1. Introduction
Local weather and regional climate are strongly influenced by the daytime convective circulations of energy, mass, and momentum within the atmospheric boundary layer (ABL), which is largely driven by fluxes between the land surface and atmosphere. While many different processes and feedback mechanisms are involved in land–atmosphere interactions (e.g., de Bruin 1983; Pielke et al. 1998; Betts 2000, 2004; Santanello et al. 2009; van Heerwaarden et al. 2009), the presence of clouds at the top of ABL plays an important roles in this system (e.g., Driedonks and Duynkerke 1989; Wetzel et al. 1996; Freedman et al. 2001; Ek and Holtslag 2004). Clouds significantly reflect and absorb incoming shortwave radiation as well as emit their own longwave thermal radiation. Also, the existence of clouds near the entrainment zone not only affects temperature and humidity within the ABL due to the entrainment process, but also increases the possibility of future precipitation.
Using a one-dimensional soil–vegetation–boundary layer model applied to a field site in Kansas and Oklahoma, Wetzel et al. (1996) concluded cloud development is controlled by two main factors: atmospheric structure (e.g., atmospheric stability, thermodynamics, humidity, etc.) and surface properties (e.g., soil moisture, vegetation type–cover, roughness, etc.). The feedback and interactions between atmospheric structure and surface properties have been intensively studied over the past decade (e.g., Betts 2000; Margulis and Entekhabi 2001; Barros and Hwu 2002; Daly et al. 2004; Ek and Holtslag 2004; Medeiros et al. 2005; Juang et al. 2007; Santanello et al. 2007, 2009; Porporato 2009; Siqueira et al. 2009; van Heerwaarden et al. 2009). For instance, using a mesoscale model to investigate the impact of land surface on summertime rainfall processes over the southern Great Plains, Barros and Hwu (2002) concluded that increased surface evaporation affecting the boundary layer stability provides a positive feedback for additional cloud and rainfall.
Two models widely used to investigate boundary layer clouds are the single-column model (SCM) and large-eddy simulation (LES), both of which have particular advantages for hydrometeorological studies. Generally including the basic parameterizations used in large-scale climate models (i.e., regional climate model or general circulation model), SCMs have been used in many studies (e.g., Jacobs and de Bruin 1992; Ek and Holtslag 2004; Juang et al. 2007; Porporato 2009; van Heerwaarden et al. 2009) investigating the processes, mechanisms, and feedbacks in land–atmosphere interactions because they do not require a large computational expense because of the simpler model structure. On the other hand, an LES model [and coupled LES–land surface model (LSM)] is able to provide a detailed and diagnostic characterization of the ABL (e.g., Stevens 2007; Savic-Jovcic and Stevens 2008; Stevens and Seifert 2008) because of its more explicit treatment of turbulent processes in the ABL and potential heterogeneity in land surface properties (Brown et al. 2002). In addition, an LES can provide a benchmark for the development of parameterizations that could be used in large-scale models. For example, Brown et al. (2002) presented an intercomparison project applying eight LES models to investigate the diurnal cycle of shallow cumulus clouds at the Atmospheric Radiation Measurement Program (ARM) Southern Great Plains (SGP) site (the Central Facility) on 21 June 1997. The same field observations on this particular day were later used in an SCM intercomparison study (Lenderink et al. 2004) to investigate the behavior of different parameterizations in cloud distribution.
Atmospheric stability 
The SCM is a useful tool for cloudy boundary layer study; however, it is highly dependent on the model parameterizations, which contain many uncertainties. In addition, an SCM is not able to resolve the detailed turbulent motions over heterogeneous surfaces (soil moisture, vegetation cover, surface characteristics, etc.), which may significantly affect the boundary layer structure and land–atmosphere interaction (Patton et al. 2005; Huang et al. 2009). SCM results also reveal some deficiencies in the cloud layer, such as too much cloud liquid water, unrealistic atmospheric profiles, and a large amount of numerical noise due to grid resolution (Lenderink et al. 2004).
Alternatively, a coupled LES–LSM can explicitly resolve both horizontal and vertical advections in turbulent dynamics of scalars and momentums. In addition, instead of using a statistical method to prescribe the cloud cover fraction, the coupled model is able to explicitly resolve the boundary layer cloud characteristics by using a microphysical scheme. Thus, in order to examine the hypothesis introduced in Ek and Holtslag (2004) using an SCM, in this study we perform a series of similar numerical experiments to further investigate and confirm the impact of atmospheric stability and near-surface soil moisture on cloud development using a recently developed coupled LES–LSM model. The use of the coupled LES–LSM allows us to explicitly resolve the surface states and fluxes, ABL properties, turbulent characteristics, entrainment processes, and more realistic cloud cover fraction and distributions. The objectives of this study are twofold. First, we would like to use the results of this study to corroborate the soil moisture–cloud feedback proposed by Ek and Holtslag (2004) in the context of SCM experiments. In that study, a statistical representation of fractional cloud cover [based on a Gaussian distribution of total water relative humidity (Ek and Mahrt 1991)] was used. Here, the LES–LSM allows for an explicit high-resolution treatment of clouds and investigation of the soil moisture–cloud feedback on cloud cover. The second objective of this work is to investigate cloud evolution in terms of other characteristics beyond fractional cover (i.e., cloud-base height and thickness), which cannot be obtained from SCM results.
