Daeun KIM,Yoon-Jin LIM,Minseok KANG,and Minha CHOI

1Department of Civil and Environmental Engineering,Hanyang University,222 Wangsimni-ro,Seongdong-gu, Seoul 04763,Republic of Korea

2Applied Meteorological Research Division,National Institute of Meteorological Research,Korean Meteorological Administration, 61 Yeouidaebang-ro,16-gil,Dongjak-gu,Seoul 63568,Republic of Korea

3National Center for AgroMeteorology,Bld.#36(RM.#109),Seoul National University,1 Gwanak-ro, Gwanak-gu,Seoul 08826,Republic of Korea

4Department of Water Resources,Graduate School of Water Resources,Sungkyunkwan University,2066 Seobu-ro, Jangan-gu,Suwon,Gyeonggi-do 16419,Republic of Korea

Land Response to Atmosphere at Different Resolutions in the Common Land Model over East Asia

Daeun KIM1,Yoon-Jin LIM2,Minseok KANG3,and Minha CHOI∗4

1Department of Civil and Environmental Engineering,Hanyang University,222 Wangsimni-ro,Seongdong-gu, Seoul 04763,Republic of Korea

2Applied Meteorological Research Division,National Institute of Meteorological Research,Korean Meteorological Administration, 61 Yeouidaebang-ro,16-gil,Dongjak-gu,Seoul 63568,Republic of Korea

3National Center for AgroMeteorology,Bld.#36(RM.#109),Seoul National University,1 Gwanak-ro, Gwanak-gu,Seoul 08826,Republic of Korea

4Department of Water Resources,Graduate School of Water Resources,Sungkyunkwan University,2066 Seobu-ro, Jangan-gu,Suwon,Gyeonggi-do 16419,Republic of Korea

Towards a better understanding of hydrological interactions between the land surface and atmosphere,land surface models are routinely used to simulate hydro-meteorological fluxes.However,there is a lack of observations available for model forcing,to estimate the hydro-meteorological fluxes in East Asia.In this study,Common Land Model(CLM)was used in offline-mode during the summer monsoon period of 2006 in East Asia,with different forcings from Asiaflux,Korea Land Data Assimilation System(KLDAS),and Global Land Data Assimilation System(GLDAS),at point and regional scales, separately.The CLM results were compared with observations from Asiaflux sites.The estimated net radiation showed good agreement,with r=0.99 for the point scale and 0.85 for the regional scale.The estimated sensible and latent heat fluxes using Asiaflux and KLDAS data indicated reasonable agreement,with r=0.70.The estimated soil moisture and soil temperature showed similar patterns to observations,although the estimated water fluxes using KLDAS showed larger discrepancies than those of Asiaflux because of scale mismatch.The spatial distribution of hydro-meteorological fluxes according to KLDAS for East Asia were compared to the CLM results with GLDAS,and the GLDAS provided online.The spatial distributions of CLM with KLDAS were analogous to CLM with GLDAS,and the standalone GLDAS data.The results indicate that KLDAS is a good potential source of high spatial resolution forcing data.Therefore,the KLDAS is a promising alternative product, capable of compensating for the lack of observations and low resolution grid data for East Asia.

Common Land Model,Korea Land Data Assimilation System,Global Land Data Assimilation System,Asiaflux,hydro-meteorological fluxes

1.Introduction

Recently,the exchange of energy and water fluxes between the land surface and the atmosphere has been recognized as a crucial factor in understanding the hydrological cycle(Betts et al.,2000;Cox et al.,2000;Pielke,2001; Friedlingstein et al.,2003;Seneviratne and St¨ockli,2006; Betts et al.,2007;St¨ockli et al.,2008;Lei et al.,2011).Previous studies have attempted to estimate land–atmosphere interactions using climate models in accordance with more accurate quantification of heat and water fluxes(Chen et al.,1996;Sellers et al.,1996;Dai et al.,2003;Lim et al.,2012).

Previous studies have considered Land Surface Models(LSMs)coupled with other models to compute hydrometeorological fluxes with high accuracy(Rihani et al., 2010).In addition,LSMs have greatly progressed in terms of calibration procedures,parameter optimization,and adoption of other LSMs(Dai et al.,2003;Choi et al.,2010).These models contain several schemes,including:dynamic vegetation(Dickinson et al.,1998);a transfer system for water, radiative and turbulent transfer;and a physiological system for stomatal processes.Previous researches on LSMs have incorporated complex lateral heterogeneity in addition to the significant advances in computational complexity(Famiglietti and Wood,1994)and increased understanding of physio-logical vegetation processes(Wang and Jarvis,1990).

©Institute of Atmospheric Physics/Chinese Academy of Sciences,and Science Press and Springer-Verlag Berlin Heidelberg 2016

The Common Land Model(CLM)has been widely used to interpret relationships between the land surface and atmosphere,considering the effects of vegetation(Arora,2002). The CLM can reproduce the complicated energy and water transportation patterns and calculate realistic results using relatively few user-defined parameters.It can also produce accurate results for fluxes based on the consideration of various vegetation and snow effects.This model is continually being developed by coupling it with other models,modifying its schemes,and increasing the level of model verification.

For instance,Whitfield et al.(2006)applied the CLM and another model(the Land Surface Process model)to a marsh wetland in the southeastern United States over a dry-down period.The CLM results showed good agreement,with low values of mean absolute error and root-mean-square error (RMSE):less than 0.33 m3m-3for volumetric water content,1.40 K for temperature,and 61.7 W m-2for surface heat fluxes.In addition,soil moisture and snow water storage were simulated using CLM which is able to estimate terrestrial water storage such as interception,runoff,and frozen soil processes.These results were compared with those of the Gravity Recovery and Climate Experiment(Niu and Yang, 2006).Abramowitz et al.(2008)evaluated CLM,the Community Atmosphere Biosphere Land Exchange model,and the Organizing Carbon and Hydrology in Dynamic Ecosystems method.Their results showed that no single model achievedthebestresultsforallofthemeasurementsatsixdifferent sites;therefore,they suggested a calibration of the few parameters for better results.Huang et al.(2008)and Meng et al.(2009)adjusted the data assimilation algorithm to improveevapotranspirationestimationscomparedwithprevious studies.Li et al.(2012)modified the scheme of soil thermal conductivity in CLM to more accurately estimate the energy and water fluxes,such as soil temperature,net radiation,latent heat flux,and sensible heat flux,using land surface process field experiment data for bare soil conditions on a loess plateau.Choi et al.(2010)obtained reliable energy and water fluxesusingCLMwithKoreafluxnetworkdataforfarmlands during the growing season.Likewise,CLM has been analyzed for model developments,scheme improvements,and model applications.

