1. Introduction
To support the mission of the National Oceanic and Atmospheric Administration (NOAA) in improving its operational climate prediction on the seasonal scale, the National Centers for Environmental Prediction (NCEP) has produced a new generation of global reanalysis: the NCEP Climate Forecast System Reanalysis (CFSR; Saha et al. 2010). CFSR is designed as a global, high-resolution (T382 global Gaussian grid, approximately 38-km resolution), coupled atmosphere–ocean–land–sea ice system to provide a best estimate of the state of these coupled domains over the period from 1979 to 2009, with a real-time extension. CFSR provides the initial conditions for historical reforecast experiments required to calibrate operational NCEP climate forecast from the Climate Forecast System (CFS). Comparing to the previous two NCEP global reanalyses—namely, the NCEP–National Center for Atmospheric Research (NCAR) reanalysis (R1; Kalnay et al. 1996) and the NCEP–Department of Education (DOE) reanalysis (R2; Kanamitsu et al. 2002)—CFSR uses a Global Land Data Assimilation System (GLDAS; Mitchell et al. 2004; Rodell et al. 2004) for the first time in a global reanalysis to perform the land surface analysis.
The land surface interacts with weather and climate through regulation of the energy and water fluxes between the land surface and the atmosphere on a range of spatial and temporal scales. Land surface is determined by land use, soil type, vegetation type, soil moisture, and the presence of snowpack. Soil moisture plays a critical role in the process of partitioning the surface energy between latent, sensible, and soil heat fluxes. Wet soil promotes evapotranspiration and higher latent heat flux. On the other hand, dry soil leads to low evapotranspiration and higher sensible heat flux. Changes in soil moisture cause changes in the soil heat capacity, conductivity, and diffusivity. These variables control the heat transmitted within the soil layer in the form of soil heat flux, and hence, affect the soil temperature and subsequently the near-surface air temperature. In the transition zone between wet and dry climate, the coupling strength between soil moisture and summertime convective precipitation in climate models has been evaluated (Koster et al. 2004).
Numerical weather and climate prediction and assimilation systems generally use prescribed or simulated soil moisture values to formulate surface layer parameterizations within a land surface model (LSM) to calculate evapotranspiration, latent, sensible, and soil heat fluxes to solve the land surface energy and water balance equations. During the past two decades, significant efforts have been made in improving land surface simulation as well as their impact on the subsequent atmospheric and hydrologic simulations and predictions (Pan and Mahrt 1987; Chen et al. 1996; Wood et al. 1997; Dickinson et al. 1998; Koster et al. 2000; Xue et al. 2001; Ek et al. 2003; Dirmeyer et al. 2006; Oleson et al. 2008). Yet, coupled land–atmospheric prediction and assimilation systems often yield nontrivial error and drift in soil moisture, soil temperature, snowpack, and surface energy and water fluxes, owing to substantial biases in the land surface forcing from the companion atmospheric model (Dirmeyer 2001; Berg et al. 2003; Dirmeyer and Zhao 2004). To improve land surface simulation, the accuracy of precipitation is most important among the other land surface forcing variables (Oki et al. 1999; Chen 2005; Guo et al. 2006; Decharme and Douville 2006; Materia et al. 2010).
In the previous two NCEP global reanalyses (R1 and R2), the Oregon State University (OSU) LSM (Mahrt and Pan 1984; Pan and Mahrt 1987) was used to calculate the land surface states and fluxes in a coupled land–atmosphere system. To overcome the deficiencies in the precipitation forcing from the atmospheric model and to prevent the calculated soil moisture from drifting too far from a preferred climatology, adjustments to soil moisture were applied. In R1, the simulated soil moisture was nudged toward a prescribed global monthly soil moisture climatology derived by Mintz and Serafini (1992). In R2, the simulated soil moisture was adjusted to compensate for differences between the simulated and observed precipitation. The observed precipitation data used in R2 was the pentad global precipitation analysis of the NCEP Climate Prediction Center (CPC) Merged Analysis of Precipitation (CMAP; Xie and Arkin 1997). Lu et al. (2005) concluded that the soil moisture content of R1 carried a too-strong seasonal variability and too-weak interannual variability. When compared with observations in selected locations, the mean value of R1 soil moisture content was close to observations during the wet season but much less in the dry season. The soil moisture content in R2 revealed more consistent interannual and mean seasonal cycles with observations but the mean values were much lower in all seasons. Dirmeyer and Zhao (2004) showed that anomaly correlations with observed monthly soil moisture were higher in R1 than R2 in many locations.
