Abstract

It is still a daunting challenge for land surface models (LSMs) to correctly represent surface heat exchange for water-limited desert steppe ecosystems. This study aims to improve the ability of the Noah LSM to simulate surface heat fluxes through addressing uncertainties in precipitation forcing conditions, rapidly evolving vegetation properties, soil hydraulic properties (SHPs), and key parameterization schemes. Three years (2008–10) of observed surface heat fluxes and soil temperature over a desert steppe site in Inner Mongolia, China, are used to verify model simulations. The proper seasonal distribution of precipitation, along with more realistic vegetation parameters, can improve the simulation of sensible heat flux (SH) and the seasonal variability of latent heat flux. Correctly representing the low-surface exchange coefficient is crucial for improving SH for short vegetation like this desert steppe site. Relating Czil, the coefficient in the Noah surface exchange coefficient calculation, with canopy height h improves the simulated SH and the diurnal range of soil temperature over the simulation compared with using the default constant Czil. The exponential water stress formulation proposed here for the Jarvis scheme improves the partitioning between soil evaporation and transpiration. It is found that the surface energy fluxes are very sensitive to SHPs. This study highlights the important role of the proper parameter values and appropriate parameterizations for the surface exchange coefficient and water stress function in canopy resistance in capturing the observed surface energy fluxes and soil temperature variations for this desert steppe site.

1. Introduction

Approximately 8.8 × 104 km2 of grassland is classified as desert steppe in Inner Mongolia, which covers almost 34.7% of northern grassland in China and accounts for 18% of the total steppe area in China (Liao and Jia 1996). The desert steppe region is the most arid of all grassland regions and water shortage is common in this area, usually coincident with high temperature and wind speed and rapid evaporation (Humphrey 1958). These environments over the desert grasslands create less stable plant communities than other types of grasslands. Climate change, especially extreme precipitation events, directly affects plant growth and the partitioning between sensible heat flux (SH) and latent heat flux (LH) over this water-limited ecosystem (Gosz 1993; Neilson 1986; Mielnick et al. 2005; Li 2001). Chen et al. (2010) found that when precipitation is insufficient, the surface sensible heat transport and soil temperature variations become dominant land surface processes in arid regions. However, during long dry spells, evapotranspiration (ET) varies according to weather and wind conditions and its significance is magnified because ET continues to deplete the limited remaining water supplies in streams and the soil (Wetzel and Chang 1987). This could lead to more severe water stress in other periods. It is critical to capture these variations in land surface models (LSMs) in order to correctly represent the hydraulic and heat characteristics of arid and semiarid regions in coupled weather and climate models.

However, difficulty in realistically simulating surface energy fluxes over arid and semiarid regions remains in most LSMs, including the Noah LSM (Gao et al. 2008; Chen et al. 2010; Zeng et al. 2012). The Noah LSM is a community model and is developed through multi-institutional collaboration (Pan and Mahrt 1987; Chen et al. 1996; Chen and Dudhia 2001; Koren et al. 1999; Ek et al. 2003). It has been widely used in the research and operational communities and has been coupled to community weather and regional climate models (Chen and Dudhia 2001; Ek et al. 2003). While there are focused efforts to validate the Noah LSM for forest (Charusombat et al. 2010; Wu et al. 2011, 2012) and tall prairie grasslands (Chen et al. 1996; Sridhar et al. 2002; Chen et al. 2003; LeMone et al. 2007, 2008), little is known about its application to desert grassland. Given the special nature and importance of grassland over the arid region, the goal of this study is to improve the application of the Noah LSM over this arid region.

The Noah LSM includes numerous physical parameterization schemes and employs physical and physiological parameters such as the green vegetation fraction (GVF), leaf area index (LAI), surface albedo, roughness length, canopy resistance, and soil hydraulic properties (SHPs). These directly or indirectly measurable parameters are assigned typical values for different soil texture and land cover or are parameterized by different empirical formulations. Improper parameter selection and inaccuracies in the model formulations likely produce significant model errors (Godfrey and Stensrud 2010). For instance, monthly climatology of GVF in Noah LSM cannot describe the interannual variability of vegetation conditions (Hong et al. 2007, 2009) and consequently cannot capture the important variations in surface characteristics (Chang and Wetzel 1991; Crawford et al. 2001; Santanello and Carlson 2001; Kurkowski et al. 2003; Matsui et al. 2005; Hong et al. 2007, 2009). Noah overestimates the surface exchange coefficient for short vegetation with the surface exchange coefficient being too large (Chen and Zhang 2009; Ma et al. 2005; Yang et al. 2003, 2008; van der Velde et al. 2009; Hong et al. 2009). The simple Jarvis scheme for calculating the canopy resistance in Noah (Chen et al. 1996) using a constant minimum canopy resistance tends to overestimate LH over grasslands (Alfieri et al. 2008; Niyogi and Raman 1997). SHPs primarily control the movement of soil water, which is a particularly important component describing the land surface and provides a key link between the atmosphere and the water and energy balances at the land surface (Wei 1995; Robock et al. 2000; Leese et al. 2001; Koster et al. 2004; Seneviratne et al. 2010), but they suffer from great uncertainties (Gutmann and Small 2007).

Therefore, we focus on the impact of aforementioned key parameterization and vegetation factors and soil properties on the simulations of surface heat exchange in Noah using long-term offline simulations and field data over a desert steppe site. For arid ecosystems, grasslands are known to be highly sensitive to precipitation variability (Knapp and Smith 2001; Heisler-White et al. 2008) and can be influenced by episodic events of extreme precipitation (Mielnick et al. 2005). Nevertheless, the quality and continuity of locally measured precipitation and gridded precipitation data present different degrees of uncertainties. Therefore, we also analyze the uncertainty in precipitation and the relevant effects on surface energy flux simulations. We also recognize that the evaluation of LSMs is limited by uncertainty in flux observations and model structures, but investigating these aspects is beyond the scope of this investigation. Specifically, we address the following questions:

  1. What is the impact of uncertainty in precipitation amount and its seasonal distribution on simulated water vapor and heat flux at a water-limited desert steppe site?

  2. How sensitive are the surface energy flux simulations to rapidly evolving vegetation and hydraulic parameters?

  3. What is the sensitivity of surface energy flux simulations to the surface exchange coefficient and to canopy resistance parameterization?

These questions are investigated with regard to peak-growing-season SH, LH, and diurnal range of soil temperature. Section 2 describes the observation site information, and section 3 depicts the design of numerical experiments that address uncertainties in vegetation parameters, the Jarvis canopy resistance scheme, surface exchange coefficient, and SHP database. Section 4 discusses the model results, followed by a summary in section 5.

2. Site description

The Desert Steppe Ecosystem Research Station (44°05′N, 113°34′E) is located in Sonid Zuoqi, Inner Mongolia, China, with a growing season spanning from late April to October (peak growth period being from June to August; Yang and Zhou 2011). Its plant community consists of the dominant grasses Stipa klemenzii and Allium polyrrhizum. Grass clusters have an average height of 0.2–0.35 m at peak growth stage, with a root depth of 0.3–0.5 m. Soil type is brown calcic and is classified as sandy loam by the international soil texture classification system. The climate of this region is arid–semiarid, temperate, and continent climate with an annual mean precipitation of 183.9 mm, and most of the annual precipitation falls in summer. Severe droughts frequently occur in spring and early summer, causing low productivity of vegetation (Zhang 1992; Yu et al. 2003). The annual mean air temperature is 3.2°C, with monthly-mean temperature of −18.7°C in January and of 22.1°C in July.

Observations used in this study were collected from 2008 to 2010. Photosynthetically active radiation (PAR) and net radiation (Rn) were measured at a height of 2.4 m above the ground, using a quantum sensor (LI-190SB, LI-COR Inc., Lincoln, Nebraska) and a four-component net radiometer (CNR-1, Kipp & Zonen, Delft, the Netherlands), respectively. Air temperature (Ta) and relative humidity (RH) were measured at two levels (2.0 and 3.4 m; HMP45C, Vaisala, Helsinki, Finland). A horizontal wind speed sensor (014A, Campbell Scientific Inc., Logan, Utah) at 2.0-m measured horizontal wind speed, and a wind set sensor (034B, Met One Inc., Grants Pass, Oregon) at 3.4-m measured horizontal wind speed and wind direction. Precipitation was measured with a tipping-bucket rain gauge (Model 52203, RM Young Inc., Traverse City, Michigan) above the canopy. Soil water content (SWC) profile was measured at the depths of 0.1, 0.2, 0.3, and 0.4 m by time domain reflectometer probes (CS616, Campbell Scientific Inc.). All meteorological data were recorded every 2 s, and half-hourly mean data were logged by a data logger (CR23X, Campbell Scientific Inc.).

Surface flux (e.g., exchange of heat, water vapor, and CO2) data were obtained with the open-path eddy covariance system installed at a height of 2.0 m. The signals were recorded at 10 Hz by a data logger (CR5000, Campbell Scientific Inc.), and the half-hourly flux data were computed by the eddy covariance method (Monteith and Unsworth 1990). The corrections to raw flux estimates include double-coordinate rotations (Wilczak et al. 2001) and density effects of heat and water vapor transfer (Webb et al. 1980). The available datasets are screened to remove any anomalous half-hourly fluxes that resulted from malfunction of the sensors and outliers due to occasional spikes in half-hourly flux values for unknown reasons (Papale et al. 2006). Details of data processing can be found in Yang and Zhou (2011) and Zhang et al. (2012). Additionally, LAI was measured using the destructive method once a month from May to September in 2008 and 2009.