In spite of the large amount of work carried out using CLM,it has not been widely used or validated for data from East Asia,due to a lack of observations.To address this,the current study utilized spatiotemporal patterns of energy and water fluxes that were estimated using CLM with different forcings,from Asiaflux and the Korea Land Data Assimilation System(KLDAS).The temporal variations in energy and water fluxes were validated using the observations from Asiaflux.The spatial distributions of the energy and water fluxes from CLM using KLDAS data were compared with those from the results of the Global Land Data Assimilation System(GLDAS)forcing,as well as standalone GLDAS data. Our research was performed to evaluate the applicability of CLM and the ability of KLDAS to act as a supplement to the current data limitations of ground-based measurement in East Asia.

2.Common Land Model

CLM was developed at the National Center for Atmospheric Research from the Project for Intercomparison of Land-surface Parameterization Schemes to determine land surface fluxes(Whitfield et al.,2006).Its development is generally recognized as a combination of three older models:the LSM designed by Kiehl et al.(1996),the Biosphere Atmosphere Transfer Scheme(Dickinson et al.,1993),and a 1994 version of the Chinese Academy of Sciences’Institute of Atmospheric Physics LSM(Dai and Zeng,1997).CLM is the most advanced type of land surface model,integrating the strengths of the aforementioned land surface models into one system(Li et al.,2012).This combination improves the performance of the previous methods and facilitates practical results using relatively simple user-defined parameters.

The model characteristics are as follows:(1)10 soil layersareusedtocalculatethesoilmoistureandtemperature;(2) uptofivesnowlayersareusedtoforecastice,snow,andsnow melt(Daietal.,2001);(3)thefreezingofsoilmoistureisconsidered;(4)a mosaic method for calculating the energy and water balance of each tile by means of a sub-grid is included (Koster and Suarez,1992);(5)the calculation of the degree of ground saturation and runoff is based on the Topographybased Hydrological Model(Beven and Kirkby,1979);(6)18 International Geosphere–Biosphere Programme(IGBP)classifications for land cover types are used;and(7)one layer is considered for photosynthesis-conductance.In this study, CLM version 2.1 was used to produce the various fluxes.

The fundamental equation of CLM is a physical governing equation that includes the water conservation equation to represents mass conservation of water phase,Eq.(1),and the energy conservation equation which represents the conservation of energy according to time rate of change in stored energy,Eq.(2)(Dai et al.,2003):

Here,V is the control volume(m-3),T is temperature(K), ρkis the intrinsic density of constituent k(kg m-3),θkis the partial volume of constituent k(m3m-3),hkis the specific enthalpy(J kg-1),Ukis mass flux(kg m-2s-1),Mk′k is phase change(kg m-3s-1),δk′kis the Kronecker delta, Skis a source or sink term,λis the thermal conductivity of the medium(W m-1K-1),and R is radiation flux(W m-2). Equation(1)shows the rate of mass change represented by the sum of mass flow,phase change,and sources or sinks. Equation(2)representstherateofchangeinstoredheatasthesum of conduction,radiation,and negative convection(Dai et al.,2001).

3.Validation sites and data

3.1.Asiaflux

In this study,the data obtained from flux networks were used to create a forcing dataset and validate the output of the model.Asiaflux is a regional network of micrometeorological tower sites that measure the exchange of carbon dioxide, water vapor,and energy between terrestrial ecosystems and the atmosphere(Baldocchi et al.,2001).The selected validation sites registered in Asiaflux are Gwangneung(GDK), Haenam(HFK),and Seto(SMF),located in southern Korea and Japan(Fig.1).This study used the GDK and HFK measurements for the same time period(2006),as they had relatively good data quality.Asiaflux uses eddy covariance techniques to measure the surface energy within the atmospheric boundary layer(Hong et al.,2003;Matumoto et al., 2008).Quality control procedures based on micrometeorological theories and statistical tests(e.g.,coordinate rotation, Webb–Pearman–Leuning correction,control of ranges and spikes,and quality flagging)were applied to improve the data quality(Kwon et al.,2009).The planar fit rotation method was used to compensate for the limitations of double rotations and triple rotations for tilt correction(Wilczak et al., 2001).When applied to flux sites,these methods can produce a stable coordinate frame to minimize over-rotation.

The characteristics of each site are described in Table 1. Observations were recorded every 30 minutes.The HFK and GDK sites are located in South Korea at(34◦32′N,126◦34′E) and(37◦45′N,127◦09′E),respectively.The annual mean precipitation at HFK is 1306 mm,with rainfall occurring predominantly in the summer months(Chun et al.,2010).The mean annual air temperature is 13.3◦C,and the land cover is seasonal cropland for rice paddies,soybeans,and sweet potatoes(Choi et al.,2010).The soil type is classified as silt loam to loam,and the terrain type is relatively flat,except in the southeast where there is a slope of approximately 4◦.The annual mean precipitation at GDK is 1132 mm,with rainfall occurring mainly in the summer months.The mean annual air temperature is 11.5◦C,the vegetation type is mixed deciduous/coniferous forest,and the soil type is sandy loam.The SMF site(35◦15′N,137◦04′E)has a complex terrain type and mixed forest vegetation.The mean annual air temperature is 15.1◦C,with a mean annual precipitation of 1615 mm.The soil type is classed as brown forest soil.The instruments for flux measurement are described in Table 2.The fluxes were measured at a height of 20.8 m at HFK,40.7 m at GDK,and 19.0 m at SMF.The study period was 23 June to 9 July 2006.

3.2.GLDAS

Fig.1.The three validation flux sites in the study area(http://www.AsiaFlux.net/;http://www.sanlim.kr/).

Table 2.Flux measurement instruments(http://www.asiaflux.net/).

GLDAS is an assimilation system that combines satellite data and observed data using advanced modeling systems,such as Mosaic,CLM,and the Noah LSM.GLDAS is a global,high-resolution,and offline modeling system that integrates satellite and ground-based observations to produce optimized surface states and fluxes(Rodell et al.,2004).The spatial resolutions of GLDAS data are 0.25◦and 1.0◦.The atmospheric forcing data in GLDAS are based on outputs from the atmospheric data assimilation systems of weather forecasts or analysis systems,including NASA’s Goddard Earth Observing System data assimilation system,the European Centre for Medium-Range Weather Forecasting,and the Air Force Weather Agency Agricultural meteorology modeling system.The land surface data came from the University of Maryland vegetation classification,the Boston University Leaf Area Index(derived from Advanced Very High Resolution Radiometer measurements),and Moderate Resolution Imaging Spectroradiometer(MODIS)data.The elevation database was compiled from a global 30-arc-s elevation topographic dataset.GLDAS data have been validated and are widely used as inputs and validation data in various research. In this study,the GLDAS data using the Noah LSM(0.25◦resolution and 3 h intervals)were used to compare the spatial distributions between KLDAS and GLDAS.We selected the GLDAS data from the Noah LSM because it has the highest resolution and a relatively short time interval.

