Meteor. Soc., 142, 2882–2896, https://doi.org/10.1002/qj.2875.
Allabakash, S., and S. Lim, 2020: Climatology of planetary boundary layer height-controlling meteorological parameters over the Korean peninsula. Remote Sens., 12, 2571, https://doi.org/10.3390/rs12162571.
Anderson, J., T. Hoar, K. Raeder, H. Liu, N. Collins, R. Torn, and A. Avellano, 2009: The data assimilation research testbed: A community facility. Bull. Amer. Meteor. Soc., 90, 1283–1296, https://doi.org/10.1175/2009BAMS2618.1.
Avolio, E., S. Federico, M. M. Miglietta, T. L. Feudo, C. R. Calidonna, and A. M. Sempreviva, 2017: Sensitivity analysis of WRF model PBL schemes in simulating boundary-layer variables in southern Italy: An experimental campaign. Atmos. Res., 192, 58–71, https://doi.org/10.1016/j.atmosres.2017.04.003.
Banks, R. F., J. Tiana-Alsina, J. M. Baldasano, F. Rocadenbosch, A. Papayannis, S. Solomos, and C. G. Tzanis, 2016: Sensitivity of boundary-layer variables to PBL schemes in the WRF model based on surface meteorological observations, lidar, and radiosondes during the HygrA-CD campaign. Atmos. Res., 176–177, 185–201, https://doi.org/10.1016/j.atmosres.2016.02.024.
Bao, J.-W., S. A. Michelson, and E. D. Grell, 2019: Microphysical process comparison of three microphysics parameterization schemes in the WRF Model for an idealized squall-line case study. Mon. Wea. Rev., 147, 3093–3120, https://doi.org/10.1175/MWR-D-18-0249.1.
Borge, R., V. Alexandrov, J. J. del Vas, J. Lumbreras, and E. Rodríguez, 2008: A comprehensive sensitivity analysis of the WRF model for air quality applications over the Iberian Peninsula. Atmos. Environ., 42, 8560–8574, https://doi.org/10.1016/j.atmosenv.2008.08.032.
Chipilski, H. G., X. Wang, and D. B. Parsons, 2020: Impact of assimilating PECAN profilers on the prediction of bore-driven nocturnal convection: A multiscale forecast evaluation for the 6 July 2015 case study. Mon. Wea. Rev., 148, 1147–1175, https://doi.org/10.1175/MWR-D-19-0171.1.
Clark, R., 2016: FP3 Ellis, KS radiosonde data, version 2.0. UCAR/NCAR - Earth Observing Laboratory, accessed 27 March 2023, https://doi.org/10.5065/D6GM85DZ.
Clark, R., and M. McAuliffe, 2015: FP3 Millersville University surface meteorology and radiation data, version 1.0. UCAR/NCAR - Earth Observing Laboratory, accessed 24 January 2024, https://doi.org/10.5065/D6CV4FTM.
Clifford, S. F., J. C. Kaimal, R. J. Lataitis, and R. G. Strauch, 1994: Ground-based remote profiling in atmospheric studies: An overview. Proc. IEEE, 82, 313–355, https://doi.org/10.1109/5.272138.
Cohen, A. E., S. M. Cavallo, M. C. Coniglio, and H. E. Brooks, 2015: A review of planetary boundary layer parameterization schemes and their sensitivity in simulating southeastern U.S. cold season severe weather environments. Wea. Forecasting, 30, 591–612, https://doi.org/10.1175/WAF-D-14-00105.1.
Coniglio, M. C., G. S. Romine, D. D. Turner, and R. D. Torn, 2019: Impacts of targeted AERI and Doppler lidar wind retrievals on short-term forecasts of the initiation and early evolution of thunderstorms. Mon. Wea. Rev., 147, 1149–1170, https://doi.org/10.1175/MWR-D-18-0351.1.