The basic framework of the coupled LES–LSM model, including the microphysical cloud process representation in the LES, is briefly introduced in section 2. Section 3 describes the two-part experimental design, including 1) baseline simulation (to verify the daytime diurnal cycle of the selected day) and 2) synthetic cases to investigate the impacts of soil moisture and atmospheric stability on ABL characteristics. Section 4 presents simulated results from the baseline case to evaluate the coupled model performance. Analysis and discussion of results from the synthetic cases are presented in section 5, followed by section 6, which summarizes the conclusions and findings of this study.
2. Numerical model
a. Coupled land–atmosphere model framework
The numerical framework we used in this study is the University of California, Los Angeles (UCLA) LES–LSM coupled model (Huang and Margulis 2010), which integrates a radiative parameterization (Fu and Liou 1992), a high-resolution LES (Stevens et al. 2005; Huang and Margulis 2009), and a force-restore LSM (Noilhan and Planton 1989) to simulate the storage and fluxes of momentum, water, energy, and an arbitrary number of scalars (if necessary) in the system between the land surface and the overlying ABL. Radiative fluxes are estimated by the k-distribution method (Lacis and Oinas 1991) with gaseous parameterizations (Fu 1991), using the δ-four-stream approximation (Liou et al. 1988) integrated with the LES dynamics. The LES model solves the three-dimensional distribution of wind velocity, liquid-water potential temperature, and mixing ratio in a finite-difference Cartesian grid, while the subgrid-scale (SGS) parameterizations use the Lagrangian averaging scale-dependent model (Bou-Zeid et al. 2005; Huang et al. 2008) to dynamically calculate the spatially and temporally varying SGS parameters for the momentum (i.e., Smagorinsky coefficient) and scalar terms (i.e., Prandtl and Schmidt numbers). This LES model uses a leapfrog forward time-differencing scheme and fourth-order centered difference to solve momentum terms and advection, respectively; a forward-in-time differencing is applied on scalar terms. Initial conditions of vertical profiles of mean atmospheric states (i.e., potential temperature, specific humidity, wind, etc.) are specified. To break the symmetry of the numerical system and to generate turbulence, the 1D data is distributed over the 3D domain by adding small random perturbation at the first time step.
The LES provides the reference-level radiative fluxes and atmospheric states [collected at the first grid level (5 m)] as forcing for the LSM to calculate surface fluxes. These turbulent fluxes are used as the lower boundary conditions for the LES to further estimate the friction velocity and scalar scales (
The time step in the LES is 0.5 s to maintain the accuracy and stability of turbulence simulation, but longer time steps are used in the radiative transfer model (5 min) and in the LSM (10 min) because the temporal variation of simulated quantities in these two models is smaller than that of turbulence. This setting had been used in a previous study (Huang and Margulis 2010), in which a 12-h daytime diurnal cycle simulation from the National Aeronautic and Space Administration (NASA)-sponsored Soil Moisture–Atmosphere Coupled Experiment 2002 (SMACEX; Kustas et al. 2005) was performed using the coupled framework. Good agreement between the simulation and tower-based and remotely sensed observations was achieved for the surface fluxes and states at both domain-averaged and footprint scales. In addition, there was a good agreement with the convective boundary layer (CBL) potential temperature and humidity profiles. The SMACEX site located in Ames, Iowa, was predominantly covered by corn and soybean crops, which is different from the Cabauw site. Thus, we first performed a baseline simulation reproducing a daytime diurnal cycle of the Cabauw site (described in section 3a) to evaluate the performance of the coupled LES–LSM model at this study site. The reader is referred to Huang and Margulis (2010) for more details about the coupled framework.
b. Cloud microphysical processes
3. Experimental design
Following the work of Ek and Holtslag (2004), the experimental cases simulated were based on observations collected near the Cabauw experimental site, which is a flat domain for a distance of more than 20 km in the central part of the Netherlands, mainly dominated by heavy clay soil with a cover of nearly uniformly distributed short grass. This experimental site contains dense field- and tower-based measurements, including land surface characteristics (i.e., soil moisture, soil properties, vegetation, etc.) and atmospheric observations (i.e., air properties, wind speed, radiative and turbulent fluxes, etc.). The data collected at the site has been extensively reported in many ABL studies (e.g., Holtslag et al. 1995; van Ulden and Wieringa 1996; Beljaars and Bosveld 1997; Ek and Holtslag 2004). For additional details about the Cabauw site and observations, see Holtslag et al. (1995), van Ulden and Wieringa (1996), and Ek and Holtslag (2004).
