Atmospheric water balance over oceanic regions as estimated
from satellite, merged, and reanalysis data
Hyo-Jin Park1, Dong-Bin Shin1 and Jung-Moon Yoo2
1
2
Department of Atmospheric Sciences, Yonsei University, Seoul, Korea
Department of Science Education, Ewha Womans University, Seoul, Korea
Submitted to Journal of Geophysical Research
November 15 2012
1 Corresponding author’s address: Dong-Bin Shin, Department of Atmospheric Sciences,
Yonsei University, 50 Yonsei-ro, Seodaemun-gu, Seoul 120-749, Korea. Tel: 82-2-2123-5685,
Fax: 82-2-365-5163. E-mail:[email protected]
1
Abstract
2
3
The column integrated atmospheric water balance over the ocean was examined using
4
satellite-based and merged datasets for the period from 2000 to 2005. The datasets for
5
the components of the atmospheric water balance include evaporation from the HOAPS,
6
GSSTF, and OAFlux, and precipitation from the HOAPS, CMAP and GPCP. The water
7
vapor tendency was derived from water vapor data of HOAPS. The product estimated
8
by Xie et al. [2008] was used for water vapor flux convergence. The atmospheric
9
balance components form the MERRA reanalysis data were also examined. Residuals
10
of the atmospheric water balance equation were estimated using nine possible
11
combinations of the datasets over the ocean between 60˚N and 60˚S. The results showed
12
that there was considerable disagreement in the residual intensities and distributions
13
from the different combinations of the datasets. In particular, the residuals in the
14
estimations of the satellite-based atmospheric budget appear to be large over the oceanic
15
areas with heavy precipitation such as the inter-tropical convergence zone, South Pacific
16
convergence zone, and monsoon regions. The lack of closure of the atmospheric water
17
cycle may be attributed to the uncertainties in the datasets and approximations in the
18
atmospheric water balance equation. Meanwhile, the anomalies of the residuals from the
19
nine combinations of the datasets are in good agreement with their variability patterns.
20
These results suggest that significant consideration is needed when applying the datasets
21
of water budget components to quantitative water budget studies, while climate
22
variability analysis based on the residuals may produce similar results.
23
24
25
1.
Introduction
26
27
The atmospheric water cycle is one of most important components of the global water
28
cycle. Large amounts of water vapor that are evaporated from the ocean are transported
29
to the continents through the atmosphere. The transported water vapor is converted into
30
precipitation that provides vital water for living things on Earth. Precipitation over
31
ocean surfaces supplies the fresh water that changes sea surface salinity and drives
32
ocean circulation. Changes in the phase of water in the atmosphere involve latent heat
33
exchanges. Latent heat is one of the major forces driving the general circulation of the
34
atmosphere. Knowledge of the atmospheric water cycle is therefore essential in order to
35
manage water resources and to understand the Earth’s climate.
36
37
The atmospheric water cycle has been investigated in many regions. The Global Energy
38
and Water Cycle Experiment (GEWEX) initiated by the World Climate Research
39
Program (WCRP) is well known for its scientific studies of the water cycle. The
40
GEWEX aims at observing, understanding, and modeling the hydrological cycle and
41
energy fluxes in order to predict global and regional climate change. Under the missions
42
of the GEWEX, projects including the Regional Hydroclimate Projects (RHPs)
43
(formerly the Continental Scale Experiment, CSE) were initiated. The focus of the
44
GEWEX was to solve the problems of closing the balance of water and energy. In
45
addition, several water cycle studies were performed at the atmospheric branch in order
46
to quantify the regional water balance, including the Mackenzie GEWEX Studies
47
(MAGS) [e.g., Stewart et al., 1998 ; Rouse et al., 2003], the Baltic Sea Experiment
48
(BALTEX) [e.g., Raschke et al., 2001 ; Ruprecht and Kahl, 2003], the Climate
49
Prediction Program for the Americas (CPPA) [e.g., Roads et al., 2003], and the Murray-
50
Darling Basin (MDB) [e.g., Draper and Mills, 2008].