Dang, R., Y. Yang, X. M. Hu, Z. Wang, and S. Zhang, 2019: A review of techniques for diagnosing the Atmospheric Boundary Layer Height (ABLH) using aerosol lidar data. Remote Sens., 11, 1590, https://doi.org/10.3390/rs11131590.
Dang, R., X. Qiu, Y. Yang, and S. Zhang, 2022: Observation system simulation experiments (OSSEs) for assimilation of the planetary boundary-layer height (PBLH) using the EnSRF technique. Quart. J. Roy. Meteor. Soc., 148, 1184–1207, https://doi.org/10.1002/qj.4254.
de Arruda Moreira, G., and Coauthors, 2018: Study of the planetary boundary layer by microwave radiometer, elastic lidar and Doppler lidar estimations in southern Iberian Peninsula. Atmos. Res., 213, 185–195, https://doi.org/10.1016/j.atmosres.2018.06.007.
Degelia, S. K., X. Wang, and D. J. Stensrud, 2019: An evaluation of the impact of assimilating AERI retrievals, kinematic profilers, rawinsondes, and surface observations on a forecast of a nocturnal convection initiation event during the PECAN field campaign. Mon. Wea. Rev., 147, 2739–2764, https://doi.org/10.1175/MWR-D-18-0423.1.
Delgado, R., and K. Vermeesch, 2016: FP2 UMBC surface weather station data, version 1.0. UCAR/NCAR - Earth Observing Laboratory, accessed 20 December 2023, https://doi.org/10.5065/D6SJ1HSG.
Du, Q., and Coauthors, 2020: Modeling diurnal variation of surface PM2.5 concentrations over East China with WRF-Chem: Impacts from boundary-layer mixing and anthropogenic emission. Atmos. Chem. Phys., 20, 2839–2863, https://doi.org/10.5194/acp-20-2839-2020.
Dudhia, J., 2010: WRF physics options. NCAR WRF Basic Tutorial, 113 pp., https://citeseerx.ist.psu.edu/document?repid=rep1&type=pdf&doi=9869b1fa62d69201c71bae3bcc4131b2ebaba969.
Dzebre, D. E. K., A. A. Acheampong, J. Ampofo, and M. S. Adaramola, 2019: A sensitivity study of Surface Wind simulations over Coastal Ghana to selected Time Control and Nudging options in the Weather Research and Forecasting Model. Heliyon, 5, e01385, https://doi.org/10.1016/j.heliyon.2019.e01385.
ECMWF, 2019: ERA5 reanalysis (0.25 degree latitude-longitude grid). National Center for Atmospheric Research. Computational and Information Systems Laboratory, accessed 15 June 2022, https://rda.ucar.edu/datasets/d633000/citation/#.
Edwards, J. M., A. C. Beljaars, A. A. Holtslag, and A. P. Lock, 2020: Representation of boundary-layer processes in numerical weather prediction and climate models. Bound.-Layer Meteor., 177, 511–539, https://doi.org/10.1007/s10546-020-00530-z.
Fast, J. D., W. I. Gustafson Jr., R. C. Easter, R. A. Zaveri, J. C. Barnard, E. G. Chapman, G. A. Grell, and S. E. Peckham, 2006: Evolution of ozone, particulates, and aerosol direct radiative forcing in the vicinity of Houston using a fully coupled meteorology-chemistry-aerosol model. J. Geophys. Res. Atmos., 111, D21305, https://doi.org/10.1029/2005JD006721.
Flynn, D., and V. Morris, 2023: Total Sky Imager (TSISKYCOVER) Atmospheric Radiation Measurement (ARM) user facility, accessed 27 May 2023, https://doi.org/10.5439/1992207.
García-Díez, M., J. Fernández, L. Fita, and C. Yagüe, 2013: Seasonal dependence of WRF model biases and sensitivity to PBL schemes over Europe. Quart. J. Roy. Meteor. Soc., 139, 501–514, https://doi.org/10.1002/qj.1976.
