ArticlePDF Available

Regional and Global Land Data Assimilation Systems: Innovations, Challenges, and Prospects

Authors:
  • SAIC at NCEP/EMC

Abstract and Figures

Since the North American and Global Land Data Assimilation Systems (NLDAS and GLDAS) were established in 2004, the development of regional and global LDASs has made significant progress. National, regional, project-based, and global LDASs are widely developed worldwide. This study summarizes and overviews the development, current status, challenges, and future prospects of these LDASs. It should be noted that this review focuses on un-coupled LDASs without specific comparison and analysis about various LDASs' performance (e.g., strengths and weaknesses). We first introduce regional and global LDASs, including their history and development, and then discuss the evaluation, validation, and application (from numerical model prediction to water resources management) of various LDASs. More importantly, we detail the challenges of LDASs including but not limited to (1) the quality of in-situ observations, satellite retrievals, reanalysis data, and soil and vegetation datasets, and (2) land surface model physical processes and parameters, single criterion and multi-criteria calibration for small watersheds and related re-gionalization, land data assimilation, and spatial incomparability problems. Finally, some prospects such as the use of land information system software, unified global LDAS system with nesting concept and hyper-resolution, and uncertainty estimates are discussed.
Content may be subject to copyright.
Regional and Global Land Data Assimilation Systems: Innovations,
Challenges, and Prospects
Youlong XIA1, Zengchao HAO2*, Chunxiang SHI3, Yaohui LI4, Jesse MENG1, Tongren XU5,
Xinying WU2, and Baoqing ZHANG6
1 I.M. Systems Group at Environmental Modeling Center (EMC), National Centers for Environmental Prediction (NCEP),
National Oceanic and Atmospheric Administration (NOAA), College Park, MD 20740, USA
2 College of Water Sciences, Beijing Normal University, Beijing 100875, China
3 National Meteorological Information Center, China Meteorological Administration, Beijing 100081, China
4 Institute of Arid Meteorology, China Meteorological Administration, Lanzhou 730020, China
5 State Key Laboratory of Earth Surface Processes and Resource Ecology, School of Natural Resources,
Faculty of Geographical Science, Beijing Normal University, Beijing 100875, China
6 Key Laboratory of Western China’s Environmental Systems of Ministry of Education, College of Earth
and Environmental Sciences, Lanzhou University, Lanzhou 730000, China
(Received November 9, 2018; in final form February 26, 2019)
ABSTRACT
Since the North American and Global Land Data Assimilation Systems (NLDAS and GLDAS) were established in
2004, the development of regional and global LDASs has made significant progress. National, regional, project-
based, and global LDASs are widely developed worldwide. This study summarizes and overviews the development,
current status, challenges, and future prospects of these LDASs. It should be noted that this review focuses on un-
coupled LDASs without specific comparison and analysis about various LDASs’ performance (e.g., strengths and
weaknesses). We first introduce regional and global LDASs, including their history and development, and then dis-
cuss the evaluation, validation, and application (from numerical model prediction to water resources management) of
various LDASs. More importantly, we detail the challenges of LDASs including but not limited to (1) the quality of
in-situ observations, satellite retrievals, reanalysis data, and soil and vegetation datasets, and (2) land surface model
physical processes and parameters, single criterion and multi-criteria calibration for small watersheds and related re-
gionalization, land data assimilation, and spatial incomparability problems. Finally, some prospects such as the use of
land information system software, unified global LDAS system with nesting concept and hyper-resolution, and uncer-
tainty estimates are discussed.
Key words: land data assimilation system (LDAS), regional and global LDASs, in-situ observation, satellite retrie-
val, land surface model
Citation: Xia, Y. L., Z. C. Hao, C. X. Shi, et al., 2019: Regional and global land data assimilation systems: Innova-
tions, challenges, and prospects. J. Meteor. Res., 33(2), 1–31, doi: 10.1007/s13351-019-8172-4.
1. Introduction
The accuracy of a land surface model (LSM) simula-
tion is largely affected by many errors that exist in sur-
face meteorological forcing data (to drive the model),
model parameters (e.g., soil and hydrologic parameters),
and model structures (e.g., specific soil layer vs. varied
soil layers). Therefore, it is a challenging task for an
LSM to produce relatively accurate initial conditions for
numerical weather and climate prediction and to simu-
late reasonable energy budget, water cycle, and biogeo-
chemical cycle. With the increase of in-situ observations
and remote sensing data, the accuracy of these products
can be improved by assimilating these data into the LSM
(Kumar et al., 2014). Based on this concept, the land data
assimilation system (LDAS) was developed by integrat-
Supported by the US Environmental Modeling Center (EMC) Land Surface Modeling Project (granted to Youlong Xia) and National
Natural Science Foundation of China (51609111, granted to Baoqing Zhang).
*Corresponding author: haozc@bnu.edu.cn.
©The Chinese Meteorological Society and Springer-Verlag Berlin Heidelberg 2019
Volume 33
Special Issue on Development and Applications of Regional and Global
Land Data Assimilation Systems APRIL 2019
ing the advanced LSM with data assimilation (DA) tech-
niques. The LDAS uses gauge-based and remotely-
sensed observations and/or reanalysis land surface met-
eorological forcing data to drive the LSM, and uses DA
techniques to ingest ground-measured or satellite-based
data into the LSM to generate the merging products.
In the beginning, the North American and Global
LDASs were mainly used to provide optimal initial states
and fluxes to parent atmospheric models to enhance
model prediction skill in coupled numerical weather/cli-
mate systems (Mitchell et al., 2004; Rodell et al., 2004).
With the development and expansion of various LDASs,
more useful products have been generated for a variety of
applications, including drought and flood monitoring, ag-
ricultural crop management and planning, wildfire early
warning, and water resources management. Based on the
development, the LDAS can be divided into an un-
coupled and a weakly coupled LDAS system. Based on
the size of the spatial domain, it can be classified as a na-
tional, regional, and global system. The uncoupled
LDAS uses given meteorological forcing data as well as
soil and vegetation parameters to drive LSMs to obtain
surface energy fluxes (e.g., sensible heat flux, latent heat
flux, net radiation, ground heat flux, etc.), water fluxes
(e.g., evaporation, transpiration, surface runoff and base-
flow, and sublimation), and land state variables (soil
moisture, soil temperature, snow water equivalent, can-
opy water storage, groundwater, and surface temperat-
ure). For the coupled LDAS, the weakly coupled system
is run to generate forecast forcing data first. Then an off-
line (uncoupled to the atmosphere) land surface model is
re-run with forecast forcing data (only the forecast pre-
cipitation is replaced with observations) to produce soil
moisture and temperature. Finally, the soil moisture and
temperature generated from the offline LSM are used as
initial and boundary conditions to re-run the weakly
coupled system.
There are many uncoupled regional systems, such as
the European LDAS (Jacobs et al., 2008), North Ameri-
can LDAS (NLDAS; Mitchell et al., 2004; Xia et al.,
2012a, b), South American LDAS (de Goncalves et al.,
2006), Central Asia and East/Southern/West Africa LDAS
(McNally et al., 2017), and so on. Besides the regional
systems, a number of uncoupled national LDASs also ex-
ist, such as the Canadian LDAS (CaLDAS; Carrera et al.,
2015), China LDAS (CLDAS; Li et al., 2007), CMA
LDAS (Shi et al., 2011), and Korean LDAS (Lim et al.,
2012); and on top of these is the global LDAS (GLDAS;
Rodell et al., 2004). Two examples of the weakly
coupled LDAS are the NCEP GLDAS (Meng et al.,
2012), the ECMWF LDAS (de Rosnay et al., 2014; de
Rosnay, 2017; Albergel et al., 2018). The regional and
national uncoupled LDASs are widely and rapidly de-
veloped as the surface meteorological forcing data, such
as gauge-observed precipitation, air temperature, and ra-
diation, can be more easily and accurately obtained from
the national and regional domains than on a global scale
(e.g., difficulty in sharing data between various coun-
tries). Furthermore, the development of regional and na-
tional LDASs has been effectively accelerated due to
more affordable computational resources for higher spa-
tial resolution LDASs at regional/national scales, more
homogeneous datasets of climate, hydrology, vegetation,
and soil characteristics, increased available high-quality
observations, as well as more governmental funding sup-
port from various countries. The advance of project-
based LDASs depends heavily on their funding agencies
and resources. However, they contribute to the develop-
ment of the regional, national, and global LDASs.