51
52
One of the major complications of the atmospheric water balance study is the selection
53
of appropriate datasets. The atmospheric water balance equation includes water cycle
54
components such as evaporation, precipitation, water flux convergence, and water
55
tendency. The equation itself is not complicated. However, accurate estimations of all of
56
the water cycle components are difficult to achieve. Water cycle components have been
57
acquired from direct measurements, numerical weather prediction (NWP) model-
58
derived products, remotely sensed observations, and residual methods using the
59
atmospheric water balance equation. Direct measurements provide relatively reliable
60
data, but have limited observation coverage. Earlier water cycle studies were usually
61
performed using radiosonde data [Rasmusson, 1967, 1968]. Despite its limited spatial
62
and temporal coverage, radiosonde data is still used over the regions with dense
63
radiosonde networks [Kanamaru and Salvucci, 2003 ; Zangvil et al., 2004]. NWP-
64
derived analysis and reanalysis data have been actively employed for the study of the
65
water budget
66
Draper and Mills, 2008], because of their ability to minimize known errors and high
67
spatiotemporal resolutions. NWP products, however, are highly dependent on model
68
physics and parameterization. With advancement in satellite instruments and retrieval
69
algorithms, more satellite observation data has become available [e.g., Bakan et al.,
70
2000]. Satellite observations, which have better coverage than in situ observations
71
particularly over the oceans, have been widely used for a complementary purpose or for
72
comparison with other observations. Many satellite-based or merged datasets for water
[e.g., Roads et al., 2003; Ruprecht and Kahl, 2003; Turato et al, 2004;
73
cycle components have been produced using such satellite observations.
74
75
In this study, the atmospheric water balance will be examined over oceans using various
76
satellite-based and merged datasets. Reanalysis datasets were also used for comparison
77
with satellite-based datasets for the atmospheric water balances over oceanic regions.
78
79
2.
80
2.1. Atmospheric water balance over the ocean
Data and methodology
81
82
The column integrated atmospheric water balance over the ocean can be expressed as
83
follows:
84
85
 
 (W  Wc )
 ( E  P)    (Q  Qc )
t
(1),
86
87
where E and P are the evaporation and precipitation, respectively. W indicates the
88

column integrated total water vapor, Q is the horizontal water vapor flux vector, and
89
 
the subscript c denotes the condensed phase of water. W  Wc and Q  Qc are defined
90
as:
91
ps
92
W  Wc   (q  qc )
p0
93
and
dp
g
(2)
94
ps
 
 dp
Q  Qc   (q  qc )v
g
p0
(3),
95
96
where g is the acceleration of gravity, p s is the pressure at the surface, p0 is the
97
pressure at the top of the atmosphere, q is the specific humidity, qc is the condensed
98

water mixing ratio, and v is the horizontal wind velocity vector.
99
100

The terms Wc / t and Qc are related to the condensed phase water and are usually
101
small enough that both the tendency of liquid and solid water in clouds and their
102
horizontal flux can be ignored in Eq. 1. Therefore, by averaging Eq. 1 in the specified
103
temporal and spatial domain of interest, the general atmospheric water balance equation
104
can be simplified to:
105
106

W
 E  P  Q
t
(4),
107
108
where the bar indicates the time average and the angular bracket denotes the area
109
average. Equation 4 shows that the excess of evaporation over precipitation is balanced
110
by the local rate change of the water vapor contents and by the horizontal net flux of
111
water vapor.
112
113
In order to examine the atmospheric water balance, we calculated the residuals using Eq.
114
4 as follow:
115
116

W
R  E  P  Q 
t
(5),
117
118
where R represents the residual of the atmospheric water balance equation. R was
119
obtained from combinations of the datasets for the components of the water cycle in Eq.
120
5. Section 2.2 describes the datasets used for this study.
121
122
2.2. Datasets
123
124
Table 1 summarizes the characteristics of the datasets for the water cycle components
125
used in this study. For evaporation (E), the monthly fields of three datasets were used.