In this study, we applied our coupled model setup to be consistent with the parameters used in previous studies at the Cabauw site. Because the spatial heterogeneity of vegetation and soil cover at the Cabauw site is small (Ek and Holtslag 2004), in this study the simulation domain is set as a homogenous flat soil surface completely covered by short grass. The numerical experiment is a 12-h daytime simulation starting at 0600 UTC on 31 May 1978. This same period was investigated in the study of a turbulence parameterization (Stull and Driedonks 1987), a one-dimensional boundary layer diffusion scheme (Holtslag et al. 1995), and cloud development under varied soil moisture using an SCM (Ek and Holtslag 2004). The horizontal simulation domain is approximately 10 × 10 km2 modeled by a 172 × 172 square grid with cells at 60-m resolution. The vertical height is about 4250 m simulated by 151 grid cells with 30-m spacing, except those near the surface where the grid resolution is reduced to 5 m. The land surface consists of uniformly distributed heavy clay (Ek and Holtslag 2004) with low hydraulic conductivity. The volumetric water content (Θ) at wilting point and at saturation are 0.23 and 0.53, respectively (Beljaars and Bosveld 1997). The vegetation cover is short grass with a specified height of 0.2 m (Holtslag et al. 1995), vegetation cover of 0.95, and leaf area index of 1.5 (Beljaars and Bosveld 1997). The momentum roughness length is set to one-tenth of the vegetation height. Finally, the relationship between the momentum roughness length 
a. Baseline case
The first simulation (i.e., baseline) in this study was performed to test the validity of the model in reproducing the daytime diurnal cycle of the Cabauw site on 31 May 1978, which has been used in Holtslag et al. (1995) and Ek and Holtslag (2004). Using the initial profiles illustrated in Ek and Holtslag (2004), soil temperature was initialized based on temperatures at two soil depths of 0.05 m (288.90 K) and 0.60 m (285.15 K). The corresponding volumetric soil moistures in the two layers were 0.33 and 0.53, respectively. Surface pressure is 1021.5 hPa (Holtslag et al. 1995). In the initial atmospheric profiles, the potential temperature starts from a value of 287.8 K at the land surface and piecewise linearly increases to 2200 m where the lapse rate above is set to a constant value of 3.5 K km−1. Similarly, specific humidity linearly decreases from a value of 7.8 g kg−1 at the ground. The profiles of potential temperature and specific humidity are illustrated in Figs. 1a and 1b (dashed lines), respectively, both of which are consistent with the profiles used in Ek and Holtslag (2004). The geostrophic wind components along the x and y directions are 
Initial atmospheric profiles of (a) potential temperature and (b) specific humidity. The dashed line is the observed (baseline) sounding (R2) from 31 May 1978 in Cabauw, the solid line represents a strong thermal stability condition (R1), and the dotted line represents a weak thermal stability condition (R3). The horizontal dashed line is the observed midafternoon atmospheric boundary layer height.
Citation: Journal of Hydrometeorology 12, 5; 10.1175/2011JHM1315.1
Initial states and parameters for the baseline case.
b. Experimental cases
In an analogous manner to Ek and Holtslag (2004) with their SCM framework, we performed a series of simulations with different soil moisture and different thermal stability using the coupled LES–LSM framework. We designed five sets (S1–S5) of simulations with soil moisture values varying from near-wilting point 
4. Results for the baseline case
The results for the baseline simulation are first presented to evaluate the performance of the coupled model for the experimental site. Results include reference-level atmospheric properties, surface fluxes, and daytime ABL structure. Here we present the simulated results compared to the observations shown in Ek and Holtslag (2004). The time series of model estimates of air temperature (Ta), specific humidity (qa), and wind speed (Ua) simulated at a reference level (i.e., the first level of the LES grid) are shown in Fig. 2a. To consider the difference between wind components along x and y directions, the model predictions presented here are spatially averaged states over a small footprint (the 4 × 2 gridcell region upwind from the location of the tower). The air temperature Ta starts at ~16°C at the beginning of the simulation and reaches a peak value of ~25°C around the midafternoon (~1430 UTC). The time series of simulated Ta using the coupled model matches the Cabauw tower observations (shown as markers) in the morning hours, but a slight difference (~1°C) is found compared to the peak value of observations. Starting from an initial value of ~7.5 g kg−1, the time series of qa approaches a local maximum around 1000 UTC and decreases around noontime before it rises again in the late afternoon. The estimates of qa are slightly higher than the observed values around midday. Small fluctuations are found for the time series of Ua in the first 2 h of simulation; however, the value is relatively stable during most of the daytime diurnal cycle before a small decrease occurs after 1600 UTC. Opposite to the result of 
Time series of model estimates on 31 May 1978 at the Cabauw site: (a) reference-level air temperature, specific humidity, and wind speed, and (b) surface energy fluxes. Markers represent observations collected at the Cabauw and the De Bilt sites (Ek and Holtslag 2004).