In the early stage, all LDAS systems do not assimilate
any satellite information to improve the state variables
due to the lack and low quality of satellite-retrieved data.
With the increasing availability of ground-based and re-
motely-sensed data (e.g., satellite, radar) and advances in
DA techniques, LDAS is becoming an efficient tool to
improving land surface model simulations and products,
and to enhancing numerical forecast skill at various
timescales (daily, weekly, sub-seasonal, and seasonal). In
this study, we overview various systems including na-
tional, regional, project-based, and global LDASs and
their initiatives, developments, and current status. The
application of their products in weather forecast and sub-
seasonal-to-seasonal (S2S) prediction (e.g., drought and
flood monitoring, agricultural crop management, wild
fire monitoring, and water resources management) is also
introduced. This review covers general topics on un-
coupled LDASs, without specific comparison and analy-
sis about various LDASs’ performance (e.g., strengths
and weaknesses). Detailed comparison and analysis of
the shortcomings and strengths of various LDASs re-
quire in-depth knowledge of the forcing used (e.g., reana-
lysis, observations, satellites, etc.), different LSMs and
physical parameterizations, and DA algorithms and the
variables assimilated. They are beyond the scope of the
present paper, but will be performed by the ongoing
works and the LDAS collaborators and developers in the
future. In this paper, Sections 2 and 3 summarize region-
al, national, project-based, and global LDAS systems, re-
spectively. Section 4 illustrates how to comprehensively
evaluate LDAS products, in particular, using NLDAS as
2Journal of Meteorological Research Volume 33
an example. The applications of LDAS products for nu-
merical weather and seasonal climate models, drought
monitoring, and agricultural crop and water resource
management are discussed in Section 5. The challenges,
discussions, and conclusions are presented in Sections 6,
7, and 8, respectively.
2. Regional, national, and project-based
LDASs
The LDASs can be divided into various categories ac-
cording to different classification standards, such as the
research domain size (e.g., national, regional, and
global), spatial resolution (high vs. coarse resolution),
application purpose (e.g., operational, research-based,
project-based, etc.), and the coupling methodology
between the land surface model and its atmospheric mod-
el (e.g., uncoupled and weakly coupled; see Fig. 1). The
standard definition of the strength of coupling in DA is
discussed in Penny et al. (2017) and Penny and Hamill
(2017). Weakly coupling represents that the assimilation
is performed for each of the components of the coupled
model independently (e.g., land DA). The interaction
between the various components (i.e., atmosphere, ocean,
land, and sea ice) is provided by the coupled forecast sys-
tem. The direct impacts of observations on the analysis
are limited to the region where the observations exist.
Cross-region impacts are produced as a secondary effect
by the integration of the coupled model forecast. Here,
we mainly discuss regional, national, and project-based
LDASs regardless of resolution, application, and coup-
ling methodology (Tables 1a, 1b).
The domain of a regional LDAS can extend over sev-
eral countries. The NLDAS, which includes the contin-
ental United States, southern Canada, and northern Mex-
ico, was initiated in 1998 as a multi-institution joint
project between several government agencies and uni-
versities (Mitchell et al., 1999) and is one of the most
successful uncoupled LDASs. After a 5-yr collaboration
and effort, the NLDAS configuration was established in
2004 (Mitchell et al., 2004). Some overview papers sum-
marized the generation and validation of NLDAS for-
cing (Cosgrove et al., 2003a; Luo et al., 2003), produc-
tion and validation of NLDAS streamflow and water
budget (Lohmann et al., 2004), evaluation of energy
budget, soil moisture, temperature, and snowpack (Pan et
al., 2003; Robock et al., 2003; Sheffield et al., 2003;
Schaake et al., 2004), and model spin-up testing (Cos-
grove et al., 2003b). Four land surface models are ex-
ecuted in uncoupled mode in NLDAS, including the
Noah (Ek et al., 2003), Mosaic (Koster and Saurez,
1994), Sacramento Soil Moisture Accounting (SAC-
SMA; Burnash et al., 1973), and Variable Infiltration Ca-
pacity (VIC; Liang et al., 1994). These models are run at
a 0.125° spatial resolution and 1-h temporal resolution to
produce water fluxes, energy fluxes, and state variables.
The purpose of the first phase of NLDAS (NLDAS-1)
is to establish an NLDAS framework, and thus it was run
for only a 3-yr period from 1 October 1996 to 30 Septe-
mber 1999. Although such a short-term run may be suffi-
cient for evaluating forcing and model outputs, it is in-
sufficient for practical applications such as drought mon-
itoring, long-term trend analysis, and crop and water re-
sources management. To overcome this weakness, the
second phase of NLDAS (NLDAS-2; Xia et al., 2012a,
b) was developed, based on the NLDAS-1 framework
and was improved with a long-term run period from 1
January 1979 to present. The precipitation forcing de-
rived from a daily gauge-only precipitation analysis is
bias-corrected for the impacts of topography on precipit-
ation using a Parameter-elevation Regression on Inde-
(a) Uncoupled LDAS
LDAS LDAS LDAS
GCM
LSM
GCM
LSM
Atmosphere
DA
Observations Error
stats
LSM Land
DA
Observations
(b) Weakly coupled LDAS
Time stepTime stepTime step
321 Error
stats
LDAS LDAS LDAS
GCM
LSM
Some of forecast variables are
replaced by the observed
GCM
LSM
Atmosphere
DA
Observations Error
stats
LSM Land
DA
Observations
Time stepTime stepTime step
321 Error
stats
Fig. 1. Schematic diagrams describing the interaction between an
LDAS and a coupled GCM model: (a) uncoupled LDAS and (b)
weakly coupled LDAS. The LDAS uses some DA techniques (e.g.,
simple insertion, ensemble Kalman filter/smoother, etc.) to integrate
past forecasts with observations to improve numerical weather and cli-
mate prediction skill (GCM: General Circulation Model; DA: Data As-
similation; LDAS: Land DA System; LSM: Land Surface Model).
APRIL 2019 Xia, Y. L., Z. C. Hao, C. X. Shi, et al. 3
pendent Slopes Model (PRISM). This daily precipitation
analysis is temporally divided to hourly precipitation us-
ing the NCEP Stage II hourly radar/gauge precipitation.
The NLDAS-2 non-precipitation forcing is generated
from the North American Regional Reanalysis (NARR;
Mesinger et al., 2006). The NARR downward shortwave
radiation is corrected by using a ratio-based bias correc-
tion technique (Berg et al., 2003) and the University of
Maryland’s Surface Radiation Budget (SRB) dataset
(Pinker et al., 2003). This is the classical uncoupled
LDAS concept (Figure 1a) where land surface models
are run using a given surface meteorological forcing. In-
spired by the development of NLDAS, many regional
LDASs, such as the European, South America, Central
Asia, and East/Southern/West Africa LDAS, were de-
veloped in the same way. More details about the devel-
opment and applications of NLDAS can be seen in EMC
website (http://www.emc.ncep.noaa.gov/mmb/nldas/) and
NASA website (https://ldas.gsfc.nasa.gov/nldas).