The Noah LSM (Ek et al. 2003) was developed with substantial upgrades from the OSU LSM as a collaborative effort between NCEP and numerous partners from the land–hydrology community over the last decade. Noah was implemented in the NCEP North American Mesoscale model (NAM) in 2003 and the NCEP Global Forecast System (GFS) in 2005 for operational regional and global medium-range weather forecasts, respectively. In CFSR, Noah is implemented in both the land–atmosphere–ocean prediction model CFS to generate the first guess fields of the land–atmosphere simulation, and in GLDAS to perform the land surface analysis. The NASA Land Information System infrastructure (LIS; Peters-Lidard et al. 2007) is used to execute the CFSR GLDAS land surface analysis using observed global precipitation analyses as direct forcing. The adjustment to soil moisture is based on land surface physics responding to the adjustment to precipitation forcing, rather than the artificial adjustments previously applied in R1 and R2.
This paper describes the design and implementation of CFSR GLDAS/LIS, with special attention to the use of observed global precipitation analyses and snow analysis and the procedure to evaluate the CFSR land surface energy and water budgets.
2. LIS infrastructure and configuration
LIS, developed primarily at NASA Goddard Space Flight Center, Hydrological Sciences Branch, in close collaboration with NCEP, is a high-performance parallel computing infrastructure for land surface modeling and assimilation with portable capability to facilitate interoperation with other Earth system modeling and assimilation systems (Peters-Lidard et al. 2007). In CFSR GLDAS, LIS is configured with the identical settings of the land surface component in a coupled land–atmosphere CFS, including the same Noah physics, the same T382 global Gaussian grid specification, and the same land–sea mask, terrain height, soil, and vegetation classes and soil and vegetation parameters (Ek et al. 2003). The global soil dataset of Zobler (1986) and the global vegetation dataset of Hansen et al. (2000) are used. GLDAS/LIS is forced with the CFSR atmospheric data assimilation output and observed precipitation. Land surface states of soil moisture, soil temperature, and snow variables from the GLDAS/LIS simulation are inserted into the CFSR prediction and assimilation systems on a 24-h basis at 0000 UTC for each day of reanalysis production. Comparing to traditional uncoupled land surface hydrologic simulation experiments (Henderson-Sellers et al. 1995; Mitchell et al. 2004; Dirmeyer et al. 2006), where LSMs are driven with atmospheric forcing fields from observation, simulation, and assimilation (but the LSMs’ simulated results do not have any impact on the prescribed atmospheric forcing), this coupled CFSR GLDAS/LIS feeds the simulated soil moisture and temperature back to the coupled land–atmosphere model and influences subsequent simulation of atmospheric forcing. The GLDAS/LIS simulation of land surface conditions and its atmospheric forcing are consistent in CFSR.
3. Forcing
The atmospheric forcing variables required to drive the Noah LSM simulation are surface downward shortwave radiation, surface downward longwave radiation, precipitation, and temperature, humidity, and wind speed at the lowest model layer above surface, air pressure of that lowest model layer and surface, and aerodynamic surface exchange coefficient. In CFSR GLDAS/LIS, most of the forcing variables are from the CFSR global atmospheric data assimilation system (GDAS) except for precipitation. The reason is that previous studies have shown nontrivial biases in the GDAS precipitation (Gottschalck et al. 2005). Such a bias over land often leads to biases in many simulated land surface variables (e.g., soil moisture, soil temperature, snowpack, evapotranspiration, runoff, and surface latent and sensible heat fluxes).