126
The first dataset was the Hamburg Ocean Atmosphere Parameters and Fluxes from
127
Satellite data (HOAPS) and the second dataset was the Goddard Satellite-based Surface
128
Turbulent Fluxes (GSSTF). The third dataset was the Objectively Analyzed Air-Sea
129
Heat Fluxes (OAFlux) project at the Woods Hole Oceanographic Institution (WHOI). In
130
HOAPS, evaporation data was obtained from all of the available Special Sensor
131
Microwave Imager (SSM/I) observation data based on the Coupled Ocean-Atmosphere
132
Response Experiment (COARE) bulk flux algorithm version 2.6a [Bradley et al., 2000;
133
Fairall et al., 1996] for latent heat flux parameterization. The near-surface wind speed
134
and specific air humidity were retrieved from the SSM/I brightness temperature and the
135
sea surface temperature from the Advanced Very High Resolution Radiometer (AVHRR)
136
measurements were used [Andersson et al., 2010]. The GSSTF latent heat flux was
137
retrieved based on the bulk flux model [Chou, 1993], using the SSM/I surface wind,
138
surface air humidity, and near surface air and sea surface temperatures from the NCEP-
139
NCAR reanalysis as the input data [Chou et al., 2003; Shie et al., 2010]. The OAFlux
140
evaporation data was derived from the blending of the satellite retrievals from the
141
SSM/I, the Quick Scatterometer (QuikSAT), the AVHRR, the Tropical Rain Measuring
142
Mission (TRMM) Microwave Imager (TMI), and the Advanced Microwave Scanning
143
Radiometer Earth Observing System (EOS) (AMSR-E), as well as the NWP reanalysis
144
outputs from the NCEP (e.g., NCEP/NCAR, NCEP/DOE reanalysis) and ECMWF (e.g.,
145
ERA40) by objective analysis techniques [Yu and Weller, 2007, Yu et al., 2008]. The
146
OAFlux products were constructed based on the COARE bulk flux algorithm version
147
3.0 [Fairall et al., 2003].
148
149
The following three monthly precipitation (P) products were used: HOAPS, the Global
150
Precipitation Climatology Project (GPCP), and the Climate Prediction Center (CPC)
151
Merged Analysis of Precipitation (CMAP). The HOAPS precipitation was retrieved
152
from a neural network algorithm that takes the SSM/I brightness temperature and the
153
precipitation from the ECMWF model as training data [Andersson et al., 2010]. Both
154
the GPCP [Adler et al., 2003; Huffman et al., 2009] and the CMAP [Xie and Arkin, 1997]
155
use multiple satellite and rain gauge datasets with some differences in their input data
156
and merging techniques [Yin et al., 2004].
157
158
For water vapor flux convergence (WVFC), the monthly WVFC estimated by Xie et al.
159
(2008, hereinafter referred to as XLT08) was used. The XLT08 algorithm estimates the
160
WVFC based on support vector regression (SVR) using the surface wind vector from
161
the Quick Scatterometer (QuikSCAT), the cloud drift wind vector from the Multi-angle
162
Imaging Spectroradiometer (MISR) and the NOAA geostationary satellites, and the
163
precipitable water from the SSM/I.
164
165
The water vapor tendency (WVT) data was derived in this study using the HOAPS
166
twice-daily total column water vapor (TPW) data. The HOAPS twice-daily TPW was
167
averaged over a pentad of days in order to create daily data (e.g., the 1/1/2005 daily
168
TPW is the averaged value of the 12/30/2004 through 1/3/2005 twice daily TPW). The
169
monthly WVT was then calculated using the following equation:
170
171
W f  Wi
 W 

 
t
 t m
(6),
172
173
where t is the one-month time interval, and i and f indicate the first and the
174
day of the m th month, respectively.
175
176
Modern-Era Retrospective analysis for Research and Applications (MERRA), which is
177
NASA’s new reanalysis method, was also used for comparison with the satellite-based
178
and merged datasets. The MERRA is generated using the Goddard Earth Observing
179
System (GEOS) atmospheric model and data assimilation system (DAS), version 5.2.0
180
[Rienecker et al., 2011]. MERRA provided the various quantities for the atmospheric
181
water cycle components.
182
183
Figure 1 is a diagram of the estimated residuals from nine possible combinations of
184
satellite and merged datasets for water cycle components. For example, the residual
185
OGXH indicated that the residual came from the combination of evaporation for
186
OAFlux, precipitation for GPCP, water vapor flux convergence for XLT08, and water
187
vapor tendency for HOAPS. All of the monthly datasets were remapped to have a 5˚ by
188
5˚ lat-lon spatial resolution for the period from 2000 to 2005. For each dataset, we
189
averaged the data values at each 5˚x5˚ grid if the number of the missing value did not
190
exceed 50% of the total number of data points. The domain is limited to oceanic regions
191
between 60˚N and 60˚S. Oceanic regions are defined as the areas where the MERRA’s
192
ocean fraction is greater than 0.9.