Examples of national LDASs include the CaLDAS,
CLDAS, and China Meteorological Administration
(CMA) LDAS (Table 1). The CaLDAS is an uncoupled
operational LDAS with an external land surface model-
ing system, which provides optimal initial conditions to
its parent atmospheric model. The CaLDAS uses the en-
semble Kalman filter (EnKF) to assimilate the observed
Table 1a. Summary of various LDAS systems and their products over the world (CaLDAS: Canadian LDAS; CLDAS: China LDAS; CMA:
China Meteorological Administration; CONUS: Continental United States; ELDAS: European LDAS; ERA-Interim: ECMWF Re-Analysis In-
terim data; ERA-5: The fifth generation ECMWF Re-Analysis; FLDAS: Famine Early Warning Systems Network LDAS; GLDAS: Global
LDAS; HRLDAS: High-Resolution LDAS; IFS: Integrated Forecast System; NASA: National Aeronautics and Space Administration; NCA: Na-
tional Climate Assessment; NLDAS: North American LDAS; SLDAS: South American LDAS)
Name Domain, resolution Data period Reference Website
Regional LDAS systems
ELDAS Europe, 0.2° 1 May–31 October 2000 Jacobs et al., 2008 N/A
NLDAS CONUS, southern Canada,
 northern Mexico, 0.125°
1 January 1979–present Mitchell et al., 2004;
 Xia et al., 2012a, b
EMC:
http://www.emc.ncep.noaa.gov/
 mmb/nldas/
NASA:
https://ldas.gsfc.nasa.gov/nldas/
SALDAS South America, 40 km 17 March 2001–16 March
2002
de Goncalves et al., 2006 N/A
National LDAS systems
CaLDAS Canada, 1 km, 40 km 1 June–30 September 2009 Carrera et al., 2015 N/A
CLDAS China, 0.25° 1 October 2002–31
 December 2006
Li et al., 2007 N/A
CMA LDAS China, 0.06° 1 January 2008–present Shi et al., 2011 N/A
Korean LDAS East Asia, 10 km 1 August 2004–12
 December 2006
Lim et al., 2012 N/A
Project-based LDAS systems
NASA FLDAS Central Asia (1 km) and
 East/Southern/West
 Africa, 0.1°~0.25°
1 January 1981–present McNally et al., 2017 https://lis.gsfc.nasa.gov/projects/
 fewsnet
NCA LDAS NLDAS domain, 0.125° 1 January 1979–31
 December 2018
Kumar et al., 2018 https://ldas.gsfc.nasa.gov/
 NCA-LDAS/
NCAR HRLDAS Central US, 4 km, 12 km 1 January 2001–1 July 2002 Chen et al., 2007 https://ral.ucar.edu/solutions/
 products/high-resolution-land-
 data-assimilation-system-hrldas
Global LDAS systems
NASA/GLDAS Globe, 0.25°, 1.0° 1 January 1979–present for
 GLDAS-1; 1 January
 1948–31 December 2008
 for GLDAS-2
Rodell et al., 2004 https://ldas.gsfc.nasa.gov/gldas/
NCEP/GLDAS Globe, ~38 km 1 January 1979–present Meng et al., 2012 https://climatedataguide.ucar.edu/
 climate-data/climate-forecast-
 system-reanalysis-cfsr
ECMWF /GLDAS Globe, 9 km, 32 km,
 79 km
Real-time for 9-km
 operational IFS; 1 January
 1950–present for 32 km
 ERA-5, 1 January
 1979–30 November 2018
 for 79-km ERA-Interim
de Rosnay et al., 2014;
 de Rosnay, 2017
https://software.ecmwf.int/wiki/
 display/LDAS/LDAS+Home
Note: Many LDASs used in the NWP systems of various operational centers and research institutes are not included in this table. Details of these
systems are provided in Section 4, as well as relevant references.
4Journal of Meteorological Research Volume 33
Table 1b. Summary of various LDAS systems over the world for their forcings, coupling ways, and application purposes (DLWR: Downward
Longwave Radiation; DSWR: Downward Shortwave Radiation; CLM2: Community Land Model version 2; CLM3.5: Community Land Model
version 3.5; CLSM: Catchment Land Surface Model; CMAP: CPC Merged Analysis of Precipitation; CMORPH: CPC MORPHing technique;
CoLM: Common Land Model; CPC: Climate Prediction Center; EDAS: Eta Data Assimilation System; EnKF: Ensemble Kalman Filter; ERA-
40: 40-yr ECMWF Re-Analysis; GDAPS: Global Data Assimilation and Prediction System; GDAS: Global Data Assimilation System; GFS:
Global Forecast System; GLDAS-1: Global LDAS phase 1; GLDAS-2: Global LDAS phase 2; GOES: Geostationary Operational Environment-
al Satellite system; ISBA: Interactions between Soil–Biosphere–Atmosphere; MERRA-2: Modern-Era Retrospective analysis for Research and
Applications Version 2; Mosaic: NASA Land Surface Model; NARR: North American Regional Reanalysis; Noah: NCEP Land Surface Model
(Noah2.5.2, Noah2.7.1, Noah2.8, and Noah3.3 indicate different versions); Noah-MP: Noah Multiphysics; P: precipitation; ps: surface pressure;
q2m: 2-m specific humidity; RH2m: 2-m relative humidity; SAC: Sacramento Soil Moisture Accounting Model; SIB2: Simple Biosphere model
version 2; SSiB: Simplified Simple Biosphere model; T2m: 2-m air temperature; TESSEL: Tiled ECMWF Scheme for Surface Exchanges over
Land; H-TESSEL: TESSEL with a revised land surface hydrology); U10m; 10-m x-direction wind speed; V10m: 10-m y-direction wind speed; VIC:
Variable Infiltration Capacity, VIC4.0.3, VIC4.0.4, and VIC4.1.2 indicate different versions)
Name Meteorological forcing Coupling strength
and DA method Land surafce model Application
purpose
Regional LDAS systems
ELDAS ERA-40 reanalysis (P, DSWR, DLWR, T2m,
 q2m, U10m, V10m, ps)
Uncoupled, no DA TESSEL Research
NLDAS NARR reanalysis (DLWR, T2m, q2m, U10m,
 V10m, ps), CPC gauge-based P, bias
 corrected DSWR
Uncoupled, no DA Noah2.8, Mosaic,
 SAC, VIC4.0.3
Operation
SALDAS NCEP GDAS (P, DSWR, DLWR, T2m, q2m,
 U10m, V10m, ps)
Uncoupled, no DA SSiB Research
National LDAS systems
CaLDAS Canadian precipitation analysis with
 observations, NWP model DSWR, DLWR,
 T2m, q2m, U10m, V10m, ps
Weakly; assimilate T2m, RH2m, snow
 depth and L-band brightness
 temperature with EnKF, 6-hourly
 assimilation cycle
ISBA Operation
China LDAS Observed P, DSWR, T2m, NASA GLDAS
 (DLWR, q2m, U10m, V10m, ps)
Uncoupled; assimilate soil moisture
 and snow with EnKF
CoLM, SiB2 Research
CMA LDAS P blended with gauge observations and
 CMORPH, analyzed T2m from stations and
 GFS product, Satellite-retrieved DSWR,
 and GFS DLWR, q2m, U10m, V10m, ps
Uncoupled, no DA CoLM, Noah-MP,
 CLM3.5
Operation
Korean LDAS Observed P and DSWR, Korean GDAPS
 (DLWR, T2m, q2m, U10m, V10m, ps)
Uncoupled, no DA Noah2.5.2 Research
Project-based LDAS systems
NASA FLDAS NCEP GDAS and NASA MERRA-2
 reanalysis (DSWR, DLWR, T2m, q2m,
 U10m, V10m, ps), P blended with gauge
 observations and satellite retrievals
Uncoupled, no DA Noah3.3, VIC4.1.2 Research
NCA LDAS NARR reanalysis (DLWR, T2m, q2m, U10m,
 V10m, ps), CPC gauge-based P, bias
 corrected DSWR
Uncoupled, no DA Research
NCAR HRLDAS Bias-corrected radar precipitation, GOES
 radiation, and EDAS reanalysis
Uncoupled, no DA Noah2.7.1 Research
Global LDAS systems
NASA/GLDAS NCEP GDAS (DLWR, T2m, q2m, U10m, V10m,
 ps), P from disaggregated CMAP, and
 satellite-retrieved DSWR and DLWR
 (the Air Force Weather Agency) for
 GLDAS-1; bias-corrected reanalysis (P,
 DSWR, DLWR, T2m, q2m, U10m,
 V10m, ps) generated from Princeton
 University (Sheffield et al., 2006)
Uncoupled, no DA CLM2, Mosaic,
 Noah2.7.1, and
 VIC4.0.4 for
 GLDAS-1; CLM3.5,
 CLSM, Noah2.7.1,
 and VIC4.0.4
 for GLDAS-2
Research
NCEP/GLDAS NCEP GDAS (DSWR, DLWR, T2m, q2m,
 U10m, V10m, ps), blended P using gauge-
 based observations, satellite-retrievals,
 and GDAS P
Weakly, direct insert method for snow
 depth, 6-hourly assimilation cycle
Noah2.7.1 Operation
ECMWF/GLDAS ECMWF Integrated Forecasting System (P,
 DSWR, DLWR, T2m, q2m, U10m, V10m, ps)
Weakly, use 2-D optimal interpolation
 for T2m, RH2m, and snow depth and
 cover analysis, assimilate soil
 moisture with simplified extended
 Kalman filter at 6-hourly
 assimilation cycle
H-TESSEL Operation
APRIL 2019 Xia, Y. L., Z. C. Hao, C. X. Shi, et al. 5
precipitation to improve precipitation forcing. The im-
proved precipitation and other forcing data generated
from the Environment Canada numerical weather model
are used to drive the Interactions between Soil-Bio-
sphere-Atmosphere (ISBA) land surface model. The
CaLDAS also assimilates L-band passive brightness air
temperature by coupling the land surface model with a
microwave radiative transfer model to reduce surface and
root zone soil moisture errors (Balsamo et al., 2007; Car-
rera et al., 2015).