3.3.KLDAS

KLDAS is an offline modeling system developed in order to generate realistic and consistent surface variables,specifically for the region of East Asia(Lim et al.,2010,2012).The purpose of the KLDAS is to provide optimum land surface fluxes,landsurfaceparameters,andinitialsurfaceconditions, such as soil moisture and soil temperature,as an alternative to in-situ data.In particular,the main goal of this system is to estimate coherent values by reducing the uncertainties in current land models,which use satellite and observation data.The KLDAS uses the same basic concepts as land data assimilation projects.In the KLDAS,the Noah LSM is derived using Global Data Assimilation and Prediction System analysis data from the Korean Meteorological Administration,cumulative rainfall data from the Global Telecommunication System,and satellite data based on downward solar radiation.In addition,the Noah LSM uses MODIS land products to determine land surface characteristics.The domain of KLDAS covers East Asia between 10◦N and 50◦N,and 110◦E and 155◦E.The KLDAS project provides hourly near-surface meteorological data,which are available to force LSMs and near real-time offline Noah LSM soil fields and flux data,with a grid spacing of 10 km.The KLDAS provides temperature, surface pressure,relative humidity,wind speed,and downward solar radiation.In East Asia,KLDAS can contribute to analyzingthewaterbudgetbyprovidinghigh-qualitydatasets in various fields.Table 3 presents the time intervals and other information on the KLDAS forcing data sources.

3.4.Forcing data

The temporal patterns of the forcing data(Asiaflux and KLDAS)for CLM at each study site were examined(figure not shown).The forcing data for CLM comprised incoming solar radiation,incoming infrared radiation,precipitation,air temperature,wind speed,atmospheric pressure,and specific humidity.The range of incoming solar radiation was approximately 0–1000 W m-2;however,the variation in incoming infraredradiationdidnotshowlargefluctuations(300–500W m-2).The patterns of increase and decrease of the incoming solar radiation between the Asiaflux and KLDAS data were similar,except when it rained.This meant that differences appeared in the precipitation amounts between the observation and KLDAS,because of scale mismatch;however the occurrence times of precipitation were similar.The range of air temperature at the study sites was 290–304 K(CLM using Asiaflux forcing)and 287–302 K(CLM using KLDAS forcing).The variation in air temperature from Asiaflux was larger than the assimilated air temperature from the KLDAS data.The air temperature and specific humidity showed similar trends in terms of variation,although there was less fluctuation in the specific humidity than there was for air temperature.Precipitation data in KLDAS were similar to theobservations;however,the KLDAS precipitation data at HFK overestimatedduringthesameperiod.Theincomingsolarradiation was underestimated when the precipitation was overestimated.Although the Asiaflux data showed higher values than the KLDAS data,both datasets generally indicated similar tendencies for each forcing variable.

Table 3.Data sources of KLDAS.

4.Model parameters and initial data settings

ToexecuteCLM,initializationandparameterizationwere needed.The necessary data for parameterization are longitude/latitude,soil texture profile,soil color index,porosity, leaf area index,height of the forcing measurements,and percentages of land cover types based on IGBP classifications 1–18.The parameterization details are given in Table 4.HFK was considered to be cropland,while GDK and SMF were mixed forest.The soil texture and porosity were predetermined soil characteristics,and the leaf area index and canopy roughness length were predefined land cover conditions.

The initial data such as soil moisture,soil temperature, and air temperature at the point scale were obtained from the Asiaflux observations.The initial data for the regional scale were provided by different sources.The soil moisture, soiltemperature,andairtemperaturewereacquiredfromKLDAS,as estimated using the Noah LSM.The parameterizations,such as the site locations,soil texture,soil color,and height of the forcing measurements,for the point scale,were taken from Asiaflux.For the regional scale,the United States Geological Survey land use/land cover classifications of KLDAS were converted into IGBP classifications(Liang et al., 2005).

5.Statistics

Table 4.Initial parameterization for point scale estimation of land cover conditions.

Statistics were used to estimate error.The accuracy of the CLM results was assessed using the RMSE and bias between flux observations and CLM results.The formulas used for the regression analysis are as follows:where n,xobservedand xsimulatedare the number of data points, observedvalues,andCLMresults,respectively.Inaddition,a simple linear regression analysis was performed to obtain regression coefficients,error terms,and correlation coefficients (r).Subsequently,regression analysis was used to determine the linearity.

A Taylor diagram is an effective way to show statistical values such as r,standard deviation,and RMSE(Taylor, 2001).The azimuthal position represents r,and the radial axis represents normalized standard deviation.The curved line at the center of the first quadrant signifies RMSE.The insitu data,the observations,represent the reference data plotted along the horizontal axis.

6.Results and discussion

6.1.Validation with observations:temporal patterns of energy fluxes

The CLM results at each site during the summer monsoon season of 2006 were assessed by plotting time series graphs of the simulations and observations(Figs.2–7).The corresponding Taylor diagrams are also presented in these figures. The statistical results are presented in Table 5.The point scale indicates the CLM results using the Asiaflux data,while the regional scale indicates the CLM results using KLDAS. 6.1.1.Net radiation

The temporal variations in simulated net radiation are presented in Fig.2.Net radiation is a fundamental quantity of energy required to change natural phenomena(i.e.,surface evapotranspiration,heating of the land surface and soil, photosynthesis,and energy storage within vegetation).The CLM reasonably and accurately reproduced the net radiation(Figs.2a–c).The net radiation values estimated using the CLM with Asiaflux and KLDAS data showed similar patterns to the observations,although the estimated net radiation from the KLDAS data was slightly underestimated. The accuracy of estimated net radiation was attributed to the reliable simulation of the albedo,reflected solar radiation, and upwelling longwave radiation in CLM(Choi et al.,2010; Li et al.,2012).During rainfall periods,there was increased disagreement between the observed and estimated net radiation for the Asiaflux and KLDAS data,primarily on days 176,182,185 and 189(Fig.2a)for HFK;days 177,181,and 188 for GDK(Fig.2b);and days 177 and 190 for SMF(Fig. 2c).The values during periods of heavy rainfall were much smaller than those during rainless periods due to the effects of precipitation on the estimated net radiation in the Asiaflux and KLDAS datasets.The range of the observed and estimated net radiation was between-80 to 800 W m-2on clear days and-80 to 400 W m-2during periods of rainfall(Fig. 2)because of the influence of clouds on the radiation budget (Ohring and Clapp,1980);net radiation can be computed more accurately under clear-sky conditions(Xiu and Pleim, 2001).In addition,relatively small values of net radiation are caused by rainfall,which affects evaporation and albedoconsiderably(Charney,1975).The r values for the point scale (using Asiaflux data)were all 0.99.The r values of the estimated net radiation with the KLDAS data ranged from 0.82 to 0.91 for all of the sites.The normalized standard devia-tions for the net radiation were approximately 1.0,indicating good agreement with observations(Fig.2d).The net radiation at the point scale generally showed higher r and lower RMSE and bias than the statistical results using the KLDAS data(Table 5);however,our statistical analysis of the KLDAS data indicated high reliability,with r values exceeding 0.82.

Table 5.Statistical error between observations and CLM estimations using Asiaflux,KLDAS,and GLDAS(r:correlation coefficient).