The energy-balance ratio (EBR) is usually used to assess the performance of the eddy covariance system (Wilson et al. 2002). It is calculated using the following equation for hourly periods where all the data (Rn, SH, LH, and G) are available:

 
formula

where G is the ground heat flux (all data in W m−2).

The discrepancy (i.e., 1 − EBR) in energy-balance closure is a bias that varies from 0% to 30% (Twine et al. 2000). With the discrepancy larger than 30%, the utility of SH and LH measurements for model validation or calibration is greatly reduced (Kustas et al. 1999; Wilson et al. 2002; Twine et al. 2000). The discrepancy over this site is 11%, 19%, and 16% for 2008, 2009, and 2010, respectively. Therefore, we did not adjust SH and LH in this study.

3. LSM and numerical experiments

The Noah LSM employs a Penman potential evaporation method to calculate LH (Mahrt and Ek 1984) and includes a four-layer soil model with the soil thickness of 0.1, 0.3, 0.6, and 1 m for each layer (Chen et al. 1996; Schaake et al. 1996). It uses a roughness Reynolds number approach for determining the ratio between the roughness length for heat (Zot) and that for momentum (Zom; Zilitinkevich 1995; Chen et al. 1997). The Noah LSM has been extended with canopy resistance as a function of soil water availability and atmospheric conditions (i.e., Jarvis scheme) and a surface runoff scheme by Chen et al. (1996). The Noah LSM generally performs well in various LSM intercomparison studies (Qu et al. 1998; Mitchell et al. 2004; Rodell et al. 2004; Kato et al. 2007) and has been coupled to and evaluated in regional weather and climate models (Chen et al. 1997; Chen and Dudhia 2001; Betts et al. 1997; Yucel et al. 1998; Ek et al. 2003).

Hourly tower measurements of Ta, mixing ratio, wind speed, wind direction, atmospheric pressure, downward shortwave radiation, and downward longwave radiation from 2008 to 2010 are used to drive the Noah LSM in 1D offline mode. To address our scientific questions, we design sensitivity tests with different precipitation datasets (section 3a) and with green vegetation fraction and associated parameters (section 3b), testing a revised surface-exchange-coefficient parameterization and Jarvis scheme (section 3c) and an ensemble of SHP databases (section 3d).

a. Precipitation datasets

It should be noted that the tipping-bucket rain gauge at this site only provides rainfall measurements but not snowfall (Fig. 1). To test the influences of different precipitation data on flux simulation, we use three additional precipitation datasets to prepare the precipitation input: 1) gauge-based daily precipitation data (0.5° × 0.5°) from the Climate Precipitation Center (CPC) of the National Oceanic and Atmospheric Administration (NOAA; Fig. 1), 2) daily precipitation from a nearby automatic weather station (AWS; approximately 30 km away from this site; Fig. 1), and 3) 3-hourly precipitation data from the Global Land Data Assimilation System (GLDAS; Rodell et al. 2004) precipitation (0.25° × 0.25°).

Fig. 1.

Monthly precipitation from four different datasets in 2008, 2009, and 2010: in situ local rainfall measurements, daily precipitation from NOAA/CPC, daily precipitation from nearby AWS, and the new adjusted daily precipitation dataset COMBINE.

Fig. 1.

Monthly precipitation from four different datasets in 2008, 2009, and 2010: in situ local rainfall measurements, daily precipitation from NOAA/CPC, daily precipitation from nearby AWS, and the new adjusted daily precipitation dataset COMBINE.

First, daily CPC and AWS precipitation data are used to fill the gap of in situ rainfall data in the wintertime. When the daily Ta is below 0°C, the average daily precipitation of these two datasets is added into the in situ rainfall measurements to yield a new adjusted daily precipitation dataset (named as COMBINE in Fig. 1). After this adjustment, the annual total precipitation increases from 136.3, 190.8, and 141.3 mm to 153.3, 205.4, and 162.6 mm for 2008, 2009, and 2010, respectively. This confirms that the majority of precipitation falls in summer over this site.

After we obtain the new adjusted daily precipitation data (i.e., COMBINE), we prepare 2-hourly precipitation forcing datasets based on COMBINE since hourly (not daily) precipitation is required to drive the Noah LSM. The adjusted GLDAS data are used as one dataset to test the uncertainty in precipitation-forcing condition in this study. For this dataset, we interpolate 3-hourly GLDAS precipitation data evenly into hourly precipitation data, and then we use the monthly total amount of COMBINE precipitation to rescale the interpolated hourly GLDAS precipitation data (GLDAS in Fig. 2). The other hourly precipitation forcing dataset is in situ hourly rainfall measurements when Ta is above 0°C plus the rescaled hourly GLDAS precipitation when Ta is below 0°C (COMB in Fig. 2). Note that the difference between GLDAS and COMB is the precipitation distribution in the summertime, but the monthly total amounts are same, and COMB for this period is actually in situ rainfall measurements. The difference between in situ rainfall measurements and COMB is the winter precipitation distribution.

Fig. 2.

Daily precipitation (mm) from June to August in (a) 2008, (b) 2009, and (c) 2010 from experiments: COMB (daily accumulated combined precipitation) and GLDAS (daily accumulated rescaled GLDAS precipitation).

Fig. 2.

Daily precipitation (mm) from June to August in (a) 2008, (b) 2009, and (c) 2010 from experiments: COMB (daily accumulated combined precipitation) and GLDAS (daily accumulated rescaled GLDAS precipitation).

The year 2010 has the driest summer (June–August) among these three years, and its total precipitation is approximately 50–60 mm lower than the other two years. There are obvious differences between GLDAS and COMB precipitation distribution (Fig. 2), and the correlation coefficients between daily accumulated GLDAS and COMB from June to August in 2008, 2009, and 2010 are 0.529, 0.923, and 0.756, respectively, indicating their difference is largest in 2008 and smallest in 2009.

b. Vegetation parameters

The Noah LSM uses a 5-yr monthly climatology as default values for GVF and therefore does not have interannual variations (Fig. 3d). Many studies (e.g., Jacquemin and Noilhan 1990; Chen et al. 1996, 2003; Betts et al. 1997; Miller et al. 2006) showed that GVF is a parameter of first-order importance to partition evaporation between vegetated and nonvegetated surfaces. Therefore, we calculate submonthly GVF for each year using satellite data following the widely used method of Gutman and Ignatov (1998) to represent the real-time evolving of vegetation [normalized difference vegetation index (NDVI)]:

 
formula

Daily NDVI is interpolated from Moderate Resolution Imaging Spectroradiometer (MODIS) MOD15A2 8-day composite data (available from the Oak Ridge National Laboratory Distributed Active Archive Center at http://daac.ornl.gov/MODIS/modis.html). The current model uses 0.07 for minimum NDVI (NDVImin) and 0.588 for maximum NDVI (NDVImax) based on the method of Barlage and Zeng (2004). These values depend on land-cover types and are different from those used in the equation of Gutman and Ignatov (1998), which was adapted from a dense vegetation assumption.

Fig. 3.

MODIS-based GVF and daily precipitation in (a) 2008, (b) 2009, and (c) 2010 and (d) default GVF in Noah LSM. Lines represent GVF and bars precipitation.

Fig. 3.

MODIS-based GVF and daily precipitation in (a) 2008, (b) 2009, and (c) 2010 and (d) default GVF in Noah LSM. Lines represent GVF and bars precipitation.

The MODIS-based GVF is around 45% (Fig. 3), generally higher than the climatology. The MODIS-based peak GVF occurred in early September of 2008, 2 months later than the other two years, reflecting a strong interannual variability of GVF. As the Fig. 3 shows, MODIS-based GVF correlates precipitation very well. During the growing season, the large increases in GVF happen after strong precipitation events, and long periods of antecedent precipitation can contribute to seasonal GVF increases.

The observed LAI and derived roughness length over the desert steppe site are much lower than the default values for grassland in Noah, but the in situ albedo is much higher than the default value. These vegetation parameters are modified based on observations over this site (see Table 1).

Table 1.

Vegetation parameters setup.

Vegetation parameters setup.
Vegetation parameters setup.

c. Parameterization scheme revision

1) Surface heat exchange coefficient parameterization

The Noah LSM employs Zilitinkevich’s formulation to represent the relationship between roughness length for heat (Zot) and momentum (Zom) as expressed in Eq. (3) below (Zilitinkevich 1995; Chen et al. 1997). It has been well documented that by using an constant empirical coefficient ,Czil, specified as 0.1 in Noah and in Chen et al. (1997), usually overestimates surface heat exchange coefficient for short vegetation (Yang et al. 2003, 2008; Ma et al. 2005; LeMone et al. 2008; Chen and Zhang 2009). Chen and Zhang (2009) formulate Czil as function of canopy height h [Eq. (4)] to improve the surface exchange coefficients and surface heat flux simulations:

 
formula

and

 
formula

where k is the von Kármán constant, Re is the roughness Reynolds number, h is given in meters. In this study, we use both the constant (of 0.1) and Eq. (4) to quantify the response of flux simulations to Czil.