The CLDAS is a research-based system, which was
developed by Chinese scientists from the Northwest In-
stitute of Eco-Environment and Resources and Center for
Excellence in Tibetan Plateau Earth Sciences of Chinese
Academy of Sciences (CAS) (Li et al., 2007; Yang et al.,
2007, 2009). The CLDAS is an uncoupled LDAS system
to run multiple land surface models with multiple data
assimilation algorithms. Along with the development of
CLDAS, the scientists from CMA and the CAS Institute
of Atmospheric Physics developed an operational CMA
LDAS to support CMA operational drought monitoring.
The uncoupled CMA LDAS runs the Community Land
Model version 3.5 (CLM3.5; Oleson et al., 2008), Com-
mon Land Model (CoLM; Dai et al., 2003), and Noah
with multiple physics options (NoahMP; Niu et al., 2011)
using observed precipitation, 2-m air temperature, and
other blended forcing data (e.g., reanalysis, satellite,
model, etc.) at a 1/16 degree spatial resolution in China.
The CMA LDAS versions 1 and 2 were operationally im-
plemented at the CMA National Meteorological Informa-
tion Center in 2013 and 2016, respectively, and were
used for agricultural drought monitoring, wild fire early
warning, and water resource management. The surface
meteorological forcing has been evaluated against inde-
pendent observations and results show good perform-
ance of CMA LDAS precipitation compared with NASA
GLDAS forcing and observations (Yang et al., 2017).
The project-based LDAS is developed for specific ap-
plications and purposes. Three examples of this type of
LDASs include the NASA Famine Early Warning Sys-
tems Network (FEWS NET) Land Data Assimilation
System (FLDAS) (McNally et al., 2017), NASA Nation-
al Climate Assessment-Land Data Assimilation System
(NCA-LDAS) (Kumar et al., 2018), and NCAR High-
Resolution Land Data Assimilation System (HRLDAS)
(Chen et al., 2007). The FLDAS uses various reanalysis
products as meteorological forcing data and the Land In-
formation System framework (LIS; Kumar et al., 2006)
to drive multiple land surface models to generate soil
moisture, evapotranspiration (ET), runoff, and other vari-
ables for developing countries for food security assess-
ment in data-sparse regions. The NCA-LDAS adopts an
NLDAS framework (Xia et al., 2017, 2018) to run mul-
tiple land surface models to produce water fluxes, en-
ergy fluxes, and state variables. Its purpose is to use ter-
restrial water storage and fluxes from NCA-LDAS to
evaluate selected water indicators. NCAR HRLDAS was
developed for providing initial state variables for the
coupled Weather Research and Forecasting (WRF) in the
western US and testing weather forecast skill improve-
ment by using optimal initial conditions. The uncoupled
HRLDAS uses the same grid, Noah land surface model,
land use, soil texture, terrain height, time-varying vegeta-
tion fields, and model parameters as those in the coupled
WRF to ensure the climatology of similar state variables.
Radar-based and reanalysis meteorological forcing data
are used by HRLDAS to drive the Noah land surface
model as they have smaller errors when compared with
WRF forecasts products (Chen et al., 2007). It should be
noted that after the scope of these projects is over, only a
few are still maintained and most of them are not further
developed or maintained.
3. Global Land Data Assimilation System
(GLDAS)
GLDAS is an extension of the study domain from re-
gional and national to global scale, which includes the
NASA GLDAS (Rodell et al., 2004), NCEP GLDAS
(Meng et al., 2012), and ECMWF GLDAS (de Rosnay et
al., 2014; de Rosnay, 2017; Albergel et al., 2018). The
NASA GLDAS is an uncoupled system which has an in-
frastructure similar to the NLDAS but for a global do-
main. The objective of GLDAS is to use advanced data
assimilation techniques to assimilate satellite- and
ground-based observational products into land surface
models to generate optimal fields of land surface states
and fluxes, although neither satellite-based data nor in-
situ observations have been assimilated yet.
NASA GLDAS phase 1 uses a combination of
NCEP’s Global Data Assimilation System (GDAS) at-
mospheric analysis fields, spatially and temporally disag-
gregated NOAA Climate Prediction Center Merged Ana-
lysis of Precipitation (CMAP) pentad dataset, and obser-
vation-based downward shortwave and longwave radi-
ation fields derived from the Air Force Weather
Agency’s AGRicultural METeorological modeling sys-
tem (AGRMET). This combination forcing is used to run
the CLM, Noah, Mosaic, and VIC models (Rodell et al.,
2004). The spatial resolution is 1 degree and the time
period is from 1979 to present. GLDAS phase 2 has two
6Journal of Meteorological Research Volume 33
components: one is forced entirely by the Princeton met-
eorological forcing data (Sheffield et al., 2006), and the
other is forced by a combination of model and observa-
tion based forcing datasets as used in GLDAS phase 1.
The second component uses Global Precipitation Clima-
tology Project (GPCP) precipitation field to replace
CMAP precipitation with the improved disaggregation
scheme, and quality control for the AGRMET dataset.
More details of the model are available at the NASA GL-
DAS website (https://ldas.gsfc.nasa.gov/gldas/). A re-
cent evaluation study in China has shown that the use of
different data sources for GLDAS-1 causes a temporal
discontinuity, while the use of forcing data from Prin-
ceton University in GLDAS-2 has overcome this prob-
lem (Wang et al., 2016). The temporal discontinuity is a
very important issue for drought monitoring and water
resource management as this will lead to a sudden
change and jump for anomaly and percentile when a con-
sistent climatology is a need for anomaly and percentile
calculation. The products from GLDAS have been used
to support global drought analysis and water resource
management for government agencies, academia, and
private sectors all over the world.
NCEP GLDAS is a weakly coupled system and its ob-
jective is to provide optimal initial conditions for the
NCEP Climate Forecast System Reanalysis (CFSR; Saha
et al., 2010) and Climate Forecast System version 2
(Saha et al., 2014). The NCEP GLDAS uses the CFSR
global atmospheric data assimilation system and ob-
served precipitation data (CPC unified global daily gauge
analysis and pentad data of CMAP) to drive the Noah
land surface model on a T382 global Gaussian grid
(about 38 km). It uses six streams to run six years simul-
taneously with a one-year spin-up period to ensure the
product completion on a manageable schedule. All six
streams are connected to produce a long-term product for
the period from1979 to 2009.
An evaluation shows that the correlation between
NCEP GLDAS simulated and observed soil moisture an-
omaly has largely increased in Illinois (Meng et al.,
2012). The errors (bias and root mean square error)
between GLDAS simulated and observed soil moisture
are largely reduced when compared to the NCEP–NCAR
reanalysis and NLDAS soil moisture simulation (Meng et
al., 2012). However, the design of the six-streams-run
leads to large discontinuity for many state variables, such
as soil moisture in the connection points (e.g., 31 Decem-
ber 1986, 31 December 1989, 31 December 1994, 31
March 1999, 31 March 2005), because 1-yr spin-up time
is not sufficient (Cosgrove et al., 2003b) compared to the
recent GLDAS one-stream run and NLDAS-2 Noah sim-
ulations (Fig. 2). Recently, the NCEP LDAS team is in-
tegrating the NLDAS with GLDAS to develop the NCEP
unified LDAS (NULDAS) by leveraging the advantages
of NLDAS and GLDAS and the community LDAS de-
velopment. The NULDAS is a multiple-model LDAS
with a soil moisture and snowpack DA of 0.04 degree
globally. Details of NLDAS, GLDAS, and NULDAS de-
velopment plans are available in the LDAS white paper
(see https://www.emc.ncep.noaa.gov/mmb/nldas/White_
Paper_for_Next_Phase_LDAS_final.pdf).