To enhance the CFSR land surface simulation, two sets of observed global precipitation analyses are utilized as alternative forcing to drive GLDAS/LIS. One is the pentad data of CMAP that is also used in R2. CMAP defines 5-day mean precipitation at a 2.5° latitude–longitude grid over the globe by merging information derived from gauge observations as well as satellite observations in infrared and passive microwave channels. The other is the CPC unified global daily gauge analysis, constructed on a 0.5° latitude–longitude over the global land through the interpolation of quality controlled gauge reports from approximately 30 000 stations collected from the Global Telecommunication System (GTS) and many other national and international collections, using the optimal interpolation (OI) algorithm of Xie et al. (2007). To support CFSR and the anticipating operational CFS, the gauge analysis and CMAP have been processed from 1 January 1979 to real time.
After reviewing the global gauge distribution, a blending approach with latitude-dependent weighting masks is designed that favors the gauge analysis in midlatitudes and the satellite-dominated CMAP in tropical latitudes where gauge count is low. In high latitudes where gauge count is also low and satellite estimates lack accuracy, the GDAS-model-generated precipitation is given the most weight. In three high gauge density regions (contiguous United States, Europe, and Australia) the gauge analysis is given the most weight.
4. Land analysis execution
The land surface analysis GLDAS/LIS executes one 24-h cycle per production day at 0000 UTC. Each day after the completion of the GDAS cycles of the day, GLDAS/LIS starts the 24-h, 0000 to 0000 UTC, Noah LSM simulation driven by the GDAS atmospheric forcing and the blended precipitation forcing. It is worth noting that GDAS runs at a 3-min time step and produces hourly outputs as instantaneous or averaged values for the hour. GLDAS/LIS interpolates the hourly output to a 15-min Noah simulation time step. Such a time step mismatch can be critical to the Noah simulation—in particular, the snow accumulating and melting processes during the periods when the air temperature fluctuates frequently above and below freezing. After the Noah simulation, an independent snow analysis is applied.
Snow analysis in CFSR also executes once per 24 h using datasets from the Air Force Weather Agency’s global snow depth model (SNODEP; Kopp and Kiess 1996) and the National Environmental Satellite, Data, and Information Service (NESDIS) Interactive Multisensor Snow and Ice Mapping System (IMS; Helfrich et al. 2007). SNODEP uses in situ observations, a Special Sensor Microwave Imager (SSM/I)-based detection algorithm, and its own climatology to produce a global analysis of daily physical snow depth at 47-km resolution. SNODEP has been operational since 1975 to present. The IMS data is a manually generated Northern Hemisphere daily snow cover analysis. Analysts use in situ surface data, geostationary and polar orbiting imagery, and microwave-based detection algorithms to determine whether an area is either snow covered or snow free. IMS data is available at 23-km resolution starting February 1997 and at 4-km resolution starting February 2004.
The IMS and SNODEP data are used to produce daily analyses of physical snow depth on the model physics grid. The data are horizontally interpolated using a “budget” method (Accadia et al. 2003) in order to preserve the total water volume. In the Southern Hemisphere and globally prior to February 1997, these analyses are created solely from the SNODEP data. In the Northern Hemisphere starting February 1997, a combination of SNODEP and IMS is used. IMS data is introduced because it more accurately depicts snow cover compared to SNODEP especially along mountain ridges (because of its higher resolution). Therefore, in regions where the IMS and SNODEP analyses do not agree, the IMS determines whether or not there is snow in the daily analysis. More specifically, if the IMS indicates snow cover, the analyzed depth is set to 2.5 cm or the SNODEP value—whichever is greater. If IMS indicates a region is snow free, the analyzed depth is set to zero. The CFSR snow field is updated at 00Z by comparing the model first guess to the daily analysis. If the model snow depth is greater than twice (or less than half) the analyzed depth, then the model depth is set to twice (half) the analyzed value, and the model snow water equivalent (SWE) is set to maintain the same SWE/depth ratio as in the model first guess. Otherwise, the model snow is simply cycled. In contrast to directly replacing the model snow with the analysis, this method results in a smoother evolution of the snowpack and reduces the artificial addition of water when the land surface model erroneously melts the snow too quickly.