6.1.2.Sensible heat flux

The estimated sensible heat flux(Fig.3)for both the CLM results with Asiaflux and KLDAS forcings was overestimated,compared with the observed sensible heat flux.The estimated sensible heat flux of CLM using the Asiaflux data showed an overestimation of approximately 400 W m-2,except during periods of rainfall(days 176,182,185 and 189 at HFK;days 177,181 and 188 at GDK;and days 177 and 190 at SMF),as shown in Figs.3a–c,respectively.Whitfield et al.(2006)and Li et al.(2012)indicated an overestimated trend for the estimated sensible heat flux of 0–200 W m-2using CLM,and Seuffert et al.(2002)noted that an overestimation of the sensible heat flux under dry soil conditions.At GDK and SMF(Figs.3b and c),the estimated sensible heat flux from CLM with Asiaflux was larger than that from CLM with KLDAS.This was probably caused by the estimation of surface temperature in the LSM.Winter and Eltahir(2010)demonstrated that the overestimated sensible heat flux stemmed from the overestimation of surface temperature as a consequence of excess shortwave radiation in the Regional Climate Model.During the rainfall periods, the estimated sensible heat flux from CLM with Asiaflux and KLDAS forcings was negative(Figs.3a–3c),and the RMSE wasmuchhigher(19–297Wm-2)thanthoseobtainedduring rainless periods(45–145 W m-2).These discrepancies were influenced by rainfall,as well as observation errors,which may have arisen from the precipitation measurements(Li et al.,2012).In particular,the results from the KLDAS data were contradictory to the observations made during periods of precipitation,which adversely affected the statistical analysis,producing RMSE values ranging from 51 to 146 W m-2and bias values ranging from-36 to 48 W m-2(results not shown).This was due to the fact that a considerable portion of the latent land heat transfers to the atmosphere during rainfall,viaevaporation,atforestsites(Mizutanietal.,1997). The negative values,which mainly occurred at night,indicate that the soil temperature was lower than the air temperature (Andersen et al.,2013);however,the r value from both the Asiaflux and KLDAS data exceeded 0.57.The standard deviations ranged from 1.0 to 3.5(Fig.3d).The r values of the sensible heat flux were 0.77–0.92,for the point scale(Table 5).The statistical errors at the regional scale showed moderate r values ranging from 0.57 to 0.76(Table 5).

6.1.3.Latent heat flux

Fig.2.Net radiation at(a)Haenam(HFK),(b)Gwangneung(GDK)and(c)Seto(SMF).(d)Taylor diagram for net radiation.

Fig.3.Sensible heat flux at(a)Haenam,(b)Gwangneung and(c)Seto.(d)Taylor diagram for sensible heat flux.

Latent and sensible heat fluxes produce changes that tend towards balancing the energy budget(St¨ockli et al.,2008).In the present research,the simulated latent heat flux(Fig.4) showed comparable trends to the observed latent heat flux. The estimated latent heat flux showed some abrupt increases and was generally higher during periods of rainfall(Figs.4a–c),which was in direct opposition to the estimated sensible heat flux.LSMs mainly show trends in which the sensible heat flux is overestimated and the latent heat flux is underestimated,except during periods of heavy rainfall(Xiu and Pleim,2001;Yang et al.,2007;Chen et al.,2010).The differences between observations and estimations during these periods may be due to inappropriate or missing observation values,caused by the physical deterioration of the measurement instruments during the East Asian summer monsoon (Shi et al.,2008).The bias of observed precipitation could also account for an underestimation of the evapotranspiration and the data discrepancies during periods of heavy rainfall, therefore(Ye et al.,2012).In addition,precipitation with water table changes has been closely linked to an overestimation of the latent heat flux in wet surface conditions(Comer et al., 2000;Admiral et al.,2006).The estimated latent heat flux from the CLM results with Asiaflux forcing was slightly high for the HFK(Fig.4a)and SMF(Fig.4c)sites.The estimated latent heat flux at the GDK and SMF sites was higher than at the HFK site because of the greater latent heat flux under the green vegetated conditions(LeMone et al.,2007).

As shown in Table 5,both the estimated sensible and latent heat fluxes at HFK showed relatively low errors compared with the two other sites.This was due to the fact that the measurement of heat flux under forest land cover is difficult,as the canopy volume and horizontal inhomogeneity scale are in constant fluctuation(Brutsaert,1982;Katul et al.,1996;Choi et al.,2012).This leads to the potential for increased error under forest conditions.The normalized standard deviations of the latent heat flux were 0.50–1.5(Fig.4d). The r values were in acceptable ranges of 0.58–0.81,for the point scale,and 0.40–0.66 for the regional scale(Table 5).

6.1.4.Ground heat flux

The ground heat flux is a major component of the energy budget.Although it has values that are relatively lower (approximately 100 W m-2)than for other energy fluxes (Liebethal et al.,2005),ground heat flux is a major component in the energy balance between the land surface and the atmosphere.This variable is a residual value from the energy balance equation(Dai et al.,2003).In the present study, the estimated ground heat flux from the CLM with Asiaflux forcing was overestimated in comparison with general observations.During the rainfall periods,both the observed and estimated ground heat fluxes had low(close to zero)values. The diurnal variation in heat flux is influenced by soil moisture and incoming solar radiation,and the net flow of heat is directed towards the soil on dry days and away from the soil during rainfall periods(Tessy Chacko and Renuka,2002).In our study,the estimated ground heat flux ranged from-60 to 100 W m-2(CLM using Asiaflux forcing)and from-50 to 50 W m-2(CLM using KLDAS forcing)(Fig.5).These ranges were wider than those from the observation data.Inparticular,the observed ground heat flux at GDK and SMF indicated minute fluctuations that differed from the patterns observed at other sites in previous studies(Choi et al.,2010; Hong and Kim,2010).The ground heat flux in forests generally shows low values because solar radiation is intercepted by vegetation(Oliver et al.,1987;Carlson and Groot,1997); however,the values of ground heat flux vary from those calculated with the energy balance closure equation(latent heat flux+sensible heat flux=net radiation-ground heat flux). In this research,the large differences between the observations and simulations showed the highest normalized standard deviation and lowest r at the GDK site(Fig.5d).

Fig.4.Latent heat flux at(a)Haenam,(b)Gwangneung and(c)Seto.(d)Taylor diagram for latent heat flux.