Citation: Journal of Hydrometeorology 12, 5; 10.1175/2011JHM1315.1
Figure 2b shows simulated footprint-averaged surface fluxes in the surface energy balance equation (SEB; 
Vertical profiles of domain-averaged ABL characteristics at four different times are illustrated in Fig. 3, which clearly shows the evolution of liquid-water potential temperature 
Vertical profiles of domain-averaged boundary layer characteristics: (a) potential temperature and (b) specific humidity. The light gray solid lines are the initial conditions shown in Fig. 1. The black dotted, dashed–dotted, dashed, and solid lines are simulation data selected at 0900, 1200, 1500, and 1800 UTC, respectively. Symbols are tower-based measurements collected at 1200 UTC and the horizontal gray dashed line (as in Fig. 1) is the observed midafternoon atmospheric boundary layer height.
Citation: Journal of Hydrometeorology 12, 5; 10.1175/2011JHM1315.1
5. Results for experimental cases
Results investigating the influence of soil moisture and atmospheric thermal stability on CBL characteristics and cloud properties using the LES–LSM coupled framework are described in the following section. The discussion includes the micrometeorological properties and surface fluxes, vertical profiles of boundary layer structure, and cloud characteristics.
a. Micrometeorological properties and surface fluxes
Figure 4 shows daily mean values of surface temperature 
Daytime mean value of spatially averaged micrometeorological properties: (a) surface temperature, (b) reference-level air temperature, (c) reference-level specific humidity, and (d) reference-level wind speed. Black, dark gray, and light gray color bars represent results from simulations with higher stability (R1), baseline (R2), and lower stability (R3) potential temperature profiles, respectively.
Citation: Journal of Hydrometeorology 12, 5; 10.1175/2011JHM1315.1
As illustrated in Fig. 4c, the value of the mean reference-level specific humidity increases from ~6 g kg−1 in the driest sets (cases S1R1, S1R2, and S1R3 with 
The daily mean value of the spatially averaged surface energy budget components, net radiation 
As in Fig. 4, but for model estimates of (a) net radiation, (b) latent heat flux, (c) sensible heat flux, and (d) ground heat flux.
Citation: Journal of Hydrometeorology 12, 5; 10.1175/2011JHM1315.1
b. Vertical profiles of boundary layer states
Because the LES explicitly resolves large-scale eddies that are responsible for the primary transport mechanisms of momentum, scalar, and energy, and only parameterizes small-scale eddies that are relatively unimportant in convective conditions, this model is able to provide a detailed evolution of boundary layer growth, which is difficult to obtain directly from experimental data and is fully parameterized in an SCM. Figure 6 shows the horizontally averaged vertical profiles of liquid-water potential temperature 
Vertical profiles of spatially averaged (a) liquid-water potential temperature, (b) specific humidity, and (c) liquid-water mixing ratio. Light gray, gray, and black lines (i.e., color from light to dark) are results with volumetric soil moisture of 0.28, 0.38, and 0.48. Dotted, dashed, and solid lines are results from simulations with R1, R2, and R3 thermal stability profiles. All data are 1-h averaged, centered at 1400 UTC.
Citation: Journal of Hydrometeorology 12, 5; 10.1175/2011JHM1315.1
Similar results are seen in Fig. 6b, which shows that the impact of atmospheric thermal stability on the mixed-layer 
c. Vertical profiles of domain-averaged fluxes
One advantage to using the LES to investigate land–atmosphere interactions is that it also provides turbulent statistics of the boundary layer characteristic, which are difficult or impossible to obtain from an SCM or field observations. Similar to Fig. 6, Fig. 7 shows the one-hour-averaged vertical profiles of domain-averaged thermal, total water vapor, and liquid-water fluxes centered at 1400 UTC. While the moister surface provides a lower sensible heat flux (Fig. 5c), all vertical profiles of thermal flux illustrated in Fig. 7a decreases from a maximum value at the surface and reaches a minimum value at the entrainment zone, representing a typical characteristic of the convective boundary layer. The impact of initial atmospheric stability on the vertical profile of thermal flux is small in cases with high soil moisture (black lines); however, it becomes more significant near the entrainment zone when the surface becomes drier (light gray lines). The values of surface thermal flux are similar in simulations of 
As in Fig. 6, but for the vertical profiles of spatially averaged (a) thermal flux, (b) total water vapor flux, and (c) liquid-water flux.
Citation: Journal of Hydrometeorology 12, 5; 10.1175/2011JHM1315.1
Values of turbulent characteristics near the entrainment zone in simulations with the driest surface (S1R1–3). In upper part, αθ (αq) is the ratio of entrainment and surface thermal (total water vapor) flux and the max Fluxql represents the maximum value of liquid-water flux illustrated in Fig. 7. In the lower part, max 
A more obvious impact of atmospheric stability on the dry surface cases is seen in the predicted moisture fluxes. Profiles of total water vapor flux illustrated in Fig. 7b illustrate that cases with a moist surface 
d. Vertical profiles of boundary layer state variance
Variance of 
As in Fig. 6, but for the vertical profiles of spatially averaged variance of (a) liquid-water potential temperature, (b) specific humidity, and (c) liquid-water mixing ratio.