After the completion of the 24-h cycle of GLDAS/LIS execution and snow analysis, the simulated soil moisture and temperature of all four Noah LSM soil layers and the two analyzed snow variables (snow liquid equivalent and physical depth) are inserted into the end-of-cycle 0000 UTC CFSR restart file (the so-called “surface file”) as the land surface initial conditions for the following 0000 UTC CFSR cycle. The amounts of soil moisture and SWE adjustments are included in the CFSR final product so their contribution to the water budget can be traced.
The goal of CFSR is to generate a 31-yr (1979–2009) global reanalysis. To ensure the production completion on a manageable schedule, the 31-yr CFSR is produced by running six simultaneous streams of analyses, covering the following periods:
Stream 1: 1 December 1978–31 December 1986.
Stream 2: 1 November 1985–31 December 1989.
Stream 5: 1 January 1989–31 December 1994.
Stream 6: 1 January 1994–31 March 1999.
Stream 3: 1 April 1998–31 March 2005.
Stream 4: 1 April 2004–31 December 2009.
5. Results
Figures 1a–d shows the global monthly mean 2-m column volumetric soil moisture climatology averaged for May and November in the period 1980–2008, for CFSR/Noah and R2/OSU, respectively. The global patterns from CFSR and R2 are generally in agreement and consistent with our current knowledge about large-scale climatology of wet (Amazon, central Africa, high-latitudes Eurasia, and southern Asia) and dry (Sahara, Mideast, and Australia) regions. Soil moisture in the permanent frozen glacial land of Greenland and Antarctic is not simulated in CFSR, and hence masked out in the figures. Users of CFSR product would find soil moisture values set as a flagged value of 1.0 in glacial areas. Comparing to R2, CFSR is relatively wetter in India, high-latitudes Eurasia, and most of the contiguous United States (CONUS), and, on the other hand, is relatively drier in Australia and central Africa. The difference in soil moisture climatology of November minus May, shown in Fig. 1e and Fig. 1f, respectively, depicts the seasonal variation corresponding to the precipitating and drying periods of various regions. In midlatitudes Northern Hemisphere, soil moisture content reaches the peak in early spring reflecting the melting of the wintertime snow accumulation. Vegetated areas start to lose soil moisture via evapotranspiration through summer and fall. In the tropics and subtropics, higher seasonal variations are found in the monsoon active regions of Central America, northern South America, the west coast of central Africa, India, and South Asia. CFSR and R2 agree well in most regions except for the regions of the U.S. Rocky Mountains, the central Africa savanna, and the east coast of India where CFSR shows a relatively weaker seasonal variation. Area-averaged seasonal cycles of monthly mean soil moisture climatology for selected regions of interests (as illustrated by red rectangles in Fig. 1e) from the two datasets are compared in Fig. 2. In the U.S. Rocky Mountain region, CFSR shows a relatively weaker seasonal variation owing to the addition of frozen soil physics from R2/OSU to CFSR/Noah that cause the bottom layer (100–200-cm depth below surface) of CFSR/Noah in this high elevation region to be nearly frozen and inactive for most of the year. If only comparing the soil moisture content from the top layer (0–10 cm), seasonal cycles of the two datasets are in better agreement (not shown). In the central Africa savanna region, CFSR also shows a relatively weaker seasonal variation since CFSR LIS is forced with a near-uniform amount of precipitation observed from April through October in this region, while R2 carries a more intensive precipitating season during the period from July to September. In India, the two seasonal cycles are very similar in both phase and amplitude except for that CFSR has higher soil moisture values in all seasons owing to LSM model preferences between R2/OSU and CFSR/Noah.
Monthly mean 2-m column volumetric soil moisture (m3 m−3) climatology (1980–2008) for (a) May CFSR, (b) May R2, (c) November CFSR, (d) November R2, (e) November–May difference of CFSR, and (f) November–May difference of R2. The glacial points for CFSR are masked out.
Citation: Journal of Hydrometeorology 13, 5; 10.1175/JHM-D-11-090.1
Seasonal cycle of monthly mean 2-m column volumetric soil moisture (m3 m−3) climatology (1980–2008) averaged for the regions of (a) the U.S. Rocky Mountains, (b) central Africa, and (c) India for CFSR (solid line) and R2 (dashed line). Regions are defined in red rectangles in Fig. 1e.