Gaustad, K., and S. Xie, 2015: Best-estimate fluxes from EBBR Measurements and Bulk Aerodynamics Calculations (30BAEBBR). Atmospheric Radiation Measurement (ARM) user facility, accessed 4 August 2023, https://doi.org/10.5439/1027268.
Geerts, B., and Coauthors, 2017: The 2015 Plains Elevated Convection At Night field project. Bull. Amer. Meteor. Soc., 98, 767–786, https://doi.org/10.1175/BAMS-D-15-00257.1.
Grell, G. A., S. E. Peckham, R. Schmitz, S. A. McKeen, G. Frost, W. C. Skamarock, and B. Eder, 2005: Fully coupled “online” chemistry within the WRF model. Atmos. Environ., 39, 6957–6975, https://doi.org/10.1016/j.atmosenv.2005.04.027.
Hall, W. D., R. M. Rasmussen, and G. Thompson, 2005: The new Thompson microphysical scheme in WRF. Preprints, 2005 WRF/MM5 User’s Workshop, Boulder, CO, NCAR, 6.1.
Hariprasad, K. B. R. R., C. V. Srinivas, A. B. Singh, S. V. B. Rao, R. Baskaran, and B. Venkatraman, 2014: Numerical simulation and intercomparison of boundary layer structure with different PBL schemes in WRF using experimental observations at a tropical site. Atmos. Res., 145–146, 27–44, https://doi.org/10.1016/j.atmosres.2014.03.023.
Hennemuth, B., and A. Lammert, 2006: Determination of the atmospheric boundary layer height from radiosonde and lidar backscatter. Bound.-Layer Meteor., 120, 181–200, https://doi.org/10.1007/s10546-005-9035-3.
Hersbach, H., and Coauthors, 2020: The ERA5 global reanalysis. Quart. J. Roy. Meteor. Soc., 146, 1999–2049, https://doi.org/10.1002/qj.3803.
Hicks, M., B. Demoz, K. Vermeesch, and D. Atkinson, 2019: Intercomparison of mixing layer heights from the National Weather Service ceilometer test sites and collocated radiosondes. J. Atmos. Oceanic Technol., 36, 129–137, https://doi.org/10.1175/JTECH-D-18-0058.1.
Holdridge, D., D. Turner, and D. Turner, 2015: FP6 Hesston, KS radiosonde data, version 1.0. UCAR/NCAR - Earth Observing Laboratory, accessed 27 March 2023, https://doi.org/10.5065/D6765CD0.
Holton, J. R., 2004: An Introduction to Dynamic Meteorology. 4th ed. Elsevier, 535 pp.
Holtslag, A. A. M., E. van Meijgaard, and W. C. de Rooy, 1995: A comparison of boundary layer diffusion schemes in unstable conditions over land. Bound.-Layer Meteor., 76, 69–95, https://doi.org/10.1007/BF00710891.
Holtslag, A. A. M., and Coauthors, 2013: Stable atmospheric boundary layers and diurnal cycles: Challenges for weather and climate models. Bull. Amer. Meteor. Soc., 94, 1691–1706, https://doi.org/10.1175/BAMS-D-11-00187.1.
Holzworth, G. C., 1964: Estimates of mean maximum mixing depths in the contiguous United States. Mon. Wea. Rev., 92, 235–242, https://doi.org/10.1175/1520-0493(1964)092<0235:EOMMMD>2.3.CO;2.
Hong, S.-Y., and H.-L. Pan, 1996: Nonlocal boundary layer vertical diffusion in a medium-range forecast model. Mon. Wea. Rev., 124, 2322–2339, https://doi.org/10.1175/1520-0493(1996)124<2322:NBLVDI>2.0.CO;2.
Hong, S.-Y., Y. Noh, and J. Dudhia, 2006: A new vertical diffusion package with an explicit treatment of entrainment processes. Mon. Wea. Rev., 134, 2318–2341, https://doi.org/10.1175/MWR3199.1.