Citation: Journal of Hydrometeorology 12, 5; 10.1175/2011JHM1315.1
Contrary to the results of 
In addition to the fact that the thermal stability has significant influence on both entrainments of heat and moisture, the results for boundary layer characteristics above also show that, even though dry soil provides less water vapor source at the surface, more liquid water at the top of the boundary layer is simulated when weak atmospheric stability is present. This result indicates the possibility of cloud development and leads to the following relative humidity analysis.
e. Analysis of evaporative mechanism in cloud formation
In this section, the LES results are analyzed using the RH tendency equation. Half-hourly data (during the period between 0800 and 1800 UTC) from simulated results are used to calculate the values of 
Evaporative fraction 
Citation: Journal of Hydrometeorology 12, 5; 10.1175/2011JHM1315.1
In the cases with stronger thermal stability (cases S1R1, S3R1, and S5R1), the nonevaporative term is smaller than 1 
An interesting finding is that the mean value of ne in the S3R3 case (solid square in Fig. 9c) is close to unity, which means the evolution of RH is independent from both evaporative and nonevaporative terms because 
f. Cloud cover fraction and cloud-base height
The ABL-top RH analysis discussed in the previous section has shown the liquid-water amount depends on not only soil moisture but also atmospheric stability. The increase/decrease of cloud development is a balance between the roles of soil moisture and thermal stability. The use of the LES–LSM model provides the explicit temporal and spatial evolution of cloud development. The results of cloud cover fraction from simulation sets with soil moisture Θ = 0.28, 0.38, and 0.48 are plotted in Fig. 10, where the upper row shows the time series of cloud cover fraction from 1000 UTC to the end of the simulation and the lower row shows the maximum value throughout the simulation. A binary setting is used to determine cloud cover. For each grid, if the amount of liquid water is larger than 0, cloud is deemed present over the entire grid cell; otherwise, no cloud is present. Thus, the cloud cover fraction is a ratio between the number of cloudy grid cells and the number of total cells over the simulation domain. Figure 10a illustrates that, for cases with strong atmospheric stability, the cloud cover fraction slightly increases before noon and reaches a maximum value around 1500–1600 UTC. The time series of cloud cover fraction in the two higher soil moisture simulation sets (i.e., Θ = 0.38 and 0.48) are similar and both are higher than that in the driest simulation set (Θ = 0.28). More clear evidence is seen in Fig. 10d, which shows that the maximum value of cloud cover fraction generally increases with increasing soil moisture. This plot of cloud cover fraction is consistent with the previous results of the RH analysis—increasing soil moisture increases the potential for cloud development for cases with strong thermal stability.
(a)–(c) Time series of cloud cover fraction and (d)–(f) the maximum value of cloud cover fraction in cases with (left) R1, (middle) R2, and (right) R3 initial potential temperature profiles. In time series plots, profiles with triangle, square, and circle markers are results selected from cases with volumetric soil moisture of 0.28, 0.38, and 0.48, respectively.
Citation: Journal of Hydrometeorology 12, 5; 10.1175/2011JHM1315.1
In Fig. 10b, the cloud cover fraction resulting from the dry surface simulation (Θ = 0.28) is generally the highest, while that from wet surface simulations is lower. The lower plot (Fig. 10e) shows that similar values for the maximum cloud cover fraction occur in the drier surface simulation sets (Θ = 0.28 and 0.38), while smaller values are seen in the moister surface simulation sets. This result matches the value of 
In addition to cloud cover fraction, the evolution of cloud-base height and cloud thickness selected from the same simulation sets (Θ = 0.28, 0.38, and 0.48) are illustrated in Fig. 11. In general, no matter what thermal stability is applied, both time series of cloud-base height and cloud thickness in all three cases start to rise around noon and reach a maximum in the midafternoon before decreasing. Results of simulations with strong thermal stability (R1) are illustrated in Fig. 11a, which shows the cloud-base height in all three cases are similar because the strong stability near the entrainment zone limits the growth of the boundary layer. Figure 11b shows the result of simulations with baseline thermal stability (R2). Less stable lapse rate allows for deeper mixed-layer development over dry surfaces than wet surfaces because of stronger surface sensible heat fluxes. As a result, the cloud-base height in the case of Θ = 0.28 is the highest during the simulation. The cloud thickness (Fig. 11e) in the cases of Θ = 0.28 and 0.38 are similar and both are higher than the result over the wettest case between noon and 1500 UTC, beyond which all three cases become similar. Figure 11c shows that both estimates of cloud-base height and cloud thickness significantly increase with decreasing soil moisture. The elevation of the cloud base over dry surfaces (i.e., Θ = 0.28) constantly rises after 1100 UTC and reaches more than 3500 m at about 1500 UTC. Meanwhile, lower cloud-base heights in cases with Θ = 0.38 and 0.48 reach the maximum heights of ~2000 and ~1500 m, respectively, before decreasing by the end of simulation. Figure 9c shows that for weaker thermal stability, a drier surface contributes more to the ABL-top RH tendency [i.e., large value of 
(a)–(c) Time series of cloud-base height and (d)–(f) cloud thickness in cases with (a),(d) R1, (b),(e) R2, and (c),(f) R3 initial potential temperature profiles. Profiles with triangle, square, and circle markers are results selected from cases with volumetric soil moisture of 0.28, 0.38, and 0.48, respectively.