2) Jarvis scheme

To calculate canopy resistance, Noah uses the simple Jarvis scheme that is an empirical formula of representing effects of LAI, PAR, soil moisture, Ta, and vapor pressure deficit (VPD):

 
formula
 
formula
 
formula
 
formula

and

 
formula

where rc,min is the minimum canopy resistance (40 s m−1 for grass), rc,max is the maximum canopy resistance (5000 s m−1), Rgl is taken as a constant of 100 W m−2, PAR is in watts per square meter, VPD is in kilopascals, Ta is in kelvins, is the soil moisture for the ith layer, di is the thickness of the ith soil layer, is the wilting point, is the field capacity soil moisture, and dtot is the total thickness of the rooting depth (Chen and Dudhia 2001). Since the root depth of grass at this desert steppe is 0.3–0.5 m, the default number of three root-zone layers for grass in Noah is replaced by two root-zone layers in this study, and consequently, the dtot is 0.4 m.

It is found that change in stomatal resistance by a factor of 4 can alter the energy fluxes by a factor of 2 for regions such as semiarid grassland in the southern United States (Niyogi and Raman 1997). Many studies were conducted to verify and improve the empirical Jarvis scheme in Noah. Chen et al. (1996) suggested that using a nonlinear soil moisture stress function could maintain a nonzero evaporation beyond the wilting point and reduce the evaporation when soil moisture is near the field capacity. Other studies focused on tuning the minimum stomatal resistance to improve LH simulations (Alfieri et al. 2008; Niyogi et al. 2009; Kumar et al. 2011).

However, over this desert steppe site, our previous study suggests replacing the linear soil moisture stress function in Eq. (7) with an exponential formulation (Zhang et al. 2014). Using the inverted Penman–Monteith function (Monteith 1965) and LAI as the scaling factor to calculate canopy resistance (Amer and Hatfield 2004; Alfieri et al. 2008) with eddy correlation flux observations from 2008 to 2010, we find that canopy resistance decreased exponentially with soil moisture; therefore, we propose the following exponential function of water stress while the other terms are kept the same as the original Jarvis scheme (Zhang et al. 2014):

 
formula

where SWC is the averaged volumetric soil water in the first 0.4-m soil depth.

d. SHPs database

Schaap and Leij (1998) present a collection of three databases that use the van Genuchten SHP model to estimate the relationship between soil moisture, hydraulic conductivity, and hydraulic head (van Genuchten 1980). An additional SHP model, the Campbell SHP model (Campbell 1974), is currently applied to most LSMs, including the Noah LSM. Gutmann and Small (2007) converted this SHPs database to obtain a new SHPs database for its application to the Campbell model. They found that the Noah LSM with the van Genuchten model (with Schaap and Leij’s data) and the Campbell model (with the converted SHPs data) showed nearly identical results. Modifying SHPs can drastically improve model behavior (Gutmann and Small 2007). Therefore, in this study, we ran the Noah LSM with 110 sets of optimized SHPs for the Campbell model to analyze their effects on surface heat fluxes.

e. Numerical experiments

A number of numerical experiments are conducted to assess the uncertainties in precipitation forcing conditions, model parameters, and parameterization schemes and to investigate pathways for improving the Noah LSM. These experiments are listed in Table 2. The control experiment (CTL) uses the in situ hourly rainfall measurements and default vegetation parameters. Experiments GLD and COMB are both based on CTL but with rescaled hourly GLDAS and combined hourly precipitation, respectively. Comparing the experiments GLD and COMB, we can quantify the effects of different precipitation distribution on energy flux simulation. The seasonal variability of LAI, roughness length, and albedo are calculated using time-varying GVF and constrained by related minimum and maximum values in Noah LSM (Table 1). Therefore, these three parameters together with GVF are used to evaluate the degree to which these parameters affect SH and LH with different precipitation data (GLD versus GLDVEG, COMB versus COMBVEG). The experiments CZIL and RC are designed to test the sensitivity of SH and LH simulations to parameterization uncertainty in surface heat exchange coefficient and canopy resistance, respectively. The experiment SOIL is the selected best run from four sets of 110 SHP experiments (based on COMB, COMBVEG, CZIL, and RC, i.e., 440 experiments altogether) using the least root-mean-square error (RMSE) between simulations and observations as the measure.

Table 2.

Summary of numerical experiments.

Summary of numerical experiments.
Summary of numerical experiments.

Model spinup is applied to each of the different set of sensitivity tests, and for each spinup experiment, the Noah LSM is run five times repeatedly with the first-year forcing conditions to ensure the equilibrium soil state.

4. Results and discussions

a. Response of simulated surface heat fluxes to precipitation

As mentioned in section 3a, the correlation coefficient between daily accumulated rescaled GLDAS and COMBINE precipitation in 2008 is smaller than 2009 (i.e., greater difference in 2008). Comparison of experiments GLD to COMB and GLDVEG to COMBVEG produces smaller correlation coefficients for SH and LH simulations in 2008 relative to 2009 (i.e., greater difference in 2008). With sufficient precipitation amount, this relationship between precipitation distribution and fluxes is expected. Greater differences in precipitation distribution should result in greater differences in fluxes simulation. This implies that seasonal shift in rainfall due to climate change would cause significant consequences in energy fluxes when the precipitation is relatively ample over this site, which has not been paid sufficient attention or even ignored in the past.

However, the larger correlation coefficients in flux simulation with a smaller correlation coefficient in precipitation in 2010 compared with 2009 indicates that the seasonal precipitation distribution has no significant impact in 2010 (Table 3). This might be attributed to the extremely low precipitation during the peak growing season in 2010, approximately 50–60 mm lower than the other two years. Guo et al. (2012) reported that the seasonal distribution of precipitation has less impact on grass growth than annual total precipitation amount when the total precipitation is extremely low.

Table 3.

Pearson product-moment correlation coefficients between fluxes simulations from June to August in 2008–10 with different precipitation datasets (daily average).

Pearson product-moment correlation coefficients between fluxes simulations from June to August in 2008–10 with different precipitation datasets (daily average).
Pearson product-moment correlation coefficients between fluxes simulations from June to August in 2008–10 with different precipitation datasets (daily average).

The difference of SH produced by different seasonal precipitation distribution is directly compensated by changes of LH (Fig. 4). The simulations with the realistic vegetation parameters in GLDVEG and COMBVEG produce the similar differences between GLD and COMB (Fig. 4).

Fig. 4.

The minimum, median, and maximum differences in daily-mean simulated (a) SH and (b) LH from different precipitation datasets without–with the vegetation (VEG) parameter modification for June–August in 2008–10. Dashed lines with open circles represent the range of differences between experiments GLD and COMB and with open triangles between GLDVEG and COMBVEG. The circles and triangles indicate the median values; the lower and upper ends of the whiskers represent the minimum and maximum, respectively.

Fig. 4.

The minimum, median, and maximum differences in daily-mean simulated (a) SH and (b) LH from different precipitation datasets without–with the vegetation (VEG) parameter modification for June–August in 2008–10. Dashed lines with open circles represent the range of differences between experiments GLD and COMB and with open triangles between GLDVEG and COMBVEG. The circles and triangles indicate the median values; the lower and upper ends of the whiskers represent the minimum and maximum, respectively.

Simulations using rescaled hourly GLDAS precipitation are generally worse than the combined hourly precipitation. Such degradation is expected, because the gridded data represent the average precipitation for a large area of 0.25° × 0.25° rather than that for a single point. We consider the data combining locally measured rainfall as the best approximation to ground truth, even though the local observations are not error free. For this specific site, high temperature and a dry environment favor quick evaporation of light rainfall from the tipping-bucket rain gauge, which may have underestimated precipitation amount. This undercatch might be one reason that the Noah LSM underestimates LH when it is relatively dry in the early growing season (not shown) and overestimates LH in the following growing season.

b. Response of fluxes to vegetation parameters

The MODIS-based GVF and the new vegetation parameters in Table 1 produce less difference in LH than in SH (Fig. 5). Because LH consists of three components in Noah,

 
formula

where Tr is the transpiration, Ec is the evaporation of precipitation intercepted by the canopy, and Eg is the bare soil evaporation. All three of these evaporation components are directly affected by GVF. Whether LH decreases or increases with higher GVF depends on which term is dominant (Miller et al. 2006). Figure 6 illustrates the response of SH and LH to GVF modifications. The relationship is negative between the GVF and SH differences but positive between GVF and LH. There is a threshold for GVF (~0.15) to balance the variations of Eg and Tr. Above this threshold, increasing GVF could affect specific evaporation component but not the total LH.

Fig. 5.

As in Fig. 4, but for normalized differences produced by GVF and vegetation parameters. Dashed lines with open circles represent (GLDVEG − GLD)/GLD and with open triangles (COMBVEG − COMB)/COMB.

Fig. 5.

As in Fig. 4, but for normalized differences produced by GVF and vegetation parameters. Dashed lines with open circles represent (GLDVEG − GLD)/GLD and with open triangles (COMBVEG − COMB)/COMB.

Fig. 6.

COMBVEG − COMB difference in daily mean GVF vs COMBVEG − COMB difference in the simulated daily mean (a) SH and (b) LH.

Fig. 6.