ECMWF LDAS is a weakly coupled global LDAS in
the ECMWF Integrated Forecast System (IFS; de Rosnay
al., 2014; de Rosnay, 2017; Albergel et al., 2018). It in-
cludes snow data analysis and soil moisture DA (Fig. 3).
The purpose of the ECMWF LDAS is to provide reason-
able initial conditions for IFS. The surface meteorologic-
al forcing data produced by IFS, combined with 2-m air
temperature and relative humidity data from surface syn-
optic observations (SYNOP), are used for soil moisture
DA and a snow data analysis to generate updated soil
moisture, snow water equivalent, and snow temperature.
These state variables are used as initial conditions for IFS
for the next analysis cycle. The LDAS runs separately
from the upper-air analysis of IFS every six-hour cycle.
The state variables will feedback into the upper air ana-
lysis and affect the short-term forecast in the next cycle.
In turn, the 4D-Var atmospheric analysis affects the
LDAS through the short-term forecast from one cycle to
1980 1985
GLDAS2
0.872623
GLDAS2.1
0.884844
CFSR
0.659042
NLDAS
SW
1990
0−200-cm soil moisture anomaly
LAT = 30_40 LON = 240_255
1995 2000 2005
0.05
0.04
0.03
0.02
0.01
0
0.01
0.02
0.03
0.04
0.05
Fig. 2. Comparison of top 2-m soil moisture anomaly (SMA; m3 m–3)
simulated by Noah model in the NCEP GLDAS and NLDAS system.
Climate Forecast System Reanalysis (CFSR) uses GLDAS1 with a six-
stream run (vertical line), GLDAS2 uses a single-stream to rerun GL-
DAS1, and GLDAS2.1 uses the newly released global 0.25°-daily pre-
cipitation data from the NCEP Climate Prediction Center. The num-
bers are anomaly correlation coefficients between top 2-m soil mois-
ture simulations. The domain of the southwestern U.S. covers a range
of 30°–40°N, 120°–105°W.
APRIL 2019 Xia, Y. L., Z. C. Hao, C. X. Shi, et al. 7
the next. More details about the ECMWF LDAS system,
including publications and presentations, are provided at
https://software.ecmwf.int/wiki/display/LDAS/LDAS+
Home.
4. Validation and evaluation of regional and
global LDAS products
After discussing the development of national, regional,
and global LDASs in Sections 2 and Section 3, here we
will use the NLDAS as a brief example to introduce
LDAS products’ validation and evaluation. In NLDAS
phase 1, Mitchell et al. (2004) summarized the valida-
tion methods, tools, and datasets of NLDAS products,
such as radiation (Luo et al., 2003), sensible heat flux,
latent heat flux, soil temperature, soil moisture (Robock
et al., 2003; Schaake et al., 2004), snow cover extent
(Sheffield et al., 2003), snow water equivalent (Pan et al.,
2003), land skin temperature (Mitchell et al., 2004), and
streamflow (Lohmann et al., 2004). Metrics including bi-
as, relative bias, root-mean-square-error (RMSE), and
Nash-Sutcliffe efficiency have been used for the valida-
tion based on different temporal scales (hourly, daily,
and yearly) and different spatial scales (in-situ sites and
grids, watershed, and the continent). For NLDAS phase
2, long-term data were generated and more metrics such
as anomaly correlation, Taylor skill score (Taylor, 2001),
and Normalized Information Contribution (NIC, Kumar
et al., 2009) have also been included for the validation.
Additional validation work for NLDAS-2 can be seen on
the NLDAS validation website (https://ldas.gsfc.nasa.
gov/nldas/NLDAS2valid.php) and publications website
(https://ldas.gsfc.nasa.gov/nldas/NLDASpublications.php).
Many datasets from in-situ observations, satellite re-
trievals, and reanalysis have been used to evaluate LDAS
products. These datasets were summarized in Table 2.
The in-situ observations include tower flux measure-
ments (e.g., radiation, sensible and latent heat flux),
gauge measurements (e.g., streamflow, soil moisture,
snow depth, and snow water equivalent). The satellite re-
trievals include precipitation, evapotranspiration, land
surface temperature, soil moisture, snow cover, and snow
water equivalent. The reanalysis products include the
NCEP–NCAR Reanalysis 1 (R1, Kalnay et al., 1996),
NCEP–DOE (Department of Energy) Reanalysis 2 (R2,
Kanamitsu et al., 2002), CFSR (Saha et al., 2010), Twen-
tieth-Century Reanalysis (20CR, Compo et al., 2011),
MERRA(Rienecker et al., 2011), MERRA-2 (Draper et
al., 2018), ECMWF Interim Re-Analysis (ERA-Interim,
Dee et al., 2014), the fifth generation ECMWF atmo-
spheric reanalysis ERA-5 (Hersbach and Dee, 2016), Ja-
panese 25- and 55-yr Reanalysis Project (JRA-25, Onogi
et al., 2007; JRA-55, Kobayashi et al., 2015), etc. (see
Table 2). These datasets have been widely used to assess
the quality and reliability of the LDAS products. After a
comprehensive evaluation procedure, the LDAS products
can be used for various applications (e.g., research, oper-
ation) in academia, governmental agencies, and private
enterprises. These applications will be discussed in the
next section.
The evaluation and validation of the LDAS products
not only support various applications, but also contribute
to the improvement of surface meteorological forcing
and the development of LSMs’ physics processes. For
example, spatial discontinuity due to missing GOES-de-
rived downward shortwave radiation can be caused by
replacing the GOES data with the model’s downward
shortwave radiation. A bias correction method has been
used to overcome this problem in NLDAS-2 (Xia et al.,
2013a). Furthermore, NLDAS-1 validation has im-
proved the performance of the Noah and VIC models by
tuning model parameters and upgrading model physics
(Troy et al., 2008; Livneh et al., 2010; Wei et al., 2014).
Therefore, the improved Noah and VIC models were
used in NLDAS-2 to generate more accurate LDAS
products. In certain cases, the weakness and failure found
in the LDAS validation procedure cannot be solved im-
mediately due to a lack of the understanding of multiple
factors, such as physical processes or model parameters.
However, such findings can motivate the development of
land surface models and the improvement of LDAS sys-
tems, including surface meteorological forcing, model
physical processes, and model structures and parameters.
Integrated forecast system
land−atmosphere interaction
Tair and RH at 2 m
soil moisture at L1, L2, L3
Soil moisture DA
(simplified extended
Kalman filter)
Surface synoptic
observation (SYNOP),
Tair and RH at 2 m
Satellite soil
moisture
observation
Snow analysis (2D OI) IMS snow
cover extent, in situ snow depth
Screen level
analysis (2D OI)
Fig. 3. A schematic diagram for ECMWF LDAS systems (Tair: 2-m
air temperature; L1, L2, L3: three soil layers; 2D OI: two-dimensional
optimal interpolation; IMS: interactive multisensor snow; RH: relative
humidity; SYNOP: surface synoptic observation)
8Journal of Meteorological Research Volume 33
5. Application of regional and global LDASs
The products (e.g., soil moisture and temperature,
snowpack, streamflow, and evapotranspiration) pro-
duced by regional and global LDASs have been widely
used in numerical weather and climate forecast models
(Case et al., 2011; Saha et al., 2014), water resources
management (Zaitchik et al., 2010), drought monitoring
and prediction (Xia et al., 2014a, b; Hao et al., 2016a),
wild fire monitoring (Lewis et al., 2012), flood monitor-
ing and management of dams (Munier et al., 2015;
McNally et al., 2017), and agricultural crop management
Table 2. Summary of in-situ observations, satellite retrievals, and reanalysis products (AmeriFlux: American Flux network; CERES: Clouds
and the Earth’s Radiant Energy System; Fluxnet: Flux Network; GOES: Geostationary Operational Environmental Satellite system;
GEWEX/SRB: Global Energy and Water Exchange project/Surface Radiation Budget; ISMN: International Soil Moisture Network; ISTI: Inter-
national Surface Temperature Initiative; LST: Land Surface Temperature; MODIS: Moderate Resolution Imaging Spectroradiometer; NCDC:
National Climate Data Center; NASMD: North American Soil Moisture Database; SCAN: Soil and Climate Analysis Network; SNOTEL: Snow
Telemetry; SWE: Snow Water Equivalent; USGS: U.S. Geological Survey; USCRN: U.S. Climate Reference Network)
In-situ observations
Data name Domain Time steps Reference Website
AmeriFlux North America Hourly, daily, monthly Novick et al., 2018 http://ameriflux.lbl.gov/
Global FluxNet Globe Daily, monthly Baldocchi et al., 2001 http://fluxnet.fluxdata.org/
USGS streamflow US Daily, monthly Xia et al., 2012b https://nwis.waterdata.usgs.gov/nwis/
Global streamflow Globe Daily, monthly Beck et al., 2015 https://www.bafg.de/GRDC/EN/Home/
 homepage_node.html
NASMD soil moisture North America Daily, monthly Quiring et al., 2016 http://soilmoisture.tamu.edu/Data/
ISMN soil moisture Globe Daily, monthly Dorigo et al., 2011 https://ismn.geo.tuwien.ac.at/
ISTI LST Globe Daily, monthly Rennie et al., 2014 ftp://ftp.ncdc.noaa.gov/pub/data/
 globaldatabank, https://www.ncdc.