Citation: Journal of Hydrometeorology 12, 5; 10.1175/2011JHM1315.1
6. Summary and conclusions
The role that soil moisture and thermal stability above the top of the boundary layer play in the land–atmosphere interaction and ABL-top cloud formation is investigated via a series of numerical experiments using a coupled LES–LSM model. In addition to reproducing the selected day at the Cabauw site, the Netherlands, 15 simulations designed from combinations of three ABL-top thermal stabilities and five soil moisture cases are performed using this coupled framework. For the baseline case, the simulated reference-level micrometeorological properties are well estimated, and only small differences occurring in specific humidity and wind speed around noon are seen in the comparison of observations collected at an eddy-covariance tower. In addition to the surface fluxes, the simulated boundary layer height from domain-averaged potential temperature and specific humidity also matches the observation in the midafternoon.
In addition to the expected results that the soil moisture provides a strong impact on latent–sensible heat flux and micrometeorological states, the impact of stability on net radiation is significant in the dry surface cases because of the ABL-top cloud development. While a higher soil moisture decreases (increases) the mixed-layer potential temperature (specific humidity), significant impacts of weak thermal stability is only seen in the dry surface cases, and this result confirms the hypothesis of Ek and Holtslag (2004). The LES–LSM coupled model provides more details in boundary layer turbulent characteristics. In simulations over drier surfaces, a weak thermal stability significantly enhances the entrainment fluxes at the top of the boundary layer, especially for moisture terms. The peak value of total water vapor (liquid water) flux near the entrainment zone in the driest surface case with weak stability is more than double (5 times) that with strong stability. Additionally, the variances of meteorological state at the top of the boundary layer can also vary significantly when the atmospheric thermal stability is weak.
The analysis of the coupled model outputs shows that a dry surface is able to provide a higher ABL-top relative humidity in the condition with weak thermal stability above the entrainment zone. The result of the coupled model not only supports the hypothesis purposed in Ek and Holtslag (2004) using a simple parameterization, but also quantifies the evaporative–nonevaporative processes to lead a relative humidity tendency. Additionally, the coupled model explicitly resolves the temporal evolution of cloud development, which illustrates that the cloud-cover fraction increases (decreases) with an increase of soil moisture when stronger (weaker) thermal stability exists above the top of the boundary layer. While the strength of convection dominates the cloud-base height, a significantly thicker cloud layer can develop over a wet surface if the ABL-top thermal stability is strong or over a dry surface if the stability is relatively weak. Results of this study also provide an explanation of the negative feedback of soil moisture on convection and convective rainfall discussed in Siqueira et al. (2009).
If the majority of ABL-top humidity to form the cloud is mainly contributed by surface moisture flux, the water vapor parcel leaving from the land surface must be raised to an elevation where the condensation of vapor occurs. With a strong ABL-top thermal stability acting as a limit on the growth of mixed layer, the water vapor parcel may not reach enough elevation for cloud development. On the other hand, if the thermal stability is weak, a stronger vertical convection over a dry surface can bring the water vapor parcel to the top of ABL to increase the possibility of cloud formation. In this study we only investigate the ABL-top stability; however, the influence of initial stability within the residual layer can also be critical. Thus, additional investigation is needed for future work. The results and analyses of this work using a coupled LES–LSM framework not only corroborate previous hypothesis of cloud cover in a single-column model study, but also provide additional information about the reactions of turbulent characteristic and entrainment flux because of the ABL-top stability and soil moisture.
Acknowledgments
This work was partially supported by the National Science Foundation under CAREER Grant EAR0348778. The authors especially thank Dr. Michael Ek in the Environmental Modeling Center at National Centers for Environmental Prediction for the invaluable discussions and comments. The help of Drs. Prakashan Korambath, Shao-Ching Huang, and Tajendra Vir Singh in the Academic Technology Services at UCLA with high performance computation and the valuable comments of three anonymous reviewers are gratefully acknowledged.
REFERENCES
Barros, A. P., and Hwu W. , 2002: A study of land-atmosphere interactions during summertime rainfall using a mesoscale model. J. Geophys. Res., 107, 4227, doi:10.1029/2000JD000254.
Beljaars, A. C. M., and Bosveld F. C. , 1997: Cabauw data for the validation of land surface parameterization schemes. J. Climate, 10, 1172–1193.
Betts, A. K., 2000: Idealized model for equilibrium boundary layer over land. J. Hydrometeor., 1, 507–523.
Betts, A. K., 2004: Understanding hydrometeorology using global models. Bull. Amer. Meteor. Soc., 85, 1673–1688.