COMBVEG − COMB difference in daily mean GVF vs COMBVEG − COMB difference in the simulated daily mean (a) SH and (b) LH.

In general, the vegetation parameter modifications in conjunction with two precipitation datasets improve the simulated SH by decreasing its averaged RMSE by ~20% compared with CTL (Table 4a), and its biases decrease obviously for the whole period (Fig. 7). Regarding the LH simulations, the biases decrease significantly for both GLDVEG and COMBVEG in 2009 and in July and August of 2010 compared with the CTL simulations (Fig. 8), but slightly increase for the other period. The averaged RMSE of LH does not change significantly over the whole period (Table 4b). The simulations with MODIS GVF and the modified vegetation parameters produce more LH compared to CTL during the period of peak MODIS GVF (i.e., July and August 2008, June and July 2009, and June 2010), but decrease LH for other period (i.e., August 2009 and July and August 2010). However, when the growth patterns are similar between MODIS GVF and default GVF, this impact is not significant (e.g., in 2008). The COMBVEG simulations still consistently overestimate midday SH and underestimate SH in late afternoon compared with observations. LH is underestimated in the early summer (e.g., June 2008 and June and July 2009) and is overestimated in the following period. This occurs despite a significantly improved characterization of the initial land surface conditions through spinup. This result is consistent with the report from Robock et al. (2003), who found that initial conditions with greater accuracy do not necessarily guarantee an improvement in model performance.

Table 4a.

RMSE and index of agreement (IOA) between simulated and observed SH [daytime, 0800–1600 local time (LT)] for the different experiments for June–August in 2008–10. The simulations were done in sequence, except that GLD and COMB, and GLDVEG, and COMBVEG were done in parallel; and SOIL is based on CZIL. Bold numbers represent improved simulations, and italic numbers represent degraded simulations compared with the previous experiment. The “Average” is the RMSE or IOA calculated over the whole period from June to August in 2008–10. The RMSE is calculated using and the IOA with , where and are simulated and observed fluxes, respectively, and is the averaged observed fluxes for each month.

RMSE and index of agreement (IOA) between simulated and observed SH [daytime, 0800–1600 local time (LT)] for the different experiments for June–August in 2008–10. The simulations were done in sequence, except that GLD and COMB, and GLDVEG, and COMBVEG were done in parallel; and SOIL is based on CZIL. Bold numbers represent improved simulations, and italic numbers represent degraded simulations compared with the previous experiment. The “Average” is the RMSE or IOA calculated over the whole period from June to August in 2008–10. The RMSE is calculated using  and the IOA with , where  and  are simulated and observed fluxes, respectively, and  is the averaged observed fluxes for each month.
RMSE and index of agreement (IOA) between simulated and observed SH [daytime, 0800–1600 local time (LT)] for the different experiments for June–August in 2008–10. The simulations were done in sequence, except that GLD and COMB, and GLDVEG, and COMBVEG were done in parallel; and SOIL is based on CZIL. Bold numbers represent improved simulations, and italic numbers represent degraded simulations compared with the previous experiment. The “Average” is the RMSE or IOA calculated over the whole period from June to August in 2008–10. The RMSE is calculated using  and the IOA with , where  and  are simulated and observed fluxes, respectively, and  is the averaged observed fluxes for each month.
Fig. 7.

Observed (OBS) and model-simulated monthly-mean diurnal cycle of SH using different precipitation datasets without–with the vegetation (VEG) parameter modifications (left) June, (middle) July, and (right) August of (a)–(c) 2008, (d)–(f) 2009, and (g)–(i) 2010. Bias for experiments CTL, GLDVEG, and COMBVEG, respectively, is also shown in each panel for each month. The time is UTC with hour 4 being 1200 LT.

Fig. 7.

Observed (OBS) and model-simulated monthly-mean diurnal cycle of SH using different precipitation datasets without–with the vegetation (VEG) parameter modifications (left) June, (middle) July, and (right) August of (a)–(c) 2008, (d)–(f) 2009, and (g)–(i) 2010. Bias for experiments CTL, GLDVEG, and COMBVEG, respectively, is also shown in each panel for each month. The time is UTC with hour 4 being 1200 LT.

Fig. 8.

As in Fig. 7, but for LH.

Fig. 8.

As in Fig. 7, but for LH.

Table 4b.

As in Table 4a, but for the LH.

As in Table 4a, but for the LH.
As in Table 4a, but for the LH.

c. Sensitivity of flux simulations to surface exchange coefficient

When compared to the simulated results in COMBVEG using the default Czil (0.1) in Eq. (3), the Czil experiment using the Czilh relationship in Eq. (4) significantly improves SH simulations. It decreases 65 W m−2 of averaged RMSE of SH, more than half of the averaged RMSE for COMBVEG simulations (Table 4a). The index of agreement also increases sharply. The Czilh relationship decreases the biases of SH dramatically and the underestimation after the late afternoon is also mitigated (Fig. 9).

Fig. 9.

As in Fig. 7, but with only for OBS, COMBVEG and CZIL.

Fig. 9.

As in Fig. 7, but with only for OBS, COMBVEG and CZIL.

Using Eq. (4) essentially reduces the surface exchange coefficient with regards to the use of the default Czil value, which is appropriate for semiarid and arid short grasses (Chen and Zhang 2009; Chen et al. 2010). For a given surface net radiation flux over semiarid regions, reducing the surface exchange coefficient implies a reduction of total heat fluxes transferred to the atmosphere but an increase of heat transferred between the surface and soil. Therefore, if the surface exchange coefficient is right, it is expected to see improvements in both surface heat fluxes and soil temperature. As demonstrated in Fig. 10, the biases in soil temperature at the surface layer are reduced. The daytime soil temperature simulation is significantly improved, with the averaged bias of 0.79°C (vs −3.11°C in COMBVEG). The nighttime soil temperature simulation is also improved but is smaller in magnitude, with the averaged bias of −1.77°C (vs −3.86°C in COMBVEG). These results are consistent with those in Zeng et al. (2012) and Zheng et al. (2012), indicating that use of the time-varying Czil as a function of vegetation growth can substantially alter the land–atmospheric surface exchange coefficient. However, the Czilh relationship still overestimates SH systematically, and it slightly deteriorates LH simulations for the whole period (Table 4), which is related to, among other factors, vegetation canopy resistance formulation.

Fig. 10.

As in Fig. 9, but for soil temperature at the depth of 5 cm.

Fig. 10.

As in Fig. 9, but for soil temperature at the depth of 5 cm.

d. Sensitivity of surface heat flux simulations to canopy resistance

The original Jarvis scheme overestimates LH for most months in the CZIL experiment (Fig. 11), consistent with previous studies (Niyogi and Raman 1997; Chen et al. 2007). The modified Jarvis scheme with an exponential soil moisture stress function [Eq. (10)] decreases LH when the precipitation is relatively sufficient but increases LH when there is not enough precipitation (e.g., June 2008 and July and August 2010). The RC simulations reduce biases of LH for 5 months (Fig. 11) and slightly improve the averaged RMSE in SH and LH compared with the CZIL simulations (Table 4). As shown in Table 5, the exponential soil water stress function increases canopy resistance effectively for the relatively wet conditions and decreases it for relatively dry periods (August 2010). Such an increase of canopy resistance limits soil moisture being extracted for transpiration, which consequently reduces plant transpiration and obviously reduces the ratio of transpiration to total evapotranspiration (not shown). However, the increase of soil moisture as a result of less transpiration benefits soil evaporation. It is found that LH overestimation is mainly caused by overestimation of vegetation transpiration in relatively densely vegetated areas (Hong et al. 2009). Therefore, the decrease of transpiration makes the partitioning between soil evaporation and transpiration more reasonable (not shown).

Fig. 11.

As in Fig. 8, but without–with the modified Jarvis scheme and for only OBS, CZIL and RC.

Fig. 11.

As in Fig. 8, but without–with the modified Jarvis scheme and for only OBS, CZIL and RC.

Table 5.

The normalized differences [(Rc − CZIL)/CZIL] in canopy resistance (Rc), Tr, Eg, and LH.

The normalized differences [(Rc − CZIL)/CZIL] in canopy resistance (Rc), Tr, Eg, and LH.
The normalized differences [(Rc − CZIL)/CZIL] in canopy resistance (Rc), Tr, Eg, and LH.