Hu, X.-M., J. W. Nielsen-Gammon, and F. Zhang, 2010: Evaluation of three planetary boundary layer schemes in the WRF model. J. Appl. Meteor. Climatol., 49, 1831–1844, https://doi.org/10.1175/2010JAMC2432.1.
Iacono, M. J., J. S. Delamere, E. J. Mlawer, M. W. Shephard, S. A. Clough, and W. D. Collins, 2008: Radiative forcing by long-lived greenhouse gases: Calculations with the AER radiative transfer models. J. Geophys. Res., 113, D13103, https://doi.org/10.1029/2008JD009944.
Janjić, Z. I., 1994: The step-mountain eta coordinate model: Further developments of the convection, viscous sublayer, and turbulence closure schemes. Mon. Wea. Rev., 122, 927–945, https://doi.org/10.1175/1520-0493(1994)122<0927:TSMECM>2.0.CO;2.
Jia, W., and X. Zhang, 2020: The role of the planetary boundary layer parameterization schemes on the meteorological and aerosol pollution simulations: A review. Atmos. Res., 239, 104890, https://doi.org/10.1016/j.atmosres.2020.104890.
Jiménez, P. A., J. Dudhia, J. F. González-Rouco, J. Navarro, J. P. Montávez, and E. García-Bustamante, 2012: A revised scheme for the WRF surface layer formulation. Mon. Wea. Rev., 140, 898–918, https://doi.org/10.1175/MWR-D-11-00056.1.
Johns, R. H., and C. A. Doswell III, 1992: Severe local storms forecasting. Wea. Forecasting, 7, 588–612, https://doi.org/10.1175/1520-0434(1992)007<0588:SLSF>2.0.CO;2.
Johnson, A., and X. Wang, 2017: Design and implementation of a GSI-based convection-allowing ensemble data assimilation and forecast system for the PECAN field experiment. Part I: Optimal configurations for nocturnal convection prediction using retrospective cases. Wea. Forecasting, 32, 289–315, https://doi.org/10.1175/WAF-D-16-0102.1.
Johnson, A., X. Wang, and S. Degelia, 2016: OU MAP WRF ARW deterministic and ensemble 2-D fields, version 1.0. UCAR/NCAR - Earth Observing Laboratory, accessed 11 September 2023, https://doi.org/10.5065/D6125R2T.
Johnson, A., X. Wang, and S. Degelia, 2017: Design and implementation of a GSI-based convection-allowing ensemble-based data assimilation and forecast system for the PECAN field experiment. Part II: Overview and evaluation of a real-time system. Wea. Forecasting, 32, 1227–1251, https://doi.org/10.1175/WAF-D-16-0201.1.
Kotthaus, S., and Coauthors, 2023: Atmospheric boundary layer height from ground-based remote sensing: A review of capabilities and limitations. Atmos. Meas. Tech., 16, 433–479, https://doi.org/10.5194/amt-16-433-2023.
Kyrouac, J., Y. Shi, and M. Tuftedal, 2015: Surface meteorological instrumentation (MET). Atmospheric Radiation Measurement (ARM) user facility, accessed 20 December 2023, https://doi.org/10.5439/1786358.
LeMone, M. A., M. Tewari, F. Chen, and J. Dudhia, 2013: Objectively determined fair-weather CBL depths in the ARW-WRF model and their comparison to CASES-97 observations. Mon. Wea. Rev., 141, 30–54, https://doi.org/10.1175/MWR-D-12-00106.1.
LeMone, M. A., and Coauthors, 2019: 100 years of progress in boundary layer meteorology. A Century of Progress in Atmospheric and Related Sciences: Celebrating the American Meteorological Society Centennial, Meteor. Monogr., No. 59, Amer. Meteor. Soc., https://doi.org/10.1175/AMSMONOGRAPHS-D-18-0013.1.
Liu, S., and X.-Z. Liang, 2010: Observed diurnal cycle climatology of planetary boundary layer height. J. Climate, 23, 5790–5809, https://doi.org/10.1175/2010JCLI3552.1.