6.2.Validation with observations:temporal patterns of water fluxes

6.2.1.Soil moisture

Soil moisture increases as a result of precipitation,and subsequently decreases steadily until the next rainfall event. Soil moisture is related to the Bowen ratio and net radiation which are energy fluxes.In particular,higher latent heat flux is associated with increased soil moisture;therefore,soil moisture plays an important role in energy balance partitioning(Jones and Brunsell,2009).The estimated soil moisture from CLM with Asiaflux forcing was underestimated at HFK (Fig.6a)and GDK(Fig.6b).In cases of underestimated soil moisture,the soil moisture in LSMs is occasionally overestimated under the occurrence of instantaneous precipitation (Liangetal.,1996).Thenormalizedstandarddeviationswere predominantly less than 1.0 at GDK(Fig.6d).At HFK,the observed soil moisture values were larger than those obtained from the other sites because of the rice paddy irrigation systems(Reshmidevi et al.,2008).The estimated soil moisture from CLM with KLDAS forcing reflected the effects of precipitation well;however,there were differences of 0.1–0.2 m3m-3between the estimations and observations at all of the study sites(Figs.6a–c).The differences in estimated soil moisture were caused by heterogeneous land cover types in the grid(Jackson et al.,2010;Choi and Hur,2012),which were rescaled for 10 km.Dai et al.(2003)attributed CLM’s under-or overestimation of soil moisture to clay content,conductivity,soil suction,and soil volumetric content at the wilting point.These aforementioned studies confirmed that soil moisture estimations tend to be lower than the in-situ data. Soil moisture results can be improved by the acquisition of more realistic initial conditions for a wide range of areas,validated for various sites that have different land and climate conditions.

6.2.2.Soil temperature

Fig.5.Ground heat flux at(a)Haenam,(b)Gwangneung and(c)Seto.(d)Taylor diagram for ground heat flux.

The diurnal fluctuation patterns for the estimated soil temperature for the CLM results with Asiaflux and KLDAS forcings alike,showed good agreement with the observed soil temperatures(Fig.7).The estimated soil temperature from CLM with Asiaflux forcing was slightly higher(0.5–1.0 K) than the observed soil temperature.However,the estimated soil temperature from CLM with KLDAS forcing showed under-or overestimations of approximately 2.5 K,probably caused by scale mismatches and the effects of other variables, such as estimated soil moisture.Li et al.(2012)referred that soil thermal conductivity and soil porosity influenced errors in estimated soil temperatures.The normalized standard deviations in the present study ranged from 0.9 to 2.7(Fig.7d). Meanwhile,the RMSE ranged from 0.79–1.62 K for CLM with Asiaflux forcing,and 1.38–2.62 K for CLM with KLDAS forcing(Table 5).

6.3.Spatial patterns of estimated energy and water fluxes

To compare the spatial distributions for each of the energy and water fluxes,the estimated fluxes from the CLM results with KLDAS and GLDAS forcings,as well as from the standalone GLDAS data(0300 UTC 1 July 2006),are shown in Figs.8–10.The time period was selected in order to show more accurate comparisons between the spatial distributions of the data from KLDAS and GLDAS for various climate conditions.

As seen in Fig.8,precipitation maps enable the analysis of precipitation effects on energy and water fluxes.Figure 8 shows similar rainfall distributions in South Korea and a part of China where heavy rainfall occurred.

Fig.6.Surface soil moisture at(a)Haenam,(b)Gwangneung and(c)Seto.(d)Taylor diagram for soil moisture.

Fig.7.Surface soil temperature at(a)Haenam,(b)Gwangneung and(c)Seto.(d)Taylor diagram for soil temperature.

Fig.8.Spatial distributions of precipitation of(a)KLDAS and(b)GLDAS,at 0300 UTC 1 July 2006.

The spatial distributions of net radiation showed similar tendencies for all the regions(Figs.9a–d).The Clouds and the Earth’s Radiant Energy System(CERES)data were analyzed with other spatial distribution maps in terms of the net radiation(Fig.9d).The CERES data are reanalysis data using MODIS and Visible and Infrared Sounder measurements.The information shown in Figs.9b and c are similar because they were produced from GLDAS.The net radiation from CERES also showed a similar distribution,with relatively high values in northern Japan and low values around Korea(Fig.9d).However,the estimated net radiation from CLM with GLDAS forcing(Figs.9b)showed higher values over East China compared with the net radiation from the standalone GLDAS data.This resulted from overestimated sensible heat flux,which appeared in East China in the absence of rainfall.The precipitation effects were strong in CLM,with relatively low values of net radiation for the regions of Korea,Japan,and southeastern China(Figs.9a and b).In addition,Figs.9b and c show that regions affected by precipitation were highly analogous.Overall,the estimated net radiation using the high resolution dataset was able to provide more detailed information according to area and weather conditions,as compared with that of the low resolution dataset.The estimated sensible heat fluxes with the KLDAS and GLDAS forcings,and the sensible heat flux from the standalone GLDAS data are shown in Figs.9e–g. Whereas the estimated sensible heat flux from CLM with GLDAS forcing and the sensible heat flux from GLDAS had relatively high values(0–50 W m-2)in rainfall areas(Figs.9f and g),the spatial distribution from CLM with KLDAS forcing presented low values(approximately-200 W m-2)(Fig. 9e).This was because CLM significantly underestimated the sensible heat flux during rainfall periods(refer to the temporal analysis section and Fig.3).Although the sensible heat flux values in each grid for CLM with KLDAS(Fig.9e)and GLDAS forcings(Fig.9f),and the standalone GLDAS data (Fig.9g)slightly differed,the spatial tendencies were similar. The sensible heat flux values ranged between approximately -200to550Wm-2.TheresultsinFig.9eshowlowervalues than the GLDAS data(Fig.9f)in central East China due to the effect of precipitation,as mentioned above.The overestimation of latent heat flux was linked to an underestimation of the sensible heat flux in accordance with the energy balance method(St¨ockli et al.,2008).This aspect also explains the low values(approximately-200 W m-2)over central East China at 0300 UTC(Fig.9e).In addition,the sensible heat fluxes from GLDAS showed data smoothing due to the characteristics of the system that produced the data,i.e.,through assimilation of the in-situ,satellite and modeling data.Thus, the GLDAS data showed smoother values of heat fluxes.The spatial distributions of the latent heat flux indicated a slightly different tendency in areas of heavy rainfall(Figs.9h–j).Figure 9g shows particularly high values,greater than 500 W m-2,for South Korea and parts of Japan and China,which were likely due to the effects of precipitation.CLM overestimated the latent heat flux during periods of rainfall,as compared with ground-based observations,which was contrary to the estimated sensible heat flux.The estimated latent heat flux ranged from-30 to 550 W m-2(Figs.9h and i).Jang et al.(2013)calculated evapotranspiration values using a retrieval algorithm designed using MODIS and KLDAS data.They identified the spatial distribution of annual evapotranspiration from 2006 to 2008,and their results appear comparable to our findings(Figs.9h–j),which show higher evapotranspiration over southern areas compared with northern areas.The spatial distributions of ground heat flux are shown in Figs.9k–m.Although the values of the ground heat flux from CLM with KLDAS forcing(Fig.9k)were generally lower than the ground heat flux from CLM with GLDAS forcing,and the standalone GLDAS data(Figs.9l and m),the spatial distributions were roughly similar.