193
194
3.
195
3.1. Analyses of atmospheric water cycle components
196
3.1.1. Evaporation and precipitation
Results
197
198
The time series of the monthly domain averages of the satellite-based and merged
199
datasets for four atmospheric water cycle components during the period from 2000 to
200
2005 are shown in Figure 2. The reanalysis data MERRA is also displayed for
201
comparison. For evaporation (Figure 2a), the domain averages of the evaporation from
202
GSSTF, HOAPS, OAFlux, and MERRA are 3.88, 3.82, 3.45, and 3.50 mm d-1,
203
respectively. The monthly mean values of evaporation from the satellite-based GSSTF
204
and HOAPS are typically greater than those from the merged and reanalysis datasets.
205
Correlations between the evaporation datasets have been also investigated. Relatively
206
high correlations were found between the OAFlux and MERRA data (0.72) and the
207
GSSTF and HOAPS data (0.70), but the correlations between the HOAPS and MERRA
208
data (0.40) and between the GSSTF and MERRA data (0.30) were much weaker. The
209
time series of the four precipitation datasets are shown in Figure 2b. The period means
210
of precipitation for CMAP, GPCP, HOAPS, and MERRA are 3.14, 3.01, 2.91, and 3.23
211
mm d-1, respectively. The correlations between the satellite-based and merged datasets
212
are generally higher than those between the reanalysis MERRA and the other datasets.
213
In particular, relatively high correlations are found between the GPCP and CMAP (0.88),
214
GPCP and HOAPS (0.86), and HOAPS and CMAP (0.79) datasets.
215
216
The variances associated with the various datasets can be obtained using the following
217
equations:
218
219
 i2 
220
and
221
 m2 
1
N
1
M
 X
N
j 1
 E( X )i 
2
j ,i
(7a)
M

i 1
2
i
(7b),
222
223
where  i2 is the variance of N different datasets of the variable X for the ith month and
224
 m2 is the mean variance over M months. E ( X ) i is the average of the X’s estimated
225
from N datasets for the ith month. For three different evaporation (GSSTF, HOAPS, and
226
OAFlux) and precipitation (GPCP, CMAP, and HOAPS) datasets, the spatial
227
distributions of  m were computed for each 5˚ x 5˚ grid over the 72 month period
228
between 2000 and 2005 (Figure 3). Large variances between evaporation datasets exist
229
over some of the oceanic dry regions such as the southeastern Pacific, the subtropics
230
over the west Pacific, and the parts of the south Atlantic near South Africa. For
231
precipitation, significant variances were found over the regions of the inter-tropical
232
convergence zone (ITCZ) and the south Pacific convergence zone (SPCZ) as well as the
233
East China Sea, South China Sea, and Bay of Bengal portions of the Indian Ocean. The
234
difference in the datasets for precipitation was significantly lower over the southeastern
235
Pacific Ocean and the Atlantic Ocean. The regions with discrepancies between the
236
datasets for precipitation were distributed over a broader area than those for evaporation.
237
The range of  m for precipitation is from 0.05 to 1.65 mm d-1 and the range is from
238
0.23 to 1.23 mm d-1 for evaporation.
239
240
3.1.2. Water vapor flux convergence and water vapor tendency
241
242
Both of the domain mean time series of the XLT08 and MERRA for WVFC (Figure 2c)
243
indicated that the water vapor flux generally diverges over oceanic areas. Their period
244
mean values were negative (-0.46 mm d-1 for XLT08 and -0.55 mm d-1 for MERRA).
245
For the period between 2000 and 2005, the WVFC of the XLT08 was typically larger
246
than that of the MERRA. The correlation between the XLT08 and MERRA was 0.76
247
mm d-1. The period mean seasonal spatial distributions of the WVFCs are illustrated in
248
Figure 4. The red colored areas (positive) indicate water vapor flux convergence and the
249
blue colored areas (negative) indicate water vapor flux divergence. The major patterns
250
of the WVFCs and their seasonal variations were well matched with those of the
251
precipitation (not shown). We also noted that the XLT08 had relatively higher values
252
than the MERRA over the ITCZ and SPCZ.