Lopez-Coto, I., M. Hicks, A. Karion, R. K. Sakai, B. Demoz, K. Prasad, and J. Whetstone, 2020: Assessment of planetary boundary layer parameterizations and urban heat island comparison: Impacts and implications for tracer transport. J. Appl. Meteor. Climatol., 59, 1637–1653, https://doi.org/10.1175/JAMC-D-19-0168.1.
Luo, H., and Coauthors, 2022: Interaction between aerosol and thermodynamic stability within the planetary boundary layer during wintertime over the North China Plain: Aircraft observation and WRF-Chem simulation. Atmos. Chem. Phys., 22, 2507–2524, https://doi.org/10.5194/acp-22-2507-2022.
Lv, Z., W. Wei, S. Cheng, X. Han, and X. Wang, 2020: Meteorological characteristics within boundary layer and its influence on PM2.5 pollution in six cities of North China based on WRF-Chem. Atmos. Environ., 228, 117417, https://doi.org/10.1016/j.atmosenv.2020.117417.
Mahrt, L., and O. Acevedo, 2023: Types of vertical structure of the nocturnal boundary layer. Bound.-Layer Meteor., 187, 141–161, https://doi.org/10.1007/s10546-022-00716-7.
McCaul, E. W., and C. Cohen, 2002: The impact on simulated storm structure and intensity of variations in the mixed layer and moist layer depths. Mon. Wea. Rev., 130, 1722–1748, https://doi.org/10.1175/1520-0493(2002)130<1722:TIOSSS>2.0.CO;2.
McCoy, R., S. Xie, Y. Zhang, and D. Cook, 2015: Quality Controlled Eddy Correlation Flux Measurement (30QCECOR). Atmospheric Radiation Measurement (ARM) user facility, accessed 4 August 2023, https://doi.org/10.5439/1989601.
Mellor, G. L., and T. Yamada, 1982: Development of a turbulence closure model for geophysical fluid problems. Rev. Geophys., 20, 851–875, https://doi.org/10.1029/RG020i004p00851.
Meng, J., Y. Gao, Y. Wang, L. Sheng, and S. Zhang, 2023: Evaluating the climate projections for China through the lens of the simulations of planetary boundary layer height. Atmos. Res., 294, 106975, https://doi.org/10.1016/j.atmosres.2023.106975.
Milovac, J., K. Warrach-Sagi, A. Behrendt, F. Späth, J. Ingwersen, and V. Wulfmeyer, 2016: Investigation of PBL schemes combining the WRF model simulations with scanning water vapor differential absorption lidar measurements. J. Geophys. Res. Atmos., 121, 624–649, https://doi.org/10.1002/2015JD023927.
Morrison, H., G. Thompson, and V. Tatarskii, 2009: Impact of cloud microphysics on the development of trailing stratiform precipitation in a simulated squall line: Comparison of one-and two-moment schemes. Mon. Wea. Rev., 137, 991–1007, https://doi.org/10.1175/2008MWR2556.1.
Muñoz-Esparza, D., R. D. Sharman, and S. B. Trier, 2020: On the consequences of PBL scheme diffusion on UTLS wave and turbulence representation in high-resolution NWP models. Mon. Wea. Rev., 148, 4247–4265, https://doi.org/10.1175/MWR-D-20-0102.1.
National Academies of Sciences, Engineering, and Medicine, 2018: Thriving on Our Changing Planet: A Decadal Strategy for Earth Observation from Space. The National Academies Press, 716 pp.
NCEP, 2015: NCEP GFS 0.25 degree global forecast grids historical archive. National Center for Atmospheric Research. Computational and Information Systems Laboratory, accessed 15 June 2022, https://doi.org/10.5065/D65DPWK.
Olson, J. B., J. S. Kenyon, W. Angevine, J. M. Brown, M. Pagowski, and K. Sušelj, 2019: A description of the MYNN-EDMF scheme and the coupling to other components in WRF–ARW. NOAA Tech. Memo. OAR GSD-61, 42 pp., https://doi.org/10.25923/n9wm-be49.