Citation: Journal of Hydrometeorology 13, 5; 10.1175/JHM-D-11-090.1
To evaluate its performance, the CFSR soil moisture is compared with the soil moisture from the North American Land Data Assimilation System phase 2 product (NLDAS, Xia et al. 2012). NLDAS is a multi-institution, multimodel product with the goal of constructing quality-controlled and spatially and temporally consistent land surface datasets from the best available observations and models output to support climatological and hydrological modeling activities. The NLDAS product covers the CONUS domain for the 30-plus-year period from 1979 to present. A 15-yr spinup has been applied in the NLDAS execution prior to its production mode. NLDAS includes four LSMs and in this paper the NLDAS/Noah product is used. Figure 3 shows the time series of monthly mean 2-m soil moisture averaged over CONUS for CFSR/Noah, R2/OSU, and NLDAS/Noah. The vertical lines in Fig. 3 indicate the beginning of each CFSR stream. The temporal anomaly correlation between CFSR and NLDAS is 0.79 and that between R2 and NLDAS is 0.69. Figure 4 shows the time series of monthly mean 2-m soil moisture averaged over the state of Illinois for CFSR, R2, and NLDAS compared with the commonly used Illinois soil moisture observation data product (Hollinger and Isard 1994). The temporal anomaly correlation between CFSR and observation is 0.61 and that between R2 and observation is 0.48. Wang et al. (2011) analyze the CFSR soil moisture time series and are concerned by the possible discontinuities between stream boundaries and possible trends within the streams. It is worth noting that both R2 and NLDAS are generated as single stream processes without any discontinuity. From Figs. 3 and 4, there is no clear evidence showing discontinuities and trends in the CFSR soil moisture time series. Further analysis is ongoing to investigate these issues.
Monthly mean 2-m column volumetric soil moisture temporal anomaly time series averaged over CONUS for CFSR (green), R2 (brown), and NLDAS (purple). Numbers indicate the anomaly correlation between reanalysis products and NLDAS. Vertical lines indicate the beginning of each CFSR stream.
Citation: Journal of Hydrometeorology 13, 5; 10.1175/JHM-D-11-090.1
Monthly mean 2-m column volumetric soil moisture temporal anomaly time series averaged over the state of Illinois for CFSR (green), R2 (brown), NLDAS (purple), and observation (black). Numbers indicate the temporal anomaly correlation between reanalysis products and observation. Vertical lines indicate the beginning of each CFSR stream.
Citation: Journal of Hydrometeorology 13, 5; 10.1175/JHM-D-11-090.1
CFSR January 1990 monthly mean (a) latent heat flux, (b) sensible heat flux, (c) snow phase change energy, (d) energy residual, and monthly accumulated (e) snow water equivalent update, (f) soil moisture update, (g) precipitation, and (h) water residual. The units for (a)–(d) are W m−2, for (e)–(g) are mm month−1, and for (h) is percentage.
Citation: Journal of Hydrometeorology 13, 5; 10.1175/JHM-D-11-090.1
6. Summary
The NCEP CFSR provides an updated global reanalysis from 1979 to 2009. It is the first time that a GLDAS strategy is employed in a coupled land–atmosphere–ocean–sea ice global reanalysis system to perform the land surface analysis. Observed global precipitation and snow fields are used to constrain the land surface simulation of soil moisture, soil temperature, and snowpack. The resulting global soil moisture field is reasonable and the land surface energy and water balance are well maintained. The CFSR product can improve the current understanding of the global land surface energy and water cycles in the earth system to further assist the research in hydrology, weather, and climate. This paper serves the purpose of providing users with documentation and detailed description of the CFSR land surface analysis component, beyond what is introduced in the CFSR overview paper of Saha et al. (2010). Also provided is the essential information for correctly collecting, from the CFSR product, all the variables required to study the CFSR land surface energy and water budgets. For more in-depth discussion of the CFSR land surface product, further evaluation on the CFSR soil moisture climatology and climatological variability and the land surface energy and water budgets are in preparation.