253
254
The WVT derived in this study was compared with the WVT from the MERRA. The
255
WVT of the MERRA includes the analysis increment tendency of water vapor that is a
256
non-physically added value during the assimilation process in order to adjust it to the
257
observation data. The domain averaged time series showed that both of the WVTs are in
258
good agreement (Figure 2d) and have a strong correlation (0.90). The domain average
259
estimates of the WVTs were significantly smaller than those of the other components of
260
the water cycle. The mean seasonal and spatial distributions of the WVT shown in
261
Figure 5 reveal that there is a significant seasonal variation between the Northern and
262
Southern Hemispheres. The distinct negative water vapor tendencies are present in the
263
Northern Hemisphere during the winter (from December to February) and apparent
264
positive water vapor tendencies exist in the Northern Hemisphere during the summer
265
(from June to August). Both of the WVTs have relatively strong positive tendencies
266
over the oceanic regions around East Asia and the northeastern Pacific near Mexico
267
during the boreal summer (Figure 5 c and d) and strong negative tendencies over the
268
Bay of Bengal during the boreal winter (Figure 5 a and b).
269
270
3.2. Residual analysis
271
3.2.1. Domain averaged residual time series
272
273
The nine residuals estimated from the combinations of satellite-based and merged
274
datasets for atmospheric water cycle components were analyzed by taking domain
275
averages (Figure 6). The monthly time series of the domain averaged residuals (Figure
276
6a) showed that there is a distinct feature between the residuals in combination with the
277
merged evaporation of the OAFlux and the residuals in combination with the satellite-
278
based evaporation from HOAPS and GSSTF. OAFlux included residuals that fluctuated
279
between small positive and negative values near zero. The period mean values of the
280
OGXH, OCXH, and OHXH residuals were -0.01, -0.14, and 0.08 mm d-1, respectively.
281
The residuals from the HOAPS and GSSTF had relatively high positive values between
282
0.23 and 0.51 mm d-1.
283
284
The residuals of the MERRA were also analyzed in order to make a comparison. In
285
MERRA, atmospheric water balance was accomplished by adding two unphysical terms
286
[Bosilovich et al., 2011]. One of the terms was the analysis increment of water vapor
287
(ANA) that has an order of magnitude of E-P. The other term was a negative filling term
288
(F) in order to ensure positive water vapor content. Since the value of F is small enough
289
to neglect, the MERRA residual can be obtained using the following equation:
290
291

W
R  E  P  Q 
 ANA
t
(8).
292
293
In order to determine the effects of ANA on the atmospheric water budget in MERRA,
294
we also took into consideration the residual calculated by excluding ANA in Eq. (8) and
295
we called this residual MERRA (ECANA). While the MERRA residual from Eq. (8) is
296
close to zero, the MERRA (ECANA) residual has relatively large negative values
297
(Figure 6a). The period mean values of the MERRA and MERRA (ECANA) residuals
298
are 0.01 and -0.29 mm d-1, respectively. The anomaly time series of the nine estimated
299
residuals seem to be in good agreement with each other (Figure 6b). Relatively high
300
correlations exist between the residuals from identical evaporation and precipitation
301
datasets except between OGXH-GHXH (0.80). The highest correlation can be found
302
between GCXH-GGXH (0.92). The correlations between the nine estimated residuals
303
are stronger than the correlations with the MERRA residual except for OHXH-HCXH
304
(0.25) and OHXH-GCXH (0.29). The MERRA (ECANA) residual anomaly time series
305
seems to have no relationship with the others. The averaged values of the nine estimated
306
residuals and their standard deviations (error bars) are also shown in Figure 6c. The
307
period-domain means for each atmospheric water budget component and associated
308
residuals are also summarized in Table 2.