Olson, J. B., T. Smirnova, J. S. Kenyon, D. D. Turner, J. M. Brown, W. Zheng, and B. W. Green, 2021: A description of the MYNN surface-layer scheme. NOAA Tech. Memo. OAR GSL-67, 26 pp., https://repository.library.noaa.gov/view/noaa/30605.
Parish, T. R., and R. D. Clark, 2017: On the initiation of the 20 June 2015 Great Plains low-level jet. J. Appl. Meteor. Climatol., 56, 1883–1895, https://doi.org/10.1175/JAMC-D-16-0187.1.
Pearson, G., F. Davies, and C. Collier, 2010: Remote sensing of the tropical rain forest boundary layer using pulsed Doppler lidar. Atmos. Chem. Phys., 10, 5891–5901, https://doi.org/10.5194/acp-10-5891-2010.
Pichugina, Y. L., S. C. Tucker, R. M. Banta, W. A. Brewer, N. D. Kelley, B. J. Jonkman, and R. K. Newsom, 2008: Horizontal-velocity and variance measurements in the stable boundary layer using Doppler lidar: Sensitivity to averaging procedures. J. Atmos. Oceanic Technol., 25, 1307–1327, https://doi.org/10.1175/2008JTECHA988.1.
Powers, J. G., and Coauthors, 2017: The weather research and forecasting model: Overview, system efforts, and future directions. Bull. Amer. Meteor. Soc., 98, 1717–1737, https://doi.org/10.1175/BAMS-D-15-00308.1.
Qu, Y., Y. Han, Y. Wu, P. Gao, and T. Wang, 2017: Study of PBLH and its correlation with particulate matter from one-year observation over Nanjing, Southeast China. Remote Sens., 9, 668, https://doi.org/10.3390/rs9070668.
Santanello, J. A., Jr., and Coauthors, 2018: Land–atmosphere interactions: The LoCo perspective. Bull. Amer. Meteor. Soc., 99, 1253–1272, https://doi.org/10.1175/BAMS-D-17-0001.1.
Sathyanadh, A., T. Prabhakaran, C. Patil, and A. Karipot, 2017: Planetary boundary layer height over the Indian subcontinent: Variability and controls with respect to monsoon. Atmos. Res., 195, 44–61, https://doi.org/10.1016/j.atmosres.2017.05.010.
Schween, J. H., A. Hirsikko, U. Löhnert, and S. Crewell, 2014: Mixing-layer height retrieval with ceilometer and Doppler lidar: From case studies to long-term assessment. Atmos. Meas. Tech., 7, 3685–3704, https://doi.org/10.5194/amt-7-3685-2014.
Seibert, P., F. Beyrich, S.-E. Gryning, S. Joffre, A. Rasmussen, and P. Tercier, 2000: Review and intercomparison of operational methods for the determination of the mixing height. Atmos. Environ., 34, 1001–1027, https://doi.org/10.1016/S1352-2310(99)00349-0.
Shi, L., and Coauthors, 2021: Impact of the planetary boundary layer on air quality simulations over the Yangtze River Delta region, China. Atmos. Environ., 263, 118685, https://doi.org/10.1016/j.atmosenv.2021.118685.
Shin, H. H., and S.-Y. Hong, 2011: Intercomparison of planetary boundary-layer parametrizations in the WRF model for a single day from CASES-99. Bound.-Layer Meteor., 139, 261–281, https://doi.org/10.1007/s10546-010-9583-z.
Shin, H. H., and J. Dudhia, 2016: Evaluation of PBL parameterizations in WRF at subkilometer grid spacings: Turbulence statistics in the dry convective boundary layer. Mon. Wea. Rev., 144, 1161–1177, https://doi.org/10.1175/MWR-D-15-0208.1.
Sivaraman, C., and D. Zhang, 2010: Planetary Boundary Layer Height (PBLHTDL). Atmospheric Radiation Measurement (ARM) user facility, accessed 27 March 2023, https://doi.org/10.5439/1726254.