Since the canopy resistance calculations are interactively linked with soil moisture as well as with the atmospheric components, we initialize the RC run with initial conditions in the CZIL experiment after spinup to avoid the effect of initial soil conditions on canopy resistance calculations. The differences are small and the pattern does not change (not shown). Some studies found that the Jarvis scheme could improve the simulations of surface energy flux by a significant tuning of minimum stomatal resistance (Rsmin; Alfieri et al. 2008; Niyogi et al. 2009). We test a higher Rsmin (100 s m−1) as in Alfieri et al. (2008), but the difference is similar (not shown). This makes us believe that the exponential water stress formulation in the Jarvis scheme does not degrade LH simulations and improves the partitioning between soil evaporation and plant transpiration.

e. Sensitivity of flux simulations to uncertainties in SHPs

Based on the experiments COMB, COMBVEG, CZIL, and RC, four sets of 110 SHP tests are conducted separately. Both SH and LH simulations are sensitive to variations in SHPs. The peak values of SH and LH simulations vary by approximately 200 and 150 W m−2, respectively (not shown). The 110 experiments based on the Czilh relationship produce the least RMSE for both SH and LH (Fig. 12c). We select one “optimal” group of SHPs that produces the minimal RMSE of LH as experiment SOIL, that is, the SOIL experiment is based on the CZIL experiment in Table 2. The biases in both SH and LH in the SOIL experiment decrease compared to the CZIL experiment (Figs. 13, 14). In fact, the averaged RMSE of both SH and LH are decreased by ~9 and 4 W m−2, respectively, compared with the CZIL experiment (Table 4). The SOIL experiment with optimal SHPs produces wetter soil and efficiently decreases SH, consistent with the study of Hong et al. (2009), showing that in sparsely vegetated areas SH overestimation is caused by low soil moisture. Regarding LH, almost all the precipitation is provided to evapotranspiration as a result of the dry nature of this semiarid site. Therefore, the precipitation should be stored in the root-zone layer for evaporation rather than for runoff. But in Noah LSM, the lower saturated soil water conductivity for sandy loam cannot maintain water in the root-zone layer. Therefore, it is not surprising that the wetter soil in the SOIL experiment favors LH in the early growing season and then constrains LH in the following period. However, all 110 groups of SHPs still overestimate SH for most months and could not change the tendency of low LH in the early summer, which may be attributed to other factors such as the forcing data error and model structure. This will be discussed further in the concluding remarks.

Fig. 12.

RMSE of simulated SH vs LH (June–August) using different SHPs based on experiments (a) COMB, (b) COMBVEG, (c) CZIL, and (d) RC.

Fig. 12.

RMSE of simulated SH vs LH (June–August) using different SHPs based on experiments (a) COMB, (b) COMBVEG, (c) CZIL, and (d) RC.

Fig. 13.

As in Fig. 7, but for the diurnal cycle of SH and for only OBS, CZIL, and SOIL.

Fig. 13.

As in Fig. 7, but for the diurnal cycle of SH and for only OBS, CZIL, and SOIL.

Fig. 14.

As in Fig. 13, but for LH.

Fig. 14.

As in Fig. 13, but for LH.

Simulations with selected optimal SHPs from 110 sensitivity tests based on COMB substantially decrease the RMSE of SH and LH (Fig. 12a) compared with the original COMB simulations (Table 4). The magnitude of this decrease is even greater than what the COMBVEG simulations produce, indicating that SHPs have a greater effect than vegetation factors on SH and LH simulations. Note that the difference between the two sets of SHP tests based on COMB and COMBVEG is the vegetation parameters and that the difference between the two sets of SHP tests based on COMBVEG and CZIL is the Czil parameterization. Therefore, the decrease in the averaged RMSE of LH is mainly attributed to vegetation parameters (Figs. 12a,b) and the decrease in the averaged RMSE of SH is primarily attributed to the Czil parameterization (Figs. 12b,c). This, to some degree, implies that although flux simulations are more sensitive to SHPs, the effects of vegetation parameter and surface exchange processes are not completely overwhelmed by SHPs.

The exponential soil water stress function slightly increases the RMSE of SH but has no apparent effect on LH (Fig. 12d). Both the wilting point and field capacity for the selected optimal SHPs from these 110 sensitivity tests based on the RC experiment are drier than the default values, but they are consistent with the reported values in the study of Zhang et al. (2005). This may suggest that the exponential soil water stress formulation is more suitable to dry soil conditions.

5. Concluding remarks

The work presented here is intended to improve the ability to represent surface heat exchange in the Noah LSM over a desert steppe site in Inner Mongolia, China. Numerous sensitivity tests with Noah are conducted to investigate the effects of different precipitation input, vegetation properties, surface-exchange-coefficient parameterization, and SHPs on the simulations of surface energy fluxes. The Noah LSM with the default vegetation parameters and SHPs shows a tendency to overestimate both SH and LH in the summertime over this site. After these parameters are modified based on in situ and satellite observations, the model RMSEs are generally reduced and the agreement between modeled results and observations is remarkably improved. This is expected, of course, because the model uses site-specific vegetation conditions that are closer to reality than that from a global lookup table.

Precipitation distributions modulate SH and LH together, and when used together with proper vegetation parameters, a more reasonable precipitation product can improve flux simulations. Changes in precipitation seasonal distribution due to climate change would cause significant consequences in energy flux simulations when the precipitation is relatively ample over this desert steppe site. However, when the site suffers from a drought (as in 2010, for instance), the difference in precipitation distribution has less influence on flux simulations.

Using Czil as a function of canopy height can substantially alter the land–atmospheric surface exchange coefficient. The Czil formulation represents an important aspect of land–atmosphere surface exchange processes, that is, how to parameterize the roughness length (Zot) for thermal and water vapor transfer in LSMs, which is fundamentally different from the roughness length (Zom) used to calculate momentum transfer. Most, if not all, LSMs discern Zot from Zom in their Monin–Obukhov similarity modeling framework, although their specific parameterization for Zot may be different. This study reveals the importance of Zot in modeling the surface exchange process for semiarid sparsely vegetated regions, and similar formulations considering vegetation phenology such as proposed here should be applicable to other LSMs. And, because canopy height is not explicitly represented in Noah, we use the z0 = 0.13h relationship to infer the canopy height of grass. Other vegetation parameters such as LAI, roughness length, emissivity, and albedo are strongly dependent on the seasonal vegetation cycle. They are all scaled by time-varying GVF in Noah. Such time-varying properties could be simulated using additional equations (Liu and Gupta 2007). Dynamic vegetation models have been used successfully to address a lot of issues in phenology and ecology (Scheiter et al. 2013). The research communities are trying to develop a more flexible approach based on how plants with different characteristics perform under given environmental conditions. The Noah community has also been in the development of dynamic vegetation schemes (Niu et al. 2011; Yang et al. 2011).

We use in situ observations to modify the Jarvis scheme. The difference between the modified Jarvis and the original Jarvis scheme is the formulation of soil moisture. As other studies suggested (Matsui et al. 2005), the evaluation of LSMs requires reliable soil and vegetation data. The uncertainties in observed soil moisture and the inconsistency between observed and simulated soil moisture could contribute to the bias between the development of exponential soil water stress function and its application in Noah. The application of this modified Jarvis scheme is appropriate to dry soil conditions like the desert steppe in this study. Whether it is suitable to other land-cover type needs more calibration with different land-cover observations in the future.

Moreover, as demonstrated by other studies (Koster and Milly 1997; Koster et al. 2009; Seneviratne et al. 2010), different LSMs may have different soil moisture dynamic regimes because of various methods in dealing with evaporation and runoff, that is, simulated soil moisture state is a model specific quantity. Generally speaking, the soil moisture dynamics in LSMs are more tuned toward capturing observed surface heat fluxes, and consequently, LSMs do not necessarily reproduce exactly the observed soil moisture. The proper utilization of observations for development and validation of LSMs is still a daunting challenge. From the SHP sensitivity tests, we could select the optimal group of SHPs, which are so-called effective values, but they may not be physically reasonable quantities. For instance, the wilting point is 0.285 m3 m−3, which is obviously higher than the default values and those reported previously (Zhang et al. 2005). This group of SHPs makes soil much moister, and the surface soil moisture is always larger than 0.30 m3 m−3. This indicates that when we calibrate an LSM, we must pay attention to the apparently good modeling results. Since the simulation errors are probably mutually offset by imperfection in parameterization (Liu and Gupta 2007), we should use observed values to set the physical range properly. Note that the model is a conceptual representation of the real system; therefore, this also indicates that if other parts of the model failed to capture the realism of the system, the model would not be able to improve flux simulations even with the correct SHP values. Accurate SHP measurements along with flux observations are important in the future for evaluating and improving LSMs.

This study, while limited in scope, highlights the importance of precipitation distribution, proper model parameter selection, surface heat exchange coefficient parameterization, canopy resistance formulation, and SHPs in semiarid regions, which enhances our understanding of model behavior in different aspects of land surface processes. The overall error in a model can be caused by the uncertainties in parameter and forcing data and model structure in LSMs (Gupta et al. 1998; Liu and Gupta 2007). The uncertainties in flux observations also increase the difficulties in the progress of calibrating LSMs. A more comprehensive analysis should be done to investigate other sources of uncertainties, such as those associated with model structure, other initial conditions, and flux observations (Rosolem et al. 2012).

Acknowledgments

The authors thank Dr. Ethan D. Gutmann for providing SHP data and for his helpful suggestions during the course of this research. This research was jointly supported by State Key Development Program of Basic Research (2010CB951303) and Strategic Priority Research Program–Climate Change: Carbon Budget and Related Issues of the Chinese Academy of Sciences (XDA05050408). We would also like to acknowledge the support from the NCAR Water Cycle Across Scale and BEACHON programs.