309
310
3.2.2. Period mean residual distribution
311
312
The nine estimated residuals show disagreements in their intensities and patterns for the
313
period mean spatial distributions between 2000 and 2005 (Figure 7). Positive residuals
314
(from green to red colored) indicate that the magnitudes of the source terms (E, WVFC)
315
for water vapor in respect to the atmosphere are larger than the magnitudes of the sink
316
term (P). Negative residuals (from blue to violet colored) indicate the sink terms are
317
larger than the source terms. The residuals from the combinations that include HOAPS
318
and GSSTF evaporation have more positive regions than the residuals that include the
319
OAFlux evaporation Most of the residuals have positive values due to larger WVFC
320
over some portions of the ITCZ where P is generally larger than E with the exception of
321
that for OHXH. Some of the residuals have widely spread positive values over the
322
southeastern Pacific except the residuals that include OAflux evaporation. Over the
323
SPCZ, there are distinct negative residuals especially in the residuals that included
324
CMAP precipitation. The regions where differences between the nine estimated
325
residuals exist can also be determined using Eq. 7. The distribution of the large
326
variances in the residuals coincides with those of evaporation and precipitation (Figure
327
9a).
328
329
The mean spatial distribution of the MERRA (ECANA) residuals (Figure 8a) had strong
330
positive values over the oceanic regions near Peru and strong negative values over the
331
West Pacific and the Bay of Bengal. The period mean of the MERRA (ECANA)
332
residual ranged from -2.64 to 2.70 mm d-1. The analysis increment for the WVT had
333
identical patterns with opposite signs as the MERRA (ECANA) in its period mean
334
spatial distribution (Figure 8b). The MERRA residual derived by taking the ANA into
335
consideration had relatively small values between -0.07 and 0.06 mm d-1 (Figure 8c).
336
337
A large quantity of the residual indicates an imbalance in the atmospheric water budget.
338
In order to determine the regions where large imbalances would appear, the mean
339
absolute errors were calculated from the nine estimated residuals for each grid box as
340
follows:
341
342
MAE i 
343
and
344
MAE m 
1 9
 Ri, j  0
9 j 1
(9a)
1 72
 MAEi
72 i 1
(9b),
345
346
where MAE i is the mean absolute error from the nine estimated residuals, Ri , j is j
347
th estimated residual from the satellite-based and merged datasets at month i , and
348
MAE m is the period mean absolute error. Figure 9(b) shows the spatial distribution of
349
MAE m . Relatively large magnitudes of residuals appeared around some of the coastal
350
areas and over the Arabian Sea. The imbalances were also generally large over the ITCZ,
351
SPCZ, and monsoon area where heavy precipitation occurs.
352
353
4.
Summary and Conclusions
354
355
This study examined column integrated atmospheric water balances based on the
356
analysis of residuals from various combinations of satellite-based and merged datasets
357
for atmospheric water cycle components. Satellite-based datasets for evaporation such
358
as HOAPS and GSSTF were used as well as the satellite-NWP reanalysis merged
359
evaporation dataset OAFlux. For precipitation, the satellite-gauge merged GPCP and
360
CMAP datasets as well as the satellite-based HOAPS were used. The water vapor flux
361
convergence by XLT08 and the water vapor tendency derived from the HOAPS TPW
362
were also used. The residuals of the MERRA were also analyzed for comparison.
363
364
The mean spatial distribution analysis for the period between 2000 and 2005 over
365
oceanic regions (60˚N-60˚S) showed that the satellite-based residual distribution and
366
their magnitudes varied with the datasets. The residuals from combinations including
367
HOAPS and GSSTF evaporation had relatively large positive values over the mid-
368
latitude oceanic regions due to the relatively high evaporation values of the HOAPS and
369
GSSTF over these areas. The values of the period mean standard deviations between the
370
residuals were significantly larger over the ITCZ, SPCZ, and monsoon regions and their
371
magnitudes ranged from 0.27 to 1.5 mm d-1. The magnitude of the imbalance in the
372
atmospheric water budget estimated from the satellite-based and merged datasets was
373
also generally larger over the oceanic areas with heavy precipitation such as the ITCZ,
374
SPCZ, and monsoon regions and had values ranging from 0.72 to 4.71 mm d-1 with a
375
domain average value of 1.54 mm d-1. The larger residuals over the ocean may be
376
attributed to errors in the datasets and the approximation of the water budget equation
377
without the condensed phase water component.
378
379
Meanwhile, the MERRA required artificially added a nonphysical analysis increment
380
term for the water vapor tendency in order to close the atmospheric water balance.
381
Analysis of the residuals from various combinations of datasets in this study indicated
382
that challenges remain in order to obtain an accurate atmospheric water budget.