Sivaraman, C., S. McFarlane, E. Chapman, M. Jensen, T. Toto, S. Liu, and M. Fischer, 2013: Planetary boundary layer height (PBL) value added product (VAP): Radiosonde retrievals. Tech. Rep. DOE/SC-ARM/TR-132, 36 pp., https://doi.org/10.2172/1808688.
Skamarock, W. C., and Coauthors, 2019: A description of the advanced research WRF model version 4. NCAR Tech. Note NCAR/TN-556+STR, 145 pp., https://opensky.ucar.edu/islandora/object/technotes:576/datastream/PDF/download/citation.pdf.
Smith, E. N., J. A. Gibbs, E. Fedorovich, and P. M. Klein, 2018: WRF Model study of the Great Plains low-level jet: Effects of grid spacing and boundary layer parameterization. J. Appl. Meteor. Climatol., 57, 2375–2397, https://doi.org/10.1175/JAMC-D-17-0361.1.
Smith, E. N., J. G. Gebauer, P. M. Klein, E. Fedorovich, and J. A. Gibbs, 2019: The Great Plains low-level jet during PECAN: Observed and simulated characteristics. Mon. Wea. Rev., 147, 1845–1869, https://doi.org/10.1175/MWR-D-18-0293.1.
Srinivas, C. V., R. Venkatesan, and A. B. Singh, 2007: Sensitivity of mesoscale simulations of land–sea breeze to boundary layer turbulence parameterization. Atmos. Environ., 41, 2534–2548, https://doi.org/10.1016/j.atmosenv.2006.11.027.
Steeneveld, G. J., L. F. Tolk, A. F. Moene, O. K. Hartogensis, W. Peters, and A. A. M. Holtslag, 2011: Confronting the WRF and RAMS mesoscale models with innovative observations in the Netherlands: Evaluating the boundary layer heat budget. J. Geophys. Res., 116, D23114, https://doi.org/10.1029/2011JD016303.
Stensrud, D. J., 2009: Parameterization Schemes: Keys to Understanding Numerical Weather Prediction Models. Cambridge University Press, 459 pp.
Stull, R. B., 1988: An Introduction to Boundary Layer Meteorology. Kluwer Academic Publishers, 666 pp.
Sullivan, R., D. Billesbach, E. Keeler, and B. Ermold, 2015: Eddy Correlation Flux Measurement System (30ECOR). Atmospheric Radiation Measurement (ARM) user facility, accessed 4 August 2023, https://doi.org/10.5439/1025039.
Tangborn, A., B. Demoz, B. J. Carroll, J. Santanello, and J. L. Anderson, 2021: Assimilation of lidar planetary boundary layer height observations. Atmos. Meas. Tech., 14, 1099–1110, https://doi.org/10.5194/amt-14-1099-2021.
Teixeira, J., and Coauthors, 2008: Parameterization of the atmospheric boundary layer: A view from just above the inversion. Bull. Amer. Meteor. Soc., 89, 453–458, https://doi.org/10.1175/BAMS-89-4-453.
Teixeira, J., and Coauthors, 2021: Toward a global planetary boundary layer observing system: The NASA PBL incubation Study Team Rep., 134 pp., https://ntrs.nasa.gov/api/citations/20230001633/downloads/AFridlindPBLTowardsReport.pdf.
Tewari, M., and Coauthors, 2004: Implementation and verification of the unified Noah land surface model in the WRF Model. 20th Conf. on Weather Analysis and Forecasting/16th Conf. on Numerical Weather Prediction, San Antonio, TX, Amer. Meteor. Soc., 14.2A, https:/www2.mmm.ucar.edu/wrf/users/physics/phys_refs/LAND_SURFACE/noah.pdf.
Thompson, G., R. M. Rasmussen, and K. Manning, 2004: Explicit forecasts of winter precipitation using an improved bulk microphysics scheme. Part I: Description and sensitivity analysis. Mon. Wea. Rev., 132, 519–542, https://doi.org/10.1175/1520-0493(2004)132<0519:EFOWPU>2.0.CO;2.