 noaa.gov/gosic
SCAN, SNOTEL, USCRN,
 NCDC soil temperature
US Hourly, daily, monthly Hu and Feng, 2003;Schaefer
 et al., 2007; Bell et al.,
 2013; Xia et al., 2013b
http://www.wcc.nrcs.usda.gov/,
 https://www.ncdc.noaa.
 gov/crn/
SNOTEL SWE and
 snow depth
US Daily Pan et al., 2003 https://www.wcc.nrcs.usda.gov/
 snow/snotel-wereports.html
Global Cryosphere Watch
 (snow depth, SWE)
Globe Daily, monthly See website https://globalcryospherewatch.org/
 reference/snow_inventory.php
Satellite retrievals
Data name Domain Temporal and
 spatial resolutions
Reference Website
Precipitation Globe Hourly to monthly,
 0.04° to 5°
Sun et al., 2018 N/A
LandFlux-EVAL ET Globe Monthly, varied
 spatial resolution
Jiménez et al., 2011 http://www.iac.ethz.ch/group/
 land-climate-dynamics/research/
 landflux-eval.html
Satellite-derived LST
 (GOES, MODIS, Landsat)
Globe GOES (hourly, 0.125°),
 MODIS (Daily, 8-day,
 monthly, 1km, 0.05°),
 Landsat (daily, 30m)
Yu et al., 2009; Li et al.,
 2013; Wan, 2014;
 Parastatidis et al., 2017
https://www.ospo.noaa.gov/Products/
 land/glst/, https://modis.gsfc.nasa.
 gov/data/dataprod/mod11.php
Satellite-derived
 snow products
Globe Daily, varied
 spatial resolution
Frei et al., 2012 https://globalcryospherewatch.org/
 reference/snow_inventory.php
Satellite-derived
 soil moisture
Globe Daily, 30 m-50 km Brocca et al., 2011 https://ismn.geo.tuwien.ac.at/satellites/
Satellite-derived radiation
 (GEWEX/SRB, CERES,
 GOES-R)
Globe SRB (3-houly, daily,
 monthly; 1°), CERES
 (monthly; 1°), GOES-R
 (hourly; 0.25°, 0.5°)
Pinker et al., 2003; Zhang
 et al., 2013; 2015; Kato
 et al., 2018
https://gewex-srb.larc.nasa.gov/,
 http://ceres.larc.nasa.gov/,
 https://www.star.nesdis.noaa.gov
 /goesr/product_sw.php
Reanalysis products
Data name Domain Temporal and spatial
 resolutions
Reference Website
Atmospheric reanalysis Globe Sub-daily, daily, monthly
 32 km to 2.5°
Bromwich and Wang, 2005;
 Uppala et al., 2005;
 Mesinger et al., 2006;
 Onogi et al., 2007; Saha
 et al., 2010; Rienecker
 et al., 2011; Dee et al.,
 2014; Laloyaux et al.,
 2016
https://climatedataguide.ucar.edu/
 data-type/atmospheric-reanalysis,
 https://climatedataguide.ucar.edu/
 climate-data/atmospheric-reanalysis-
 overview-comparison-tables
APRIL 2019 Xia, Y. L., Z. C. Hao, C. X. Shi, et al. 9
(Liu et al., 2017). A schematic diagram briefly summar-
izing scientific and practical applications of regional and
global LDAS products, which can provide support for
operational forecasters, decision makers, and research
scientists in governmental agencies, academia, and the
private sectors, is shown in Fig. 4. More details are dis-
cussed below.
5.1 Numerical weather forecast and seasonal
climate prediction
The land surface is a low-level boundary condition of
the atmosphere, which affects lower-level atmospheric
activities, such as redistribution of energy and water va-
por, boundary layer height growth, and convention cloud
formulation. An LSM is usually used to simulate land
surface processes, such as the continental hydrological
cycle and the interaction between atmosphere and land
surface on various temporal and spatial scales. The LSM
is a component in the coupled GCM (General Circula-
tion Model), which is the main tool to providing numer-
ical weather and seasonal climate forecasts. Therefore,
accurate initial states, such as soil moisture and temperat-
ure, snow water equivalent, snow cover, snow temperat-
ure, and snow density, can improve numerical weather
and climate prediction. The atmosphere and land surface
can interact by exchange of energy (i.e., sensible and lat-
ent flux) and water vapor (i.e., evapotranspiration). The
LSM is driven by low-level atmospheric forcing data,
such as precipitation, downward shortwave and long-
wave radiation, 2-m air temperature and specific humid-
ity, 10-m wind speed, and surface pressure. As the
coupled GCM produces substantial biases for precipita-
tion and downward shortwave radiation when compared
with observations, these biases often yield significant er-
rors and drift in soil moisture, soil temperature, and en-
ergy and water fluxes. These errors and drift in turn af-
fect atmospheric circulation, redistribution of energy and
water vapor, as well as convective activity. The major
purpose of regional and global LDASs is to ingest in-situ
and remotely-sensed data into the LSM model to en-
hance the accuracy of initial states and energy/water
fluxes to improve land-atmosphere interaction and the
GCM forecast ability.
Several operational centers have used their national
and global LDAS systems to enhance the prediction skill
of their coupled weather and seasonal climate forecast
systems. The NCEP ingested the soil moisture and tem-
perature generated from GLDAS (Meng et al., 2012) in-
to the CFSR reanalysis (Saha et al., 2010) and CFSv2
forecast (Saha et al., 2014) to enhance seasonal simula-
tion and forecast skill. The overall results in the CFSR
show large improvements when compared with the
NCEP-DOE R2 analysis and the observations. Sensitiv-
ity tests show that initial states generated from GLDAS
improve CFS forecast skill when compared with initial
states generated from R2 (Yang et al., 2011). The CaL-
DAS was developed to provide better representative land
surface initial states for the environmental prediction and
assimilation systems of Environmental Canada to en-
hance their forecast skill (Carrera et al., 2015).
The ECMWF LDAS system includes a 2-m air tem-
perature and relative humidity analysis, snow depth and
temperature analysis, soil temperature analysis, and soil
moisture DA. It first uses a 2-dimensional OI method to
produce 2-m air temperature and relative humidity, and
then the two produced fields are used as forcing in the
data assimilation of snow depth analysis and soil mois-
ture. Next, the data generated from previous steps are
used for the soil and snow temperature analysis. Finally,
all initial states (i.e., soil moisture and temperature, snow
water equivalent, snow temperature, snow density, etc.)
generated from this LDAS are provided to the ECMWF
IFS (de Rosnay, 2017). In the research community, Case
et al. (2011) improved numerical weather predictions of
summertime precipitation over the southeastern United
States by using LDAS-generated land surface states. Os-
uri et al. (2017) used LDAS-generated realistic initial
conditions to drive an NWP model to improve the predic-
tion of severe thunderstorms over the Indian monsoon re-
gion. Santanello et al. (2016) investigated the impact of
soil moisture assimilation on coupled land-atmosphere
prediction, and the results showed the potential for high-
er-resolution soil moisture assimilation applications in
weather and climate research. More details about the
Regional and global LDAS systems (LSM, DA)
(Energy fluxes, water fluxes, state variables)
Drought
monitoring and
prediction
Numerical
model
prediction
Agriculture
and wild fire
Flood
monitoring and
water resource
Meteorological,
hydrological,
and agricultural
Weather,
S2S, and
climate
Crop
management and
planning, wild
fire warning
Flood,
precipitation/snow,
lake/river runoff,
and irrigation
Fig. 4. A schematic diagram of regional and global LDAS outputs
and their applications (S2S: seasonal to sub-seasonal).