REFERENCES

REFERENCES
Alfieri
,
J. G.
,
D.
Niyogi
,
P. D.
Blanken
,
F.
Chen
,
M. A.
LeMone
,
K.
Mitchell
,
M. B.
Ek
, and
A.
Kumar
,
2008
:
Estimation of the minimum canopy resistance for croplands and grasslands using data from the 2002 International H2O Project
.
Mon. Wea. Rev.
,
136
,
4452
4469
,
doi:10.1175/2008MWR2524.1.
Amer
,
K. H.
, and
J. L.
Hatfield
,
2004
:
Canopy resistance as affected by soil and meteorological factors in potato
.
Agron. J.
,
96
,
978
985
,
doi:10.2134/agronj2004.0978.
Barlage
,
M.
, and
X.
Zeng
,
2004
:
The effects of observed fractional vegetation cover on the land surface climatology of the community land model
.
J. Hydrometeor.
,
5
,
823
830
,
doi:10.1175/1525-7541(2004)005<0823:TEOOFV>2.0.CO;2.
Betts
,
A. K.
,
F.
Chen
,
K. E.
Mitchell
, and
Z. I.
Janjic
,
1997
:
Assessment of the land surface and boundary layer models in two operational versions of the NCEP Eta Model using FIFE data
.
Mon. Wea. Rev.
,
125
,
2896
2916
,
doi:10.1175/1520-0493(1997)125<2896:AOTLSA>2.0.CO;2.
Campbell
,
G.
,
1974
:
Simple method for determining unsaturated conductivity from moisture retention data
.
Soil Sci.
,
117
,
311
314
,
doi:10.1097/00010694-197406000-00001.
Chang
,
J. T.
, and
P. J.
Wetzel
,
1991
:
Effects of spatial variations of soil moisture and vegetation on the evolution of a prestorm environment: A numerical case study
.
Mon. Wea. Rev.
,
119
,
1368
1390
,
doi:10.1175/1520-0493(1991)119<1368:EOSVOS>2.0.CO;2.
Charusombat
,
U.
,
D.
Niyogi
,
A.
Kumar
,
X.
Wang
,
F.
Chen
,
A. B.
Guenther
,
A. A.
Turnipseed
, and
K.
Alapaty
,
2010
:
Evaluating a new deposition velocity module in the Noah land surface model
.
Bound.-Layer Meteor.
,
137
,
271
290
,
doi:10.1007/s10546-010-9531-y.
Chen
,
F.
, and
J.
Dudhia
,
2001
:
Coupling an advanced land surface–hydrology model with the Penn State–NCAR MM5 modeling system. Part I: Model implementation and sensitivity
.
Mon. Wea. Rev.
,
129
,
569
585
,
doi:10.1175/1520-0493(2001)129<0569:CAALSH>2.0.CO;2.
Chen
,
F.
, and
Y.
Zhang
,
2009
:
On the coupling strength between the land surface and the atmosphere: From viewpoint of surface exchange coefficients
.
Geophys. Res. Lett.
,
36
,
L10404
,
doi:10.1029/2009GL037980.
Chen
,
F.
, and
Coauthors
,
1996
:
Modeling of land-surface evaporation by four schemes and comparison with FIFE observations
.
J. Geophys. Res.
,
101
(
D3
),
7251
7268
,
doi:10.1029/95JD02165.
Chen
,
F.
,
Z.
Janjić
, and
K.
Mitchell
,
1997
:
Impact of atmospheric surface layer parameterization in the new land-surface scheme of the NCEP Mesoscale Eta numerical model
.
Bound.-Layer Meteor.
,
85
,
391
421
,
doi:10.1023/A:1000531001463.
Chen
,
F.
,
D. N.
Yates
,
H.
Nagai
,
M. A.
LeMone
,
K.
Ikeda
, and
R. L.
Grossman
,
2003
:
Land-surface heterogeneity in the Cooperative Atmosphere Surface Exchange Study (CASES-97). Part-I: Comparing modeled surface flux maps with surface-tower and aircraft measurements
.
J. Hydrometeor.
,
4
,
196
218
,
doi:10.1175/1525-7541(2003)4<196:LSHITC>2.0.CO;2.
Chen
,
F.
, and
Coauthors
,
2007
:
Description and evaluation of the characteristics of the NCAR high-resolution data assimilation system
.
J. Appl. Meteor. Climatol.
,
46
,
694
713
,
doi:10.1175/JAM2463.1.
Chen
,
Y.
,
K.
Yang
,
D.
Zhou
,
J.
Qin
, and
X.
Guo
,
2010
:
Improving the Noah land surface model in arid regions with an appropriate parameterization of the thermal roughness length
.
J. Hydrometeor.
,
11
,
995
1006
,
doi:10.1175/2010JHM1185.1.
Crawford
,
T. M.
,
D. J.
Stensrud
,
F.
Mora
,
J. W.
Merchant
, and
P. J.
Wetzel
,
2001
:
Value of incorporating satellite-derived land cover data in MM5/PLACE for simulating surface temperatures
.
J. Hydrometeor.
,
2
,
453
468
,
doi:10.1175/1525-7541(2001)002<0453:VOISDL>2.0.CO;2.
Ek
,
M. B.
,
K. E.
Mitchell
,
Y.
Lin
,
E.
Rogers
,
P.
Grummann
,
V.
Koren
,
G.
Gayno
, and
J. D.
Tarpley
,
2003
:
Implementation of Noah land surface model advances in the National Centers for Environmental Prediction operational Mesoscale Eta Model
.
J. Geophys. Res.
,
108
,
8851
,
doi:10.1029/2002JD003296.
Gao
,
Y.
,
F.
Chen
,
M.
Barlage
,
W.
Liu
,
Y.
Ran
,
H.
Li
,
H.
Peng
, and
M.
Ma
,
2008
:
Enhancement of land surface information and its impact on atmospheric modeling in the Heihe River basin, northwest China
.
J. Geophys. Res.
,
113
,
D20S90
,
doi:10.1029/2008JD010359.
Godfrey
,
C. M.
, and
D. J.
Stensrud
,
2010
:
An empirical latent heat flux parameterization for the Noah land surface model
.
J. Appl. Meteor. Climatol.
,
49
,
1696
1713
,
doi:10.1175/2010JAMC2180.1.
Gosz
,
J. R.
,
1993
:
Ecotone hierarchies
.
Ecol. Appl.
,
3
,
369
376
.
Guo
,
Q.
,
Z. M.
Hu
,
S. G.
Li
,
X. R.
Li
,
X. M.
Sun
, and
G. R.
Yu
,
2012
:
Spatial variations in aboveground net primary productivity along a climate gradient in Eurasian temperate grassland: Effects of mean annual precipitation and its seasonal distribution
.
Global Change Biol.
,
18
,
3624
3631
,
doi:10.1111/gcb.12010.
Gupta
,
H. V.
,
S.
Sorooshian
, and
P. O.
Yapo
,
1998
:
Toward improved calibration of hydrologic models: Multiple and noncommensurable measures of information
.
Water Resour. Res.
,
34
,
751
763
,
doi:10.1029/97WR03495.
Gutman
,
G.
, and
A.
Ignatov
,
1998
:
The derivation of the green vegetation fraction from NOAA/AVHRR for use in numerical weather prediction models
.
Int. J. Remote Sens.
,
19
,
1533
1543
,
doi:10.1080/014311698215333.
Gutmann
,
E. D.
, and
E. E.
Small
,
2007
:
A comparison of land surface model soil hydraulic properties estimated by inverse modeling and pedotransfer functions
.
Water Resour. Res.
,
43
,
W05418
,
doi:10.1029/2006WR005135.
Heisler-White
,
J. L.
,
A. K.
Knapp
, and
E. F.
Kelly
,
2008
:
Increasing precipitation event size increases aboveground net primary productivity in a semi-arid grassland
.
Oecologia
,
158
,
129
140
,
doi:10.1007/s00442-008-1116-9.
Hong
,
S.
,
V.
Lakshmi
, and
E. E.
Small
,
2007
:
Relationship between vegetation biophysical properties and surface temperature using multi-sensor satellite data
.
J. Climate
,
20
,
5593
5606
,
doi:10.1175/2007JCLI1294.1.
Hong
,
S.
,
V.
Lakshmi
,
E. E.
Small
,
F.
Chen
,
M.
Tewari
, and
K. W.
Manning
,
2009
:
Effects of vegetation and soil moisture on the simulated land surface processes from the coupled WRF/Noah model
.
J. Geophys. Res.
,
114
,
D18118
,
doi:10.1029/2008JD011249.
Humphrey
,
R. R.
,
1958
:
The desert grassland. A history of vegetational change and an analysis of causes
.
Bot. Rev.
,
24
,
193
252
,
doi:10.1007/BF02872568.
Jacquemin
,
B.
, and
J.
Noilhan
,
1990
:
Sensitivity study and validation of a land surface parameterization using the HAPEXMOBILHY dataset
.
Bound.-Layer Meteor.
,
52
,
93
134
,
doi:10.1007/BF00123180.
Kato
,
H.
,
M.
Rodell
,
F.
Beyrich
,
H.
Cleugh
,
E.
van Gorsel
,
H.
Liu
, and
T. P.
Meyers
,
2007
:
Sensitivity of land surface simulations to model physics, land characteristics, and forcings, at four CEOP Sites