Thompson, G., P. R. Field, R. M. Rasmussen, and W. D. Hall, 2008: Explicit forecasts of winter precipitation using an improved bulk microphysics scheme. Part II: Implementation of a new snow parameterization. Mon. Wea. Rev., 136, 5095–5115, https://doi.org/10.1175/2008MWR2387.1.
Tiedtke, M., 1989: A comprehensive mass flux scheme for cumulus parameterization in large-scale models. Mon. Wea. Rev., 117, 1779–1800, https://doi.org/10.1175/1520-0493(1989)117<1779:ACMFSF>2.0.CO;2.
Tran, T., H. Tran, M. Mansfield, S. Lyman, and E. Crosman, 2018: Four dimensional data assimilation (FDDA) impacts on WRF performance in simulating inversion layer structure and distributions of CMAQ-simulated winter ozone concentrations in Uintah Basin. Atmos. Environ., 177, 75–92, https://doi.org/10.1016/j.atmosenv.2018.01.012.
Träumner, K., C. Kottmeier, U. Corsmeier, and A. Wieser, 2011: Convective boundary-layer entrainment: Short review and progress using Doppler lidar. Bound.-Layer Meteor., 141, 369–391, https://doi.org/10.1007/s10546-011-9657-6.
Tucker, S. C., C. J. Senff, A. M. Weickmann, W. A. Brewer, R. M. Banta, S. P. Sandberg, D. C. Law, and R. M. Hardesty, 2009: Doppler lidar estimation of mixing height using turbulence, shear, and aerosol profiles. J. Atmos. Oceanic Technol., 26, 673–688, https://doi.org/10.1175/2008JTECHA1157.1.
Turner, D., 2016: FP6 ARM surface meteorology data, version 1.0. UCAR/NCAR - Earth Observing Laboratory, accessed 24 January 2024, https://doi.org/10.5065/D6RR1WN0.
UCAR/NCAR, 2015: FP1 ARM central facility radiosonde data, version 1.0. UCAR/NCAR Earth Observing Laboratory, accessed 27 March 2023, https://doi.org/10.26023/35HZ-2W5G-Z20H.
Vermeesch, K., 2015: FP2 Greensburg, KS radiosonde data, version 1.0. UCAR/NCAR - Earth Observing Laboratory, accessed 27 March 2023, https://doi.org/10.5065/D6FQ9TPH.
Xie, S., and Q. Tang, 2004: Station-based surface products (ARMBESTNS). Atmospheric Radiation Measurement (ARM) user facility, accessed 27 May 2023, https://doi.org/10.5439/1350681.
Xie, B., J. C. H. Fung, A. Chan, and A. Lau, 2012: Evaluation of nonlocal and local planetary boundary layer schemes in the WRF model. J. Geophys. Res., 117, D12103, https://doi.org/10.1029/2011JD017080.
Zhang, D., B. Ermold, and V. Morris, 1997: Ceilometer (CEIL). Atmospheric Radiation Measurement (ARM) user facility, accessed 27 May 2023, https://doi.org/10.5439/1181954.
Zhang, Y., and S. Li, 2019: Climatological characteristics of planetary boundary layer height over Japan. Int. J. Climatol., 39, 4015–4028, https://doi.org/10.1002/joc.6056.
Zhong, S., H. In, and C. Clements, 2007: Impact of turbulence, land surface, and radiation parameterizations on simulated boundary layer properties in a coastal environment. J. Geophys. Res., 112, D13110, https://doi.org/10.1029/2006JD008274.
Zhong, T., N. Wang, X. Shen, D. Xiao, Z. Xiang, and D. Liu, 2020: Determination of planetary boundary layer height with lidar signals using maximum limited height initialization and range restriction (MLHI-RR). Remote Sens., 12, 2272, https://doi.org/10.3390/rs12142272.