10 Journal of Meteorological Research Volume 33
ECMWF LDAS system, including publications and
presentations can be found in its website (https://soft-
ware.ecmwf.int/wiki/display/LDAS/LDAS+Home). Note
that the current NCEP and ECMWF GLDASs are
coupled to atmospheric GCM models every six hours
rather than at each time step. Strictly speaking, such a
coupling is generally called weak or quasi coupling. The
coupling at each time step is expected in the future.
Although the weakly coupled LDAS system has been
used in numerical weather and seasonal climate systems,
when compared with the development and applications
of uncoupled LDAS, its application is still at an initial
stage. More experiments and investigations about the im-
pact of LDAS-generated initial state on weather and cli-
mate prediction are needed in the future.
5.2 Drought monitoring and prediction
Drought is generally defined as a prolonged period of
water deficit or water imbalance. Its characterization is
commonly based on different drought indices. Drought
can be categorized into four different types, including
meteorological, agricultural, hydrological, and socioeco-
nomical drought (Heim, 2002). The monitoring of differ-
ent types of drought can be achieved by computing
drought indictors for a relatively long record. Traditional
drought monitoring is generally based on in-situ meteor-
ological observations, including precipitation and tem-
perature. Drought indicators, such as the Palmer Drought
Severity Index (PDSI) (Palmer, 1965) or the standard-
ized precipitation index (SPI; McKee et al., 1993), have
been developed based on these observations and em-
ployed for drought monitoring. In recent decades, ad-
vancements in remote sensing (Mu et al., 2013;
AghaKouchak et al., 2015) and land surface modeling
(Nijssen et al., 2014; Xia et al., 2014a) have advanced
drought monitoring, which provides important monitor-
ing information at regional or even global scales based
on a variety of land surface variables (e.g., soil moisture,
runoff, etc.). However, certain challenges still exist in
drought monitoring based on remote sensing (e.g., errors
in the estimations) or land surface simulations (e.g., un-
certainties in the model simulations even with the same
forcing products)( Hao et al., 2017).
To address these challenges, data assimilation has
been commonly used to merge in-situ observations, re-
mote sensing products, land surface model simulations,
and climate forecasts for drought monitoring and predic-
tion (Hao et al., 2018). A variety of LDASs developed in
recent decades, including the NLDAS and GLDAS,
provide multiple state and flux variables that have been
widely applied for the drought monitoring (Xia et al.,
2014 a b; Hao et al., 2016a). Meanwhile, data products
from LDAS also provide useful information for drought
prediction by providing predictors or initial conditions
(Hao et al., 2016b). For example, based on the recently
developed Coupled Land and Vegetation Data Assimila-
tion System (CLVDAS), initial conditions for both soil
moisture and leaf area index (LAI) can be obtained for
the prediction of eco-hydrological droughts (Sawada and
Koike, 2016).
5.3 Agricultural crop management
Crop yields are important for socioeconomic develop-
ment in different regions around the world, and the ac-
curate estimation of crop yields plays an important role
for informed decision-making. Traditionally, agricultural
crop management, including crop condition monitoring
and crop yield predictions, is generally based on crop
models or remote sensing observations. For the crop
model, the status of crop growth can be simulated for
timely crop management and decision-making to maxim-
ize crop yields. Remote sensing can offer comparatively
accurate and reliable input variables for the crop models
to improve simulation results.
However, the single crop model is not capable of ex-
panding a single point to regional scale, due to the diffi-
culty of parameters needed, while remote sensing based
products may fall short in time continuity and agricul-
ture rationale (Li et al., 2011). The DA technique
provides an important tool in integrating crop models and
remote sensing observations to address these two chal-
lenges with improved model simulations. The applica-
tion of LDAS in agriculture mainly focuses on two as-
pects, including the monitoring of crop growth condition
(Ma et al., 2008; Machwitz et al., 2014) and crop yield
forecasting (deWit and van Diepen, 2007; Dente et al.,
2008; Li et al., 2014), both of which are crucial for agri-
cultural operations and management, for they can
provide reference information for precise irrigation, fer-
tilization, irrigation and forecast of agricultural pro-
ductivity.
A commonly adopted scheme for the DA application
is to assimilate different types of observations from re-
mote sensing products with crop models. In the study of
monitoring crop growth, many researchers have used as-
similated remote sensing data to improve the ability of
crop models to simulate crop growth, such as leaf area
index (LAI; Pellenq and Boulet, 2004), biomass (Mach-
witz et al., 2014), soil water (Nouvellon et al., 2001), and
so on. Similarly, a lot of studies have improved the crop
yield estimates by assimilating vegetation indices into
crop models, such as LAI (Xie et al., 2017; Mokhtari et
APRIL 2019 Xia, Y. L., Z. C. Hao, C. X. Shi, et al. 11
al., 2018), biomass (Jin et al., 2017), and soil moisture
(Chakrabarti et al., 2017). A simple example of using the
LDAS to aid agricultural crop planning is shown in Fig. 5.
In this process of combining the assimilation of re-
mote sensing data with crop models, both remote sens-
ing techniques and crop models play a vital role. To sim-
ulate crop growth conditions accurately, researchers have
made progress in improving crop models such as the
WOFOST (World Food Studies; van Diepen et al., 1989)
and DSSAT (Decision Support System for Agrotechno-
logy Transfer; Jones et al., 2003). Meanwhile, some new
satellite products (e.g., Gaofen-1, Feng et al., 2016;
SPOT-6, Henry et al., 2017) as well as new tools for
gaining information about crop dynamic change (e.g., the
Unmanned Aerial Vehicles) have emerged. In addition,
the development of radar technology has also provided
data on crop growth status for assimilation. Generally, al-
though combing DA with crop models has meant some
significant improvements in the accuracy of crop models
in crop growth simulation and yield prediction, there re-
mains much scope for further improvements. For in-
stance, how to combine different crop models as well as
acquire remote sensing data of higher spatial and tempor-
al resolutions are still important for more accurate simu-
lations of crop growth and yields.
5.4 Flooding warning and water resources management
Hydrological models have been among the most valu-
able tools in simulating hydrological processes at region-
al and global scales. However, these models are subject
to a suite of uncertainties in the forcing data and model
structures or parameters. Quantifying and reducing un-
certainties are thus needed to obtain an accurate and reli-
able hydrological forecast. Traditionally, the data assim-
ilation technique provides promising potential to im-
prove the hydrological model with improved accuracy
and the quantification of uncertainty through integrating
model simulations and observations (Liu et al., 2012;
NLDAS top 1 m soil moisture
as of September 20, 2011
Major crop areas
Percentile
Corn
>98
95
90
80
70
30
20
10
5
2
<2
ND
NLDAS top 1 m soil moisture
as of September 20, 2011
Major crop areas
Percentile
Corn
>98
95
90
80
70
30
20
10
5
2
<2
ND
NLDAS top 1 m soil moisture
as of September 20, 2011
Major crop areas
Percentile
Winter wheat
>98
95
90
80
70
30
20
10
5
2
<2
ND
(a)
(b) (c)
Fig. 5. Applications of three-model averaged NLDAS top 1-m soil moisture for agricultural crop planning in the USDA. (a) Used in conjunc-
tion with USDA crop-area shape files for crop-weather assessment to assess the dryness depicted in the Corn Belt of US, (b) adverse conditions
afflicting cotton from Texas’ drought to excessive wetness in the mid-Atlantic, and (c) dire planting prospects for winter wheat on the southern
Plains of US.
12 Journal of Meteorological Research Volume 33
Khaki et al., 2018).
As one of the popular data assimilation methods, Kal-
man filter methodology was introduced into flood fore-
casting probably in the 1970s, which was an early use of
data assimilation methods in hydrologic studies. There-
after, the Kalman filter received extensive attention in the
application of hydrological forecasting (e.g., Kitanidis
and Bras, 1980). Along with the development of en-
semble theory, the ensemble Kalman filter method was
developed in the late 1990s and has been widely used in
flood warnings and other hydrologic predictions
(Komma et al., 2008; Fan et al., 2017). Many of these ap-
plications are involved in the assimilation of gauge ob-
servations into hydrological models to improve stream-
flow prediction. The advancement of the technology of
satellite remote sensing, geographic information systems
and rain detection with digital radar has promoted the de-
velopment of assimilating remote sensing products (e.g.,
precipitation, snow cover, or soil moisture) into hydrolo-
gic models (Crow and Wood, 2003; Reichle and Koster,
2005; Liu et al., 2017). Though hydrologic DA has been
shown to be a useful tool for reducing predictive uncer-
tainty and improving prediction accuracy, its application
to operational hydrologic forecasting and water resource
management is rather limited, partly due to a lack of a
clear mechanism to quantify the uncertainty in observa-
tions and forecast models and to merge data/models in an
efficient and transparent way for forecasters (Liu et al.,
2012).