.
J. Meteor. Soc. Japan
,
85A
,
187
204
,
doi:10.2151/jmsj.85A.187.
Knapp
,
A. K.
, and
M. D.
Smith
,
2001
:
Variation among biomes in temporal dynamics of aboveground primary production
.
Science
,
291
,
481
484
,
doi:10.1126/science.291.5503.481.
Koren
,
V.
,
J.
Schaake
,
K.
Mitchell
,
Q. Y.
Duan
, and
F.
Chen
,
1999
:
A parameterization of snowpack and frozen ground intended for NCEP weather and climate models
.
J. Geophys. Res.
,
104
,
19 569
19 585
,
doi:10.1029/1999JD900232.
Koster
,
R. D.
, and
P. C. D.
Milly
,
1997
:
The interplay between transpiration and runoff formulations in land surface schemes used with atmospheric models
.
J. Climate
,
10
,
1578
1591
,
doi:10.1175/1520-0442(1997)010<1578:TIBTAR>2.0.CO;2.
Koster
,
R. D.
, and
Coauthors
,
2004
:
Regions of strong coupling between soil moisture and precipitation
.
Science
,
305
,
1138
1140
,
doi:10.1126/science.1100217.
Koster
,
R. D.
,
Z.
Guo
,
R.
Yang
,
P. A.
Dirmeyer
,
K.
Mitchell
, and
M. J.
Puma
,
2009
:
On the nature of soil moisture in land surface models
.
J. Climate
,
22
,
4322
4335
,
doi:10.1175/2009JCLI2832.1.
Kumar
,
A.
,
F.
Chen
,
D.
Niyogi
,
J. G.
Alfieri
,
K.
Mitchell
, and
M.
Ek
,
2011
:
Evaluation of a photosynthesis-based canopy resistance formulation in the Noah land surface model
.
Bound.-Layer Meteor.
,
138
,
263
284
,
doi:10.1007/s10546-010-9559-z.
Kurkowski
,
N. P.
,
D. J.
Stensrud
, and
M. E.
Baldwin
,
2003
:
Assessment of implementing satellite-derived land cover data in the Eta model
.
Wea. Forecasting
,
18
,
404
416
,
doi:10.1175/1520-0434(2003)18<404:AOISDL>2.0.CO;2.
Kustas
,
W. P.
,
J. H.
Prueger
,
K. S.
Humes
, and
P. J.
Starks
,
1999
:
Estimation of surface heat fluxes at field scale using surface layer versus mixed-layer atmospheric variables with radiometric temperature observations
.
J. Appl. Meteor.
,
38
,
224
238
,
doi:10.1175/1520-0450(1999)038<0224:EOSHFA>2.0.CO;2.
Leese
,
J.
,
T.
Jackson
,
A.
Pitman
, and
P.
Dirmeyer
,
2001
:
GEWEX/BAHC International Workshop on Soil Moisture Monitoring, Analysis, and Prediction for Hydrometeorological and Hydroclimatological Applications
.
Bull. Amer. Meteor. Soc.
,
82
,
1423
1430
,
doi:10.1175/1520-0477(2001)082<1423:MSGBIW>2.3.CO;2.
LeMone
,
M. A.
,
F.
Chen
,
J. G.
Alfieri
,
M.
Tewari
,
B.
Geerts
,
Q.
Miao
,
R. L.
Grossman
, and
R. L.
Coulter
,
2007
:
Influence of land cover and soil moisture on the horizontal distribution of sensible and latent heat fluxes in southeast Kansas during IHOP_2002 and CASES-97
.
J. Hydrometeor.
,
8
,
68
87
,
doi:10.1175/JHM554.1.
LeMone
,
M. A.
,
M.
Tewari
,
F.
Chen
,
J. G.
Alfieri
, and
D.
Niyogi
,
2008
:
Evaluation of the Noah land-surface model using data from a fair-weather IHOP_2002 day with heterogeneous surface fluxes
.
Mon. Wea. Rev.
,
136
,
4915
4941
,
doi:10.1175/2008MWR2354.1.
Li
,
X.-R.
,
2001
:
Study on shrub community diversity of Ordos Plateau, Inner Mongolia, northern China
.
J. Arid Environ.
,
47
(
3
),
271
279
,
doi:10.1006/jare.2000.0707.
Liao
,
G.
, and
Y.
Jia
,
1996
: Grassland Resource of China (in Chinese). Chinese Science and Technology Publisher, 608 pp.
Liu
,
Y.
, and
H. V.
Gupta
,
2007
:
Uncertainty in hydrologic modeling: Toward an integrated data assimilation framework
.
Water Resour. Res.
,
43
,
W07401
, doi:10.1029/2006WR005756.
Ma
,
Y.
,
S.
Fan
,
H.
Ishikawa
,
O.
Tsukamoto
,
T.
Yao
,
T.
Koike
,
H.
Zuo
,
H.
Hu
, and
Z.
Su
,
2005
:
Diurnal and inter-monthly variation of land surface heat fluxes over the central Tibetan Plateau area
.
Theor. Appl. Climatol.
,
80
,
259
273
,
doi:10.1007/s00704-004-0104-1.
Mahrt
,
L.
, and
K.
Ek
,
1984
:
The influence of atmospheric stability on potential evaporation
.
J. Climate Appl. Meteor.
,
23
,
222
234
,
doi:10.1175/1520-0450(1984)023<0222:TIOASO>2.0.CO;2.
Matsui
,
T.
,
V.
Lakshmi
, and
E. E.
Small
,
2005
:
The effects of satellite-derived vegetation cover variability on simulated land–atmosphere interactions in the NAMS
.
J. Climate
,
18
,
21
40
,
doi:10.1175/JCLI3254.1.
Mielnick
,
P.
,
W. A.
Dugas
,
K.
Mitchell
, and
K.
Havstad
,
2005
:
Long-term measurements of CO2 flux and evapotranspiration in a Chihuahuan desert grassland
.
J. Arid Environ.
,
60
,
423
436
,
doi:10.1016/j.jaridenv.2004.06.001.
Miller
,
J.
,
M.
Barlage
,
X.
Zeng
,
H.
Wei
,
K.
Mitchell
, and
D.
Tarpley
,
2006
:
Sensitivity of the NCEP/Noah land surface model to the MODIS green vegetation fraction dataset
.
Geophys. Res. Lett.
,
33
,
L13404
,
doi:10.1029/2006GL026636.
Mitchell
,
K.
, and
Coauthors
,
2004
:
The multi-institution North American Land Data Assimilation System (NLDAS): Utilizing multiple GCIP products and partners in a continental distributed hydrological modeling system
.
J. Geophys. Res.
,
109
,
D07S90
,
doi:10.1029/2003JD003823.
Monteith
,
J. L.
,
1965
:
Evaporation and environment
.
Symp. Soc. Exp. Biol.
,
19
,
205
234
.
Monteith
,
J. L.
, and
M. H.
Unsworth
,
1990
: Principles of Environmental Physics. 2nd ed. Edward Arnold, 291 pp.
Neilson
,
R.
,
1986
:
High-resolution climatic analysis and southwest biogeography
.
Science
,
232
,
27
34
,
doi:10.1126/science.232.4746.27.
Niu
,
G. Y.
, and
Coauthors
,
2011
:
The Community Noah land surface model with multi-parameterization options (Noah-MP): 1. Model description and evaluation with local-scale measurements
.
J. Geophys. Res.
,
116
, D12109,
doi:10.1029/2010JD015139.
Niyogi
,
D. S.
, and
S.
Raman
,
1997
:
Comparison of four different stomatal resistance schemes using FIFE observations
.
J. Appl. Meteor.
,
36
,
903
917
,
doi:10.1175/1520-0450(1997)036<0903:COFDSR>2.0.CO;2.
Niyogi
,
D. S.
,
K.
Alapaty
,
S.
Raman
, and
F.
Chen
,
2009
:
Development and evaluation of a coupled photosynthesis-based gas exchange evapotranspiration model (GEM) for mesoscale weather forecasting applications
.
J. Appl. Meteor. Climatol.
,
48
,
349
368
,
doi:10.1175/2008JAMC1662.1.
Pan
,
H. L.
, and
L.
Mahrt
,
1987
:
Interaction between soil hydrology and boundary-layer development
.
Bound.-Layer Meteor.
,
38
,
185
202
,
doi:10.1007/BF00121563.
Papale
,
D.
, and
Coauthors
,
2006
:
Towards a standardized processing of net ecosystem exchange measured with eddy covariance technique: Algorithms and uncertainty estimation
.
Biogeosciences
,
3
,
571
583
,
doi:10.5194/bg-3-571-2006.
Qu
,
W.
, and
Coauthors
,
1998
:
Sensitivity of latent heat flux from PILPS land-surface schemes to perturbations of surface air temperature
.
J. Atmos. Sci.
,
55
,
1909
1927
,
doi:10.1175/1520-0469(1998)055<1909:SOLHFF>2.0.CO;2.
Robock
,
A.
,
K. Y.
Vinnikov
,
G.
Srinivasan
,
J. K.
Entin
,
S. E.
Hollinger
,
N. A.
Speranskaya
,
S.
Liu
, and
A.
Namkhai
,
2000
:
The global soil moisture data bank
.
Bull. Amer. Meteor. Soc.
,
81
,
1281
1299
,
doi:10.1175/1520-0477(2000)081<1281:TGSMDB>2.3.CO;2.
Robock
,
A.
, and
Coauthors
,
2003