The soil moisture patterns(Figs.10a–c)were fairly similar;however,the map shown in Fig.10b presents dryer patterns compared with the other results.We assume that the overestimated soil temperature(Fig.10e)led to the underestimated soil moisture in Fig.10b.Nevertheless,the estimated soil moisture from CLM with GLDAS forcing was comparable to the rainfallarea in Fig.10c.In addition,the dry and wet patterns were highly similar.The upper left side shows drier patterns on account of the local arid characteristics,and the southern side was wetter than the northern side due to rainfall effects(Figs.10a–c).In Fig.10a,the result from CLM with KLDAS forcing indicates a detailed spatial distribution compared with the other maps,due to the higher spatial resolution than the GLDAS data.Not found in the reference list.Please check.presented an estimated surface soil moisture map for China,Korea,and part of Japan for July and August 2006 using CLM.Their soil moisture values resemble the estimated soil moisture reported in the present study,with higher soil moisture values found in Korea,southern Japan and southern China,and lower soil moisture values calculated for the more arid central region of China.The rainfall areas showed relatively high values of soil moisture,at approximately 0.45–0.50 m3m-3.The soil moisture in Figs.10a–c ranged from 0.01 to 0.44 m3m-3(Fig.10a),from 0.1 to 0.5 m3m-3(Fig.10b),and from 0.01 to 0.46 m3m-3(Fig.10c).In Figs. 10d–f,the estimated soil temperature from CLM with KLDAS and GLDAS forcings as well as that from the standalone GLDAS data,are presented.The estimated soil temperature from CLM with KLDAS forcing generally showed a distribution of no significant difference.However,the information in Fig.10e shows higher values in the upper part of themap compared with Fig.10d.As seen on the left side of Fig. 10f,the soil temperature from GLDAS was higher than from CLM with KLDAS forcing(Fig.10d).However,the general patterns of estimated soil temperature were similar between CLM with KLDAS forcing and the standalone GLDAS data. The soil temperature values varied between 278 and 305 K (Fig.10d),280 and 310 K(Fig.10e),and 280 and 309 K (Fig.10f).

Fig.9.Spatial distributions for energy fluxes from(a,e,h and k)CLM with KLDAS forcing,(b,f,i and l) CLM with GLDAS forcing,(c,g,j,and m)standalone GLDAS data,and(d)CERES,at 0300 UTC 1 July 2006.

Fig.9.(Continued.)

Overall,the spatial distribution patterns of the energy and water fluxes from CLM with KLDAS and GLDAS forcings were similar.The estimated fluxes from CLM with GLDAS forcing displayed a closer tendency with the standalone GLDAS data because the forcing data were identical.However,the estimated energy fluxes from CLM with KLDAS forcing also showed a similar spatial distribution to those based on the standalone GLDAS data.Therefore,KLDAS provides a good alternative meteorological forcing,as compared with GLDAS data.

7.Conclusion

Fig.10.Spatial distributions for water fluxes from(a,d)CLM with KLDAS forcing,(b,e)CLM with GLDAS forcing,and(c,f)standalone GLDAS data,at 0300 UTC 1 July 2006.

Energy and water fluxes were estimated using CLM,operated in offline mode,with different forcing data from Asiaflux(forthepointscale)andKLDAS(fortheregionalscale). The CLM estimations were validated with observations from Asiaflux sites in East Asia.The temporal variations in energy and water fluxes from CLM with Asiaflux and KLDAS forcings agreed well with the ground-based observations.The CLM results with KLDAS forcing also reflected the effects of precipitation accurately.There were more discrepancies betweentheobservationsandtheCLMestimationsusingKLDAS forcing than those with Asiaflux forcing,due to the coarse resolution of KLDAS being unable to accurately reflect the heterogeneity of land cover types or soil textures on the regional scale.However,statistical analysis indicated that results using KLDAS showed high reliability at the point scale.In addition,the spatial distributions for each variable from CLM with KLDAS forcing were similar to those from CLMwithGLDASforcing,aswellasthestandaloneGLDAS data,used as reference data.In particular,KLDAS showed several advantages over GLDAS.For instance,it is able to provide data at relatively high spatial resolution(10 km)forEast Asia,as well as spatially continuous data at intervals of 1 h.

The present study is helpful because it demonstrates that, through use of CLM with KLDAS data,which have a relatively high resolution and short interval compared with GLDAS data,the analysis of energy and water fluxes in East Asia can be increased from one to two dimensions.In future work,the accuracy of estimated energy and water fluxes using CLM with high spatial resolution data should be further improved via parameter optimization and the acquisition of more precise initial data across a broader range of areas throughout East Asia.

Acknowledgements.This research was supported by Space Core Technology Development Program through the National Research Foundation of Korea(NRF)funded by the Ministry of Science,ICT and Future Planning(NRF-2014M1A3A3A02034789). This research was supported by Basic Science Research Program through the National Research Foundation of Korea(NRF)funded by the Ministry of Education(NRF-2013R1A1A2A10004743).We sincerely thank the Asiaflux for their data collection efforts.Dr. Minseok KANG’s contribution to this research was supported by the Korea Meteorological Administration Research and Development Program under Grant Weather Information Service Engine(WISE) project,KMA-2012-0001-A.

REFERENCES

Abramowitz,G.,R.Leuning,M.Clark,and A.Pitman,2008: Evaluating the performance of land surface models.J.Climate,21,5468–5481.

Admiral,S.W.,P.M.Lafleur,and N.T.Roulet,2006:Controls on latent heat flux and energy partitioning at a peat bog in eastern Canada.Agricultural and Forest Meteorology,140,308–321.

Andersen,T.K.,D.E.Radcliffe,and J.M.Shepherd,2013:Quantifying surface energy fluxes in the vicinity of inland-tracking tropical cyclones.Journal Applied Meteorology and Climatology,52,2797–2808.

Arora,V.,2002:Modeling vegetation as a dynamic component in soil–vegetation–atmosphere transfer schemes and hydrological models.Rev.Geophys.,40,3-1–3-26.

AsiaFlux website,cited 2013:Site information.[Available online at http://www.Asiaflux.net/.Accessed 20 Oct 2013]

Baldocchi,D.,and Coauthors,2001:FLUXNET:A new tool to study the temporal and spatial variability of ecosystem-scale carbon dioxide,water vapor,and energy flux densities.Bull. Amer.Meteor.Soc.,82,2415–2434.

Betts,R.A.,P.M.Cox,and F.I.Woodward,2000:Simulated responses of potential vegetation to doubled-CO2climate change and feedbacks on near-surface temperature.Global Ecology and Biogeography,9,171–180.

Betts,A.K.,R.L.Desjardins,and D.Worth,2007:Impact of agriculture,forest and cloud feedback on the surface energy budget in BOREAS.Agricultural and Forest Meteorology,142, 156–169.

Beven,K.J.,and M.J.Kirkby,1979:A physically based variable contributingareamodelofbasinhydrology.HydrologicalSciences Bulletin,24,43–69.

Brutsaert,W.,1982:Evaporation into the Atmosphere:Theory, History,and Applications.Springer,299.