Bou-Zeid, E., Meneveau C. , and Parlange M. B. , 2005: A scale-dependent Lagrangian dynamic model for large eddy simulation of complex turbulent flows. Phys. Fluids, 17, 025105, doi:10.1063/1.1839152.
Brown, A. R., and Coauthors, 2002: Large-eddy simulation of the diurnal cycle of shallow cumulus convection over land. Quart. J. Roy. Meteor. Soc., 128, 1075–1094.
Brutsaert, W., 1982: Evaporation into the Atmosphere: Theory, History and Applications. Springer, 299 pp.
Daly, E., Porporato A. , and Rodriguez-Iturbe I. , 2004: Modeling photosynthesis, transpiration, and soil water balance: Hourly dynamics during interstorm periods. J. Hydrometeor., 5, 546–558.
de Bruin, H. A. R., 1983: A model for the Priestley-Taylor parameter α. J. Climate Appl. Meteor., 22, 572–578.
Driedonks, A. G. M., and Duynkerke P. G. , 1989: Current problems in the stratocumulus-topped atmospheric boundary layer. Bound.-Layer Meteor., 46, 275–303.
Ek, M. B., and Mahrt L. , 1991: A formulation for boundary-layer cloud cover. Ann. Geophys., 9, 716–724.
Ek, M. B., and Mahrt L. , 1994: Daytime evolution of relative humidity at the boundary layer development. Mon. Wea. Rev., 122, 2709–2721.
Ek, M. B., and Holtslag A. A. M. , 2004: Influence of soil moisture on boundary layer cloud development. J. Hydrometeor., 5, 86–99.
Feingold, G., Walko R. L. , Stevens B. , and Cotton W. R. , 1998: Simulations of marine stratocumulus using a new microphysical parameterization scheme. Atmos. Res., 47–48, 505–528.
Freedman, J. M., Fitzjarrald D. R. , Moore K. E. , and Sakai R. K. , 2001: Boundary layer clouds and vegetation–atmosphere feedbacks. J. Climate, 14, 180–197.
Fu, Q., 1991: Parameterization of radiative processes in vertically nonhomogeneous multiple scattering atmospheres. Ph.D. dissertation, University of Utah, 259 pp.
Fu, Q., and Liou K. N. , 1992: On the correlated k-distribution method for radiative transfer in nonhomogeneous atmospheres. J. Atmos. Sci., 49, 2139–2156.
Garratt, J. R., 1992: The Atmospheric Boundary Layer. Cambridge University Press, 316 pp.
Holtslag, A. A. M., van Meijgaard E. , and de Rooy W. C. , 1995: A comparison of boundary layer diffusion schemes in unstable conditions over land. Bound.-Layer Meteor., 76, 69–95.
Huang, H.-Y., and Margulis S. A. , 2009: On the impact of surface heterogeneity on a realistic convective boundary layer. Water Resour. Res., 45, W04425, doi:10.1029/2008WR007175.
Huang, H.-Y., and Margulis S. A. , 2010: Evaluation of a fully coupled large-eddy simulation–land surface model and its diagnosis of land-atmosphere feedbacks. Water Resour. Res., 46, W06512, doi:10.1029/2009WR008232.
Huang, H.-Y., Stevens B. , and Margulis S. A. , 2008: Application of dynamic subgrid-scale models for large-eddy simulation of the daytime convective boundary layer over heterogeneous surfaces. Bound.-Layer Meteor., 126, 327–348.
Huang, J., Lee X. , and Patton E. G. , 2009: Dissimilarity of scalar transport in the convective boundary layer in inhomogeneous landscapes. Bound.-Layer Meteor., 130, 327–345.
Jacobs, C., and de Bruin H. , 1992: The sensitivity of regional transportation to land-surface characteristics: Significance of feedback. J. Climate, 5, 683–698.
Jacobson, M. Z., 1999: Effects of soil moisture on temperatures, winds, and pollutant concentrations in Los Angeles. J. Appl. Meteor., 38, 607–616.
Juang, J.-Y., Porporato A. , Stoy P. C. , Siqueira M. S. , Oishi A. C. , Detto M. , Kim H.-S. , and Katul G. G. , 2007: Hydrologic and atmospheric controls on initiation of convective precipitation events. Water Resour. Res., 43, W03421, doi:10.1029/2006WR004954.
Khairoutdinov, M., and Kogan Y. , 2000: A new cloud physics parameterization in a large-eddy simulation model of marine stratocumulus. Mon. Wea. Rev., 128, 229–243.
Kogan, Y. L., Khairoutdinov M. , Lilly D. K. , Kogan Z. N. , and Lin Q. , 1995: Modeling of stratocumulus cloud layers in a large-eddy simulation model with explicit microphysics. J. Atmos. Sci., 52, 2923–2940.
Kustas, W. P., Hatfield J. L. , and Prueger J. H. , 2005: The Soil Moisture–Atmosphere Coupling Experiment (SMACEX): Background, hydrometeorological conditions, and preliminary findings. J. Hydrometeor., 6, 791–804.