: Evaluation of the North American Land Data Assimilation System over the southern Great Plains during the warm season. J. Geophys. Res.,108, 8846, doi:10.1029/2002JD003245.
Rodell
,
M.
, and
Coauthors
,
2004
:
The Global Land Data Assimilation System
.
Bull. Amer. Meteor. Soc.
,
85
,
381
394
,
doi:10.1175/BAMS-85-3-381.
Rosolem
,
R.
,
H. V.
Gupta
,
W. J.
Shuttleworth
,
L. G. G.
Gonçalves
, and
X.
Zeng
,
2012
:
Towards a comprehensive approach to parameter estimation in land surface parameterization schemes
.
Hydrol. Processes
, 27, 2075–2097, doi:10.1002/hyp.9362.
Santanello
,
J. A.
, and
T. N.
Carlson
,
2001
:
Mesoscale simulation of rapid soil drying and its implications for predicting daytime temperature
.
J. Hydrometeor.
,
2
,
71
88
,
doi:10.1175/1525-7541(2001)002<0071:MSORSD>2.0.CO;2.
Schaake
,
J. C.
,
V. I.
Koren
,
Q. Y.
Duan
,
K.
Mitchell
, and
F.
Chen
,
1996
:
A simple water balance model (SWB) for estimating runoff at different spatial and temporal scales
.
J. Geophys. Res.
,
101
(
D3
),
7461
7475
,
doi:10.1029/95JD02892.
Schaap
,
M.
, and
F.
Leij
,
1998
:
Database-related accuracy and uncertainty of pedotransfer functions
.
Soil Sci.
,
163
,
765
779
,
doi:10.1097/00010694-199810000-00001.
Scheiter
,
S.
,
L.
Langan
, and
S. I.
Higgins
,
2013
:
Next-generation dynamic global vegetation models: Learning from community ecology
.
New Phytol.
, 198, 957–967, doi:10.1111/nph.12210.
Seneviratne
,
S. I.
,
T.
Corti
,
E. L.
Davin
,
M.
Hirschi
,
E. B.
Jaeger
,
I.
Lehner
,
B.
Orlowsky
, and
A. J.
Teuling
,
2010
:
Investigating soil moisture–climate interactions in a changing climate: A review
.
Earth Sci. Rev.
,
99
,
125
161
,
doi:10.1016/j.earscirev.2010.02.004.
Sridhar
,
V.
,
R. L.
Elliott
,
F.
Chen
, and
J. A.
Brotzge
,
2002
:
Validation of the NOAH/OSU land surface model using surface flux measurements in Oklahoma
.
J. Geophys. Res.
,
107
,
4418
,
doi:10.1029/2001JD001306.
Twine
,
T. E.
, and
Coauthors
,
2000
:
Correcting eddy-covariance flux underestimates over a grassland
.
Agric. For. Meteor.
,
103
,
279
300
,
doi:10.1016/S0168-1923(00)00123-4.
van der Velde
,
R.
,
Z.
Su
,
M.
Ek
,
M.
Rodell
, and
Y.
Ma
,
2009
:
Influence of thermodynamic soil and vegetation parameterizations on the simulation of soil temperature states and surface fluxes by the Noah LSM over a Tibetan plateau site
.
Hydrol. Earth Syst. Sci.
,
13
,
759
777
,
doi:10.5194/hess-13-759-2009.
van Genuchten
,
M.
,
1980
:
A closed-form equation for prediction of the hydraulic properties of unsaturated soils
.
Soil Sci. Soc. Amer. J.
,
44
,
892
898
,
doi:10.2136/sssaj1980.03615995004400050002x.
Webb
,
E. K.
,
G. I.
Pearman
, and
R.
Leuning
,
1980
:
Correction of flux measurements for density effects due to heat and water vapor transfer
.
Quart. J. Roy. Meteor. Soc.
,
106
,
85
100
,
doi:10.1002/qj.49710644707.
Wei
,
M. Y.
, Ed.,
1995
: Soil moisture: Report of a workshop held in Tiburon, California, 25–27 January 1994. NASA Conf. Publ. 3319, 77 pp.
Wetzel
,
P.
, and
J. T.
Chang
,
1987
:
Concerning the relationship between evapotranspiration and soil moisture
.
J. Climate Appl. Meteor.
,
26
,
18
27
,
doi:10.1175/1520-0450(1987)026<0018:CTRBEA>2.0.CO;2.
Wilczak
,
J. M.
,
S. P.
Oncley
, and
S. A.
Stage
,
2001
:
Sonic anemometers tilt correction algorithms
.
Bound.-Layer Meteor.
,
99
,
127
150
,
doi:10.1023/A:1018966204465.
Wilson
,
K.
, and
Coauthors
,
2002
:
Energy balance closure at FLUXNET sites
.
Agric. For. Meteor.
,
113
,
223
243
,
doi:10.1016/S0168-1923(02)00109-0.
Wu
,
Z.
,
X.
Wang
,
F.
Chen
,
A.
Turnipseed
,
A. B.
Guenther
,
D.
Niyogi
,
U.
Charusombat
, and
B.
Xia
,
2011
:
Evaluating the calculated dry deposition velocities of reactive nitrogen oxides and ozone from two community models over a temperate deciduous forest
.
Atmos. Environ.
,
45
,
2663
2674
,
doi:10.1016/j.atmosenv.2011.02.063.
Wu
,
Z.
, and
Coauthors
,
2012
:
Evaluation and improvements of two community models in simulating dry deposition velocities for peroxyacetyl nitrate (PAN) over a coniferous forest
.
J. Geophys. Res.
,
117
,
D04310
,
doi:10.1029/2011JD016751.
Yang
,
F.
, and
G. S.
Zhou
,
2011
:
Characteristics and modeling of evapotranspiration over a temperate desert steppe in Inner Mongolia, China
.
J. Hydrol.
,
396
,
139
147
,
doi:10.1016/j.jhydrol.2010.11.001.
Yang
,
K.
,
T.
Koike
, and
D.
Yang
,
2003
:
Surface flux parameterization in the Tibetan Plateau
.
Bound.-Layer Meteor.
,
106
,
245
262
,
doi:10.1023/A:1021152407334.
Yang
,
K.
, and
Coauthors
,
2008
:
Turbulent flux transfer over bare-soil surfaces: Characteristics and parameterizations
.
J. Appl. Meteor. Climatol.
,
47
,
276
290
,
doi:10.1175/2007JAMC1547.1.
Yang
,
Z. L.
,
G. Y.
Niu
,
K. E.
Mitchell
, and
Coauthors
,
2011
: The community Noah land surface model with multiparameterization options (Noah-MP): 2. Evaluation over global river basins. J. Geophys. Res.,116, D12110, doi:10.1029/2010JD015140.
Yu
,
F.
,
K. P.
Price
,
J.
Ellis
, and
P.
Shi
,
2003
:
Response of seasonal vegetation development to climatic variations in eastern central Asia
.
Remote Sens. Environ.
,
87
,
42
54
,
doi:10.1016/S0034-4257(03)00144-5.
Yucel
,
I.
,
W. J.
Shuttleworth
,
J.
Washburne
, and
F.
Chen
,
1998
:
Evaluating NCEP Eta model derived data against observations
.
Mon. Wea. Rev.
,
126
,
1977
1991
,
doi:10.1175/1520-0493(1998)126<1977:ENEMDD>2.0.CO;2.
Zeng
,
X.
,
Z.
Wang
, and
A.
Wang
,
2012
:
Surface skin temperature and the interplay between sensible and ground heat fluxes over arid regions
.
J. Hydrometeor.
,
13
,
1359
1370
,
doi:10.1175/JHM-D-11-0117.1.
Zhang
,
F.
,
G. S.
Zhou
,
Y.
Wang
,
F. L.
Yang
, and
C.
Nilsson
,
2012
:
Evapotranspiration and crop coefficient for a temperate desert steppe ecosystem using eddy covariance in Inner Mongolia, China
.
Hydrol. Processes
,
26
,
379
386
,
doi:10.1002/hyp.8136.
Zhang
,
G.
,
G. S.
Zhou
,
F.
Chen
, and
Y.
Wang
,
2014
:
Variability of canopy resistance over a desert steppe site in Inner Mongolia, China
.
Adv. Atmos. Sci.
, in press.
Zhang
,
X.
,
1992
: Northern China. Grasslands and Grassland Sciences in Northern China, J. Ellis, Ed., National Academy Press, 33–54.
Zhang
,
Y.
,
E.
Munkhtsetseg
,
T.
Kadota
, and
T.
Ohata
,
2005
:
An observational study of ecohydrology of a sparse grassland at the edge of the Eurasian cryosphere in Mongolia
.
J. Geophys. Res.
,
110
,
D14103
,
doi:10.1029/2004JD005474.
Zheng
,
W.
,
H.
Wei
,
Z.
Wang
,
X.
Zeng
,
J.
Meng
,
M.
Ek
,
K.
Mitchell
, and
J.
Derber
,
2012
: Improvement of daytime land surface skin temperature over arid regions in the NCEP GFS model and its impact on satellite data assimilation. J. Geophys. Res.,117, D06117, doi:10.1029/2011JD015901.
Zilitinkevich
,
S. S.
,
1995
: Non-local turbulent transport: Pollution dispersion aspects of coherent structure of convective flows. Air Pollution Theory and Simulations, H. Power, N. Moussiopoulos, and C. A. Brebbia, Eds., Vol. I, Air Pollution III, Computational Mechanics Publications, 53–60.