Carlson,D.W.and A.Groot,1997:Microclimate of clear-cut, forest interior,and small openings in trembling aspen forest. Agric.For.Meteorol.,87,313–329.

Charney,J.G.,1975:Dynamics of deserts and drought in the Sahel.Quart.J.Roy.Meteor.Soc.,101,193–202.

Chen,F.,and Coauthors,1996:Modeling of land-surface evaporation by four schemes and comparison with FIFE observations. J.Geophys.Res.,101,7251–7268.

Chen,Y.Y.,K.Yang,D.G.Zhou,J.Qin,and X.F.Guo,2010:Improving the Noah land surface model in arid regions with an appropriate parameterization of the thermal roughness length. Journal of Hydrometeorology,11,995–1006.

Choi,M.,S.O.Lee,and H.Kwon,2010:Understanding of the Common Land Model performance for water and energy fluxes in a farmland during the growing season in Korea.Hydrological Processes,24,1063–1071.

Choi,M.,and Y.Hur,2012:A microwave-optical/infrared disaggregationforimprovingspatialrepresentationofsoilmoisture using AMSR-E and MODIS products.Remote Sensing of Environment,124,259–269.

Choi,M.,W.P.Kustas,and R.L.Ray,2012:Evapotranspiration models of different complexity for multiple land cover types. Hydrological Processes,26,2962–2972.

Chun,J.A.,R.A.Cooke,M.S.Kang,M.Choi,D.Timlin,and S. W.Park,2010:Runoff losses of suspended sediment,nitrogen,and phosphorus from a small watershed in Korea.Journal of Environment Quality,39,981–990.

Clapp,R.B.,and G.M.Hornberger,1978:Empirical equations for some soil hydraulic properties.Water Resour.Res.,14,601–604.

Collatz,G.J.,J.T.Ball,C.Grivet,and J.A.Berry,1991:Physiological and environmental regulation of stomatal conductance,photosynthesisandtranspiration:Amodelthatincludes a laminar boundary layer.Agricultural and Forest Meteorology,54,107–136.

Comer,N.T.,P.M.Lafleur,N.T.Roulet,M.G.Letts,M.Skarupa, and D.Verseghy,2000:A test of the Canadian Land Surface Scheme(CLASS)for a variety of wetland types.Atmos.-Ocean,38,161–179.

Cox,P.M.,R.A.Betts,C.D.Jones,S.A.Spall,and I.J.Totterdell, 2000:Acceleration of global warming due to carbon-cycle feedbacks in a coupled climate model.Nature,408,184–187.

Dai,Y.J.,and Q.C.Zeng,1997:A land surface model(IAP94) for climate studies part I:Formulation and validation in off-line experiments.Adv.Atmos.Sci.,14,433–460,doi: 10.1007/s00376-997-0063-4.

Dai,Y.J.,X.Zeng,and R.E.Dickinson,2001:The Common Land Model(CLM):Technical documentation and user’s guide. Georgia Institute of Technology,Atlanta.

Dai,Y.J.,and Coauthors,2003:The common land model.Bull. Amer.Meteor.Soc.,84,1013–1023.

Dickinson,R.E.,A.Henderson-Sellers,and P.J.Kennedy,1993: Biosphere-atmosphere transfer scheme(BATS)version 1e as coupled to the NCAR community climate model.NCAR Technical Note NCAR/TN-387+STR.

Dickinson,R.E.,M.Shaikh,R.Bryant,and L.Graumlich,1998: Interactive canopies for a climate model.J.Climate,11, 2823–2836.

Famiglietti,J.S.,and J.S.Wood,1994:Application of multiscale water and energy balance models on a tallgrass prairie.Water Resour.Res.,30,3079–3093.

Farouki,O.T.,1981:Thermal properties of soils.United States Army Cold Regions Research and Engineering Laboratory, Monograph No.81-1,136 pp.

Forest news website,cited 2013:Site information.[Available online at http://www.sanlim.kr/.]

Friedlingstein,P.,J.-L.Dufresne,P.M.Cox,and P.Rayner,2003: How positive is the feedback between climate change and the carbon cycle?Tellus B,55,692–700.

Hong,J.,H.C.Lee,J.Kim,B.Kim,C.Cho,and S.Lee,2003: Inferring regional scale surface heat flux around FK Koflux site:From one point tower measurement to MM5 mesoscale model.Korean Journal of Agricultural and Forest Meteorology,5,138–149.(in Korean)

Hong,J.,and J.Kim,2010:Numerical study of surface energy partitioning on the Tibetan plateau:Comparative analysis of two biosphere models.Biogeosciences,7,557–568.

Huang,C.L.,X.H.Li,J.M.Wang,and J.Gu,2008:Assimilation of remote sensing data products into Common Land Model for evapotranspiration forecasting.Proc.8th International Symposium on Spatial Accuracy Assessment in Natural Resources and Environmental Sciences,Shanghai,China: World Academic Union,234–241.

Jackson,T.J.,and Coauthors,2010:Validation of advanced microwave scanning radiometer soil moisture products.IEEE Trans.Geosci.Remote Sens.,48,4256–4272.

Jang,K.,S.Kang,Y.J.Lim,S.Jeong,J.Kim,J.S.Kimball,and S. Y.Hong,2013:MonitoringdailyevapotranspirationinNortheast Asia using MODIS and a regional Land Data Assimilation System.J.Geophys.Res.,118,12 927–12 940.

Jones,A.R.,and N.B.Brunsell,2009:Energy balance partitioning and net radiation controls on soil moisture-precipitation feedbacks.Earth Interactions,13,1–25.

Katul,G.,C.-I.Hsieh,R.Oren,D.Ellsworth,and N.Phillips, 1996:Latent and sensible heat flux predictions from a uniform pine forest using surface renewal and flux variance methods.Bound.-Layer Meteor.,80,249–282.

Kiehl,J.T.,J.J.Hack,G.B.Bonan,B.A.Boville,B.P.Briegleb, D.L.Williamson,and P.J.Rasch,1996:Description of the NCAR Community Climate Model(CCM3).NCAR Technical Note NCAR/TN-420+STR,Note 152.

Koster,R.D.,and M.J.Suarez,1992:Modeling the land surface boundary in climate models as a composite of independent vegetation stands.J.Geophys.Res.:Atmos.,97,2697–2715.

Kwon,H.,T.–Y.Park,J.Hong,J.–H.Lim,and J.Kim,2009: Seasonality of net ecosystem carbon exchange in two major plant functional types in Korea.Asia-Pacific Journal of Atmospheric Sciences,45,149–163.

Lei,H.M.,D.W.Yang,Y.J.Shen,Y.Liu,and Y.C.Zhang,2011: Simulation of evapotranspiration and carbon dioxide flux in the wheat-maize rotation croplands of the North China Plain using the Simple Biosphere Model.Hydrological Processes, 25,3107–3120.

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.Journal of Hydrometeorology,8, 68–87.