Lacis, A. A., and Oinas V. , 1991: A description of the correlated k distributed method for modeling nongray gaseous absorption, thermal emission, and multiple scattering in vertically inhomogeneous atmospheres. J. Geophys. Res., 96, 9027–9063.
Lenderink, G., and Coauthors, 2004: The diurnal cycle of shallow cumulus clouds over land: A single-column model intercomparison study. Quart. J. Roy. Meteor. Soc., 130, 3339–3364.
Liou, K.-N., Fu Q. , and Ackerman T. P. , 1988: A simple formulation of the delta-four-stream approximation for radiative transfer parameterizations. J. Atmos. Sci., 45, 1940–1947.
Margulis, S. A., and Entekhabi D. , 2001: Feedback between the land surface energy balance and atmospheric boundary layer diagnosed through a model and its adjoint. J. Hydrometeor., 2, 599–620.
Medeiros, B., Hall A. , and Stevens B. , 2005: What controls the mean depth of the PBL? J. Climate, 18, 3157–3172.
Noilhan, J., and Planton S. , 1989: A simple parameterization of land surface processes for meteorological models. Mon. Wea. Rev., 117, 536–549.
Patton, E. G., Sullivan P. P. , and Moeng C.-H. , 2005: The influence of idealized heterogeneity on wet and dry planetary boundary layers coupled to the land surface. J. Atmos. Sci., 62, 2078–2097.
Pielke, R. A., Sr., Dalu G. , Eastman J. , Vidale P. L. , and Zeng X. , 1998: Boundary layer processes and land surface interactions on the mesoscale. Clear and Cloudy Boundary Layers, A. A. M. Holtslag and P. G. Duynkerke, Eds., Royal Netherlands Academy of Arts and Sciences, 155–176.
Porporato, A., 2009: Atmospheric boundary-layer dynamics with constant Bowen ratio. Bound.-Layer Meteor., 132, 227–240.
Santanello, J. A., Jr., Friedl M. A. , and Kustas W. P. , 2005: An empirical investigation of convective planetary boundary layer evolution and its relationship with the land surface. J. Appl. Meteor., 44, 917–932.
Santanello, J. A., Jr., Friedl M. A. , and Ek M. B. , 2007: Convective planetary boundary layer interactions with the land surface at diurnal time scales: Diagnostics and feedbacks. J. Hydrometeor., 8, 1082–1097.
Santanello, J. A., Jr., Peters-Lidard C. D. , Kumar S. V. , Alonge C. , and Tao W.-K. , 2009: A modeling and observational framework for diagnosing local land–atmosphere coupling on diurnal time scales. J. Hydrometeor., 10, 577–599.
Savic-Jovcic, V., and Stevens B. , 2008: The structure and mesoscale organization of precipitating stratocumulus. J. Atmos. Sci., 65, 1587–1605.
Seifert, A., and Beheng K. D. , 2001: A double-moment parameterization for simulating autoconversion, accretion and self collection. Atmos. Res., 59–60, 265–281.
Seifert, A., and Beheng K. D. , 2006: A two-moment cloud microphysics parameterization for mixed-phase clouds. Part 1: Model description. Meteor. Atmos. Phys., 92, 45–66.
Siqueira, M., Porporato A. , and Katul G. , 2009: Soil moisture feedbacks on convection triggers: The role of soil–plant hydrodynamics. J. Hydrometeor., 10, 96–112.
Stevens, B., 2007: On the growth of layers of non-precipitating cumulus convection. J. Atmos. Sci., 64, 2916–2931.
Stevens, B., and Seifert A. , 2008: Understanding macrophysical outcomes of microphysical choices in simulations of shallow cumulus convection. J. Meteor. Soc. Japan, 86A, 143–162.
Stevens, B., and Coauthors, 2005: Evaluation of large-eddy simulations via observations of nocturnal marine stratocumulus. Mon. Wea. Rev., 133, 1443–1462.
Stull, R. B., and Driedonks A. G. M. , 1987: Applications of the transilient turbulence parameterization to atmospheric boundary-layer simulations. Bound.-Layer Meteor., 40, 209–239.
Tennekes, H., and Driedonks A. G. M. , 1981: Basic entrainment equations for the ABL. Bound.-Layer Meteor., 20, 515–531.
van Heerwaarden, C. C., de Arellano J. V. , Moene A. F. , and Holtslag A. A. M. , 2009: Interaction between dry-air entrainment, surface evaporation and convective boundary-layer development. Quart. J. Roy. Meteor. Soc., 135, 1277–1291.
van Ulden, A. P., and Wieringa J. , 1996: Atmospheric boundary layer research at Cabauw. Bound.-Layer Meteor., 78, 39–69.
Wetzel, P. J., Argentini S. , and Boone A. , 1996: Role of land surface in controlling daytime cloud amount: Two case studies in the GCIP-SW area. J. Geophys. Res., 101, 7359–7370.