Acknowledgments
This work is supported by NCEP for the Climate Forecast System Reanalysis and NOAA Climate Program Office for the NCEP Core Project for the Climate Prediction Program for the Americas (CPPA). We thank Paul Dirmeyer, Jordan Alpert, Xingren Wu, and anonymous reviewers for review and comments.
REFERENCES
Accadia, C., Mariani S. , Casaioli M. , and Lavagnini A. , 2003: Sensitivity of precipitation forecast skill scores to bilinear interpolation and a simple nearest-neighbor average method on high-resolution verification grids. Wea. Forecasting, 18, 918–932.
Berg, A. A., Famiglietti J. S. , Walker J. , and Houser P. R. , 2003: Impact of bias correction to reanalysis products on simulation of North American soil moisture and hydrological fluxes. J. Geophys. Res., 108, 4490, doi:10.1029/2002JD003334.
Chen, F., 2005: Variability in global land surface energy budgets during 1987–1988 simulated by an off-line land surface model. Climate Dyn., 25, 667–684.
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.
Decharme, B., and Douville H. , 2006: Uncertainties in the GSWP-2 precipitation forcing and their impacts on regional and global hydrological simulations. Climate Dyn., 27, 695–713.
Dickinson, R. E., Shaick M. , Bryant R. , and Graumlich L. , 1998: Interactive canopies for a climate model. J. Climate, 11, 2823–2836.
Dirmeyer, P. A., 2001: Climate drift in a coupled land–atmosphere model. J. Hydrometeor., 2, 89–100.
Dirmeyer, P. A., and Zhao M. , 2004: Flux replacement as a method to diagnose coupled land–atmosphere model feedback. J. Hydrometeor., 5, 1034–1048.
Dirmeyer, P. A., Gao X. , Zhao M. , Guo Z. , Oki T. , and Hanasaki N. , 2006: GSWP2: Multimodel analysis and implications for our perception of the land surface. Bull. Amer. Meteor. Soc., 87, 1381–1397.
Ek, M., Mitchell K. E. , Lin Y. , Rogers E. , Grunmann P. , Koren V. , Gayno G. , and Tarpley J. D. , 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.
Gottschalck, J., Meng J. , Rodell M. , and Houser P. , 2005: Analysis of multiple precipitation products and preliminary assessment of their impact on Global Land Data Assimilation System land surface states. J. Hydrometeor., 6, 573–598.
Guo, Z., Dirmeyer P. A. , Hu Z.-Z. , Gao X. , and Zhao M. , 2006: Evaluation of the Second Global Soil Wetness Project soil moisture simulations: 2. Sensitivity to external meteorological forcing. J. Geophys. Res., 111, D22S03, doi:10.1029/2006JD007845.
Hansen, M. C., DeFries R. S. , Townshend J. R. G. , and Sohlberg R. , 2000: Global land cover classification at 1 km spatial resolution using a classification tree approach. Int. J. Remote Sens., 21, 1331–1364.
Helfrich, S. R., McNamara D. , Ramsay B. H. , Baldwin T. , and Kasheta T. , 2007: Enhancements to, and forthcoming developments in the Interactive Multisensor Snow and Ice Mapping System (IMS). Hydrol. Processes, 21, 1576–1586, doi:10.1002/hyp.6720.
Henderson-Sellers, A., Pitman A. J. , Love P. K. , Irannejad P. , and Chen T. H. , 1995: The Project for Intercomparison of Land Surface Parameterization Schemes (PILPS): Phases 2 and 3. Bull. Amer. Meteor. Soc., 76, 489–503.
Hollinger, S. E., and Isard S. A. , 1994: A soil moisture climatology of Illinois. J. Climate, 7, 822–833.
Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77, 437–471.
Kanamitsu, M., Ebisuzaki W. , Woollen J. , Yang S. K. , Hnilo J. J. , Fiorino M. , and Potter G. L. , 2002: NCEP–DOE AMIP-II Reanalysis (R-2). Bull. Amer. Meteor. Soc., 83, 1631–1643.
Kopp, T. J., and Kiess R. B. , 1996: The Air Force Global Weather Central snow analysis model. Preprints, 15th Conf. on Weather Analysis and Forecasting, Norfolk, VA, Amer. Meteor. Soc., 220–222.