Li,Z.–C.,Z.–G.Wei,C.Wang,Z.–Y.Zheng,H.Wei,and H.Liu, 2012:Simulation and improvement of common land model on the bare soil of Loess Plateau underlying surface.Environmental Earth Sciences,66,1091–1097.

Liang,X.,F.W.Eric,and D.P.Lettenmaier,1996:Surface soil moisture parameterization of the VIC-2L model:Evaluation and modification.Global and Planetary Change,13,195–206.

Liang,X.–Z.,H.I.Choi,K.E.Kunkel,Y.J.Dai,E.Joseph,J.X. L.Wang,and P.Kumar,2005:Surface boundary conditions for mesoscale regional climate models.Earth Interactions,9, 1–28.

Liebethal,C.,B.Huwe,and T.Foken,2005:Sensitivity analysis for two ground heat flux calculation approaches.Agricultural and Forest Meteorology,132,253–262.

Lim,Y.-J.,K.-Y.Byun,T.-Y.Lee,and J.Kim,2010:Evaluation of evapotranspiration using Korea Land Data Assimilation System.Korean Journal of Agricultural and Forest Meteorology, 12,298–306.

Lim,Y.–J.,K.–Y.Byun,T.–Y.Lee,H.Kwon,J.Hong,and J.Kim, 2012:A land data assimilation system using the MODIS-derived land data and its application to numerical weather prediction in East Asia.Asia-Pacific Journal of Atmospheric Sciences,48,83–95.

Matumoto,K.,and Coauthors,2008:Energy consumption and evapotranspiration at several boreal and temperate forests in the Far East.Agricultural and Forest Meteorology,148, 1978–1989.

Meng,C.L.,Z.–L.Li,X.Zhan,J.C.Shi,and C.Y.Liu,2009: Land surface temperature data assimilation and its impact on evapotranspiration estimates from the Common Land Model. Water Resour.Res.,45,W02421.

Mizutani,K.,K.Yamanoi,T.Ikeda,and T.Watanabe,1997:Applicability of the eddy correlation method to measure sensible heat transfer to forest under rainfall conditions.Agricultural and Forest Meteorology,86,193–203.

Niu,G.Y.,and Z.L.Yang,2006:Assessing a land surface model’s improvements with GRACE estimates.Geophys.Res.Lett., 33,L07401.

Ohring,G.,and P.Clapp,1980:The effect of changes in cloud amount on the net radiation at the top of the atmosphere.J. Atmos.Sci.,37,447–454.

Oliver,S.A.,H.R.Oliver,J.S.Wallace,and A.M.Roberts,1987: Soil heat flux and temperature variation with vegetation,soil type and climate.Agricultural and Forest Meteorology,39, 257–269.

Pielke,R.A.,2001:Influence of the spatial distribution of vegetationandsoilsonthepredictionofcumulusconvectiverainfall. Riv.Geophys.,39,151–177.

Reshmidevi,T.V.,R.Jana,and T.I.Eldho,2008:Geospatial estimation of soil moisture in rain-fed paddy fields using SCSCN-based model.Agricultural Water Management,95,447–457.

Rihani,J.F.,R.M.Maxwell,and F.K.Chow,2010:Coupling groundwater and land surface processes:Idealized simulations to identify effects of terrain and subsurface heterogeneity on land surface energy fluxes.Water Resour.Res.,46, W12523.

Rodell,M.,and Coauthors,2004:The global land data assimilation system.Bull.Amer.Meteor.Soc.,85,381–394.

Sellers,P.J.,and Coauthors,1996:A revised land surface parameterization(SiB2)for atmospheric GCMs.Part I:Model formulation.J.Climate,9,676–705.

Seneviratne,S.I.,and R.St¨ockli,2006:The role of landatmosphere interactions for climate variability in Europe.Climate Variability and Extremes During the Past 100 Years,S.Br¨onnimann,J.Luterbacher,T.Ewen,H.F.Diaz,R.S.Stolarski,and U.Neu,Eds.,Springer,33,179–193.

Seuffert,G.,P.Gross,and C.Simmer,2002:The influence of hydrologic modeling on the predicted local weather:Two-way coupling of a mesoscale weather prediction model and a land surface hydrologic model.Journal of Hydrometeorology,3, 505–523.

Shi,T.T.,D.X.Guan,A.Z.Wang,J.B.Wu,C.J.Jin,and S.J. Han,2008:Comparison of three models to estimate evapotranspiration for a temperate mixed forest.Hydrological Processes,22,3431–3443.

St¨ockli,R.,and Coauthors,2008:Use of FLUXNET in the Community Land Model development.J.Geophys.Res.,113, G01025.

Taylor,K.E.,2001:Summarizing multiple aspects of model performance in a single diagram.J.Geophys.Res.,106,7183–7192.

Tessy Chacko,P.,and G.Renuka,2002:Temperature mapping, thermal diffusivity and subsoil heat flux at Kariavattom of Kerala.Journal of Earth System Science,111,79–85.

Wang,Y.P.,and P.G.Jarvis,1990:Description and validation of an array model-MAESTRO.Agricultural and Forest Meteorology,51,257–280.

Whitfield,B.,J.M.Jacobs,and J.Judge,2006:Intercomparison study of the land surface process model and the common land model for a Prairie Wetland in Florida.Journal of Hydrometeorology,7,1247–1258.

Wilczak,J.M.,S.P.Oncley,and S.A.Stage,2001:Sonic anemometer tilt correction algorithms.Bound.-Layer Meteor., 99,127–150.

Winter,J.M.,and E.A.B.Eltahir,2010:The sensitivity of latent heat flux to changes in the Radiative Forcing:A framework forcomparingmodelsandobservations.J.Climate,23,2345–2356.

Xiu,A.,and J.E.Pleim,2001:Development of a land surface model.Part I:Application in a mesoscale meteorological model.J.Appl.Meteor.,40,192–209.

Yang,K.,T.Watanabe,T.Koike,X.Li,H.Fujii,K.Tamagawa,Y. Ma,and H.Ishikawa,2007:Auto-calibration system developed to assimilate AMSR-E data into a Land Surface Model for estimating soil moisture and the surface energy budget.J. Meteor.Soc.Japan,85A,229–242.

Ye,B.,D.Yang,and L.Ma,2012:Effect of precipitation bias correction on water budget calculation in Upper Yellow River, China.Environ.Res.Lett.,7,025201.

Zou,L.,W.Gao,T.Wu,Q.Lu,and Y.Zhang,2008:Numerical simulation by the Common Land Model(CLM)of the soil moisture over China during the summer of 2006.In Optical Engineering+Applications,708 314–708 314.

Kim,D.,Y.-J.Lim,M.Kang,and M.Choi,2016:Land response to atmosphere at different resolutions in the Common Land Model over East Asia.Adv.Atmos.Sci.,33(3),391–408,

10.1007/s00376-015-5059-x.

19 February 2015;revised 28 August 2015;accepted 6 September 2015)

∗Minha CHOI

Email:mhchoi@skku.edu