Due to the errors/uncertainties in the atmospheric for-
cing (initial conditions), model structure, and parameters,
the ensemble forecasting approach has gained much pop-
ularity and provided a new idea for flood forecasting and
water resource management, which prolongs the effect-
ive prediction period and improves the forecast accuracy
(Cloke and Pappenberger, 2009; Doycheva et al., 2017).
However, a number of challenges also exist in ensemble
hydrological prediction, such as providing consistent en-
semble forcing and improving uncertainty estimation in
observations, which calls for advances in many elements
in the ensemble forecast system (Seo et al., 2014). DA
has been integrated with the ensemble hydrological fore-
casting for a variety of applications including the uncer-
tainty quantification. For example, Wang et al. (2018) re-
duced the uncertainties of hydrologic ensemble forecast-
ing by pre-processing and post-processing assimilation
data. Recently, new assimilation schemes have been de-
veloped to address certain limitations in DA and im-
prove hydrological forecasting. For example, to address
the challenge of computational demand with high dimen-
sional systems in hydrological applications, Khaki et al.
(2018) proposed a nonparametric data assimilation
scheme (i.e., Kalman–Takens filter) for the hydrological
application, which successfully improves the estimated
water storage with less computational time. A review of
the current progress, challenges, and opportunities of
data assimilation in hydrologic forecasting was provided
by Liu et al. (2012).
6. Challenges and future prospects in LDAS
The national, regional, and global LDASs have been
successfully developed over the past two decades and
their products have been widely applied in a variety of
applications including numerical model prediction and
drought and flood monitoring. However, there are still
many challenges that may limit their further develop-
ment and applications. These challenges come from poor
data quality, a lack of scientific understanding of model
physical processes and parameters, and spatial scale mis-
matches between in-situ sites and model grid cells. These
challenges are discussed as follows.
6.1 In-situ and remotely-sensed data and
reanalysis products
In this section, we first discuss general issues in com-
monly used data products in hydrometeorology as they
directly affect LDASs’ development including the
products’ evaluation/validation. In-situ measurements
and remotely-sensed data are the basis for LDAS devel-
opment, including LSM development, the DA process,
product evaluation, and LSM parameter calibration.
High-quality observations are crucial for LDAS systems
development and applications. However, the quality of
observed data is affected by missing data, malfunction-
ing instruments and sensors, different measurement
standards, and limitations of measurement techniques. In
particular, the measurement error and uncertainty are un-
known for ground measurements.
6.1.1 In-situ observations
(a) Soil moisture
For measured soil moisture, Dorigo et al. (2013) and
Quiring et al. (2016) summarized various causes of data
quality issues and developed automated quality control
methods to flag erroneous daily data measurements (Xia
et al., 2015b;Liao et al., 2019). Gruber et al. (2016) used
a triple collocation method to estimate the average net-
work error and sensor error with very limited data. These
efforts support the community goal to reasonably incor-
porate in-situ soil moisture observations into LDAS. For
soil temperature, a quality control method (Hu et al.,
2002) is used to flag erroneous daily soil temperature
APRIL 2019 Xia, Y. L., Z. C. Hao, C. X. Shi, et al. 13
data for 292 U.S. cooperative stations (Hu and Feng,
2003). The quality-controlled data have been used to
evaluate soil temperature simulations in NLDAS-2 (Xia
et al., 2013b). It should be noted that the International
Soil Moisture Network (ISMN) has integrated most of
the in-situ soil moisture observational network measure-
ments into a single dataset, which is very helpful for
global LDAS development and applications. However,
in-situ soil temperature measurements are still archived
in different countries and by different organizations.
Therefore, an effort to integrate these in-situ soil temper-
ature measurements into a single dataset as ISMN is
needed in the future.
(b) Fluxes
There are very limited tower flux measurements over
the globe with a lot of missing data due to instrument/
sensor failures, in particular during the cold season
(September to April). In addition, because net radiation,
sensible and latent heat fluxes, and ground heat flux are
measured separately, their energy budgets are not bal-
anced (Wilson et al., 2002; Foken, 2008). Even though
the energy balance closure at some sites has been correc-
ted on a monthly time scale (Jung et al., 2010), the cor-
rection and reduction of this closure error are still uncer-
tain for all timescales, in particular at hourly and daily
timescales (Shuttleworth, 2007). Some corrected and un-
corrected flux data have been used to crudely evaluate
NLDAS-2 ET simulations (Xia et al., 2015a).
(c) Streamflow
Streamflow measured from gauges has been widely
used to evaluate/calibrate LSMs and hydrological mod-
els (Lohmann et al., 2004; Troy et al., 2008; Xia et al.,
2012b). In general, the measurement errors come from
observations of stream stage, periodic measurements of
streamflow, and development of a rating curve to con-
vert stage to discharge (Hamilton and Moore, 2012).
Harmel et al. (2006) briefly summarized the uncertainty
range in streamflow measurements by various methods
for small watersheds (varying from 4% to over 100%).
Furthermore, as it is an integrated variable for a given
watershed/basin, the data quality was largely affected by
many human activities such as dam building, regulation
of water use, and irrigation. In particular, for long-term
(> 30 yr) measurement data, these human activities also
result in data discontinuity before and after the dams are
built. Although NLDAS developers have identified over
1000 small-medium size basins ( 10000 km2) with small
human effects, there are still data from gauges that have
some human affects (Xia et al., 2012b). To solve this
problem, it is a reasonable choice that either the data er-
ror is adjusted based on known human activities or these
activities are included in the LSM as physical processes.
(d) Snowfall, snow depth, and snow water equivalent
In situ snow depth and snow water equivalent are two
important variables in the development of regional and
global LDASs. Systematic measurement errors caused by
wind include wetting loss, evaporative loss, capping of
gauge orifice, and blowing snow (Rasmussen et al.,
2012). A double fence intercomparison experiment or-
ganized by World Meteorological Organization (WMO)
has shown that the catch efficiency for most widely used
gauges has been reduced from 100% to 20% (depending
on the gauge type) when wind speed is increased from 0
to 8 m/s (Yang et al., 1998). Therefore, wind bias adjust-
ments for the measurement of snowfall in windy environ-
ments are the first step. In general, wind causes an under-
catch of snowfall measurement. In contrast, blowing
snow can result in overestimation of snow depth and
snow water equivalent. Recently, bias correction for
wind has been applied to adjust gauge precipitation,
snowfall, snow depth, and snow water equivalent
(Ungersböck et al., 2001; Yang et al, 2005) to reduce the
systematic bias of these measured data.
6.1.2 Remotely-sensed data
Although in-situ data are measured directly by vari-
ous instruments, these data are station dependent. There
are few stations in remote regions such as Africa and the
polar areas. Remotely-sensed data related to the water
cycle (e.g., precipitation, snowpack, soil moisture, ET,
etc.) are grid-based data with temporal and spatial con-
tinuity (Table 2). However, most remotely-sensed data
are not directly measured from satellites and they need to
be retrieved from satellite images using either empirical
or physical-based algorithms. Therefore, errors come
from not only satellite image errors but also uncertain-
ties in the parameters used in these algorithms, which
need to be calibrated using limited in-situ observational
data. Derin and Yilmaz (2014) indicated that remotely-
sensed precipitation generally has difficulty in represent-
ing high spatiotemporal variability in areas with com-
plex topography when the precipitation is controlled by
orography. Infrared-based retrievals cannot capture light
precipitation events and underestimate orographic rain-
fall amounts, whereas passive-microwave based retriev-
als have difficulty in detecting orographic precipitation,
especially during the cold season when the precipitation
is snowfall. A comparative analysis for global precipita-
tion products, including satellite retrievals, was per-
formed and the details about the assessment of satellite-
retrieved precipitation quality can be seen in Sun et al.
14 Journal of Meteorological Research Volume 33