Koster, R., Suarez M. , and Heiser M. , 2000: Variance and predictability of precipitation at seasonal-to-interannual timescales. J. Hydrometeor., 1, 26–46.
Koster, R., and Coauthors, 2004: Regions of strong coupling between soil moisture and precipitation. Science, 305, 1138–1140, doi:10.1126/science.1100217.
Lu, C., Kanamitsu M. , Roads J. , Ebisuzaki W. , Mitchell K. , and Lohmann D. , 2005: Evaluation of soil moisture in the NCEP–NCAR and NCEP–DOE global reanalyses. J. Hydrometeor., 6, 391–408.
Mahrt, L., and Pan H.-L. , 1984: A two-layer model of soil hydrology. Bound.-Layer Meteor., 29, 1–20.
Materia, S., Dirmeyer P. A. , Guo Z. , Alessandri A. , and Navarra A. , 2010: The sensitivity of simulated river discharge to land surface representation and meteorological forcings. J. Hydrometeor., 11, 334–351.
Mintz, Y., and Serafini Y. , 1992: A global monthly climatology of soil moisture and water balance. Climate Dyn., 8, 13–27.
Mitchell, K. E., 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.
Oki, T., Nishimura T. , and Dirmeyer P. , 1999: Assessment of annual runoff from land surface models using Total Runoff Integrating Pathways (TRIP). J. Meteor. Soc. Japan, 77, 235–255.
Oleson, K. W., and Coauthors, 2008: Improvements to the Community Land Model and their impact on the hydrological cycle. J. Geophys. Res., 113, G01021, doi:10.1029/2007JG000563.
Pan, H.-L., and Mahrt L. , 1987: Interaction between soil hydrology and boundary-layer development. Bound.-Layer Meteor., 38, 185–220.
Peters-Lidard, C. D., and Coauthors, 2007: High performance earth system modeling with NASA/GSFC’s Land Information System. Innovations Syst. Software Eng., 3, 157–165, doi:10.1007/s11334-007-0028-x.
Rodell, M., and Coauthors, 2004: The Global Land Data Assimilation System. Bull. Amer. Meteor. Soc., 85, 381–394.
Saha, S., and Coauthors, 2010: The NCEP Climate Forecast System Reanalysis. Bull. Amer. Meteor. Soc., 91, 1015–1057.
Wang, W., Xie P. , Yoo S.-H. , Xue Y. , Kumar A. , and Wu X. , 2011: An assessment of the surface climate in the NCEP climate forecast system reanalysis. Climate Dyn., 37, 1601–1620, doi:10.1007/s00382-010-0935-7.
Wood, E. F., Lettenmaier D. P. , Liang X. , Nijssen B. , and Wetzel S. W. , 1997: Hydrological modeling of continental-scale basins. Annu. Rev. Earth Planet. Sci., 25, 279–300.
Xia, Y., and Coauthors, 2012: Continental-scale water and energy flux analysis and validation for the North American Land Data Assimilation System project phase 2 (NLDAS-2): 1. Intercomparison and application of model products. J. Geophys. Res., 117, D03110, doi:10.1029/2011JD016048.
Xie, P., and Arkin P. A. , 1997: Global precipitation: A 17-year monthly analysis based on gauge observations, satellite estimates, and numerical model outputs. Bull. Amer. Meteor. Soc., 78, 2539–2558.
Xie, P., Chen M. , Yatagai A. , Hayasaka T. , Fukushima Y. , and Yang S. , 2007: A gauge-based analysis of daily precipitation over East Asia. J. Hydrometeor., 8, 607–626.
Xue, Y., Zeng F. J. , Mitchell K. E. , Janjic Z. , and Rogers E. , 2001: The impact of land surface processes on simulations of the U.S. hydrological cycle: A case study of the 1993 flood using the SSiB land surface model in the NCEP Eta regional model. Mon. Wea. Rev., 129, 2833–2860.
Zobler, L., 1986: A world soil file for global climate modeling. NASA Tech. Memo. 87802, 32 pp.