383
Therefore, we recommend that careful consideration is used when applying datasets for
384
water budget components to water budget studies. However, similar residual anomalies
385
suggest that climate variability analysis based on the residuals may not be greatly
386
affected by a specific dataset.
387
388
389
Acknowledgments
390
This work was funded by the Korea Meteorological Administration Research and
391
Development Program under Grant CATER 2012-2063.
392
393
394
395
396
References
397
Adler, R. F., and Coauthors (2003), The version-2 global precipitation climatology
398
project (GPCP) monthly precipitation analysis (1979-present), J. Hydrometeorol.,
399
4(6), 1147-1167.
400
Andersson, A., K. Fennig, C. Klepp, S. Bakan, H. Graßl, and J. Schulz (2010), The
401
Hamburg Ocean Atmosphere Parameters and Fluxes from Satellite Data - HOAPS-
402
3, Earth Syst. Sci. Data Discuss., 3, 143–194.
403
404
405
406
Bakan, S., V. Jost and K. Fennig (2000), Satellite derived water balance climatology for
the North Atlantic: First results, Phys. Chem. Earth, 25, 121–128.
Bosilovich, M., F. Robertson, and J. Chen (2011), Global energy and water budgets in
MERRA, J. Clim., 24(22), 5721–5739.
407
Bradley, E.F., C.W. Fairall, J.E. Hare, and A.A.Grachev (2000), An old and improved
408
bulk algorithm for air-sea fluxes: COARE 2.6, A. Preprints, 14th Symp. On
409
Boundary Layer and Turbulence, Aspen, CO, Amer. Meteor. Soc., 294 – 296.
410
Chou, S.-H. (1993), A comparison of airborne eddy correlation and bulk aerodynamic
411
methods for ocean-air turbulent fluxes during cold-air outbreaks, Bound.-Layer
412
Meteor., 64, 75-100.
413
—, S.-H., E. Nelkin, J. Ardizzone, R. M. Atlas, and C.-L. Shie (2003), Surface turbulent
414
heat and momen-tum fluxes over global oceans based on the Goddard satellite
415
retrieval, version 2 (GSSTF2), J. Clim., 16, 3256-3273.
416
Draper, C., and G. Mills (2008), The atmospheric water balance over the semiarid
417
Murray-Darling
418
doi:10.1175/2007JHM889.1.
River
basin,
J.
Hydrometeorol.,
9(3),
521–534,
419
Fairall, C. W., E. F. Bradley, D. P. Rogers, J. B. Edson, and G. S. Young (1996), Bulk
420
parameterization of airsea fluxes for Tropical Ocean–Global Atmosphere Coupled–
421
Ocean Atmosphere Response Experiment, J. Geophys. Res., 101(C2), 3747–3764.
422
—, —, J. E. Hare, A. A. Grachev, and J. B. Edson (2003), Bulk parameterization of air-
423
sea fluxes: Updates and verificaiton for the COARE algorithm, J. Clim., 16, 571–
424
591.
425
Huffman, G. J., R. F. Adler, D. T. Bolvin, and G. Gu (2009), Improving the global
426
precipitation record: GPCP version 2.1, Geophys. Res. Lett., 36, L17808,
427
doi:10.1029/2009GL040000.
428
Kanamaru, H., and G. Salvucci (2003), Adjustments for wind sampling errors in an
429
estimate of the atmospheric water budget of the Mississippi River basin. J.
430
Hydrometeorol., 4, 518–529.
431
Raschke, E., and Coauthors (2001), The Baltic Sea Experiment (BALTEX): A European
432
contribution to the investigation of the energy and water cycle over a large drainage
433
basin, Bull. Am. Meteorol. Soc., 82(11), 2389 – 2413.
434
Rasmusson, E. (1967), Atmospheric water vapour transport and the water balance of
435
North America: Part I. Characteristics of the water vapour flux field, Mon. Wea.
436
Rev., 95(7), 403–426.
437
438
439
— (1968), Atmospheric water vapor transport and the water balance of North America
II. Large-scale water balance investigations, Mon. Wea. Rev., 96, 720–734.
Rienecker, M. R., and Coauthors (2011), MERRA: NASA’s Modern-Era Retrospective
440
Analysis
441
doi:10.1175/JCLI-D-11-00015.1.
442
443
444
445
446
for
Research
and
Applications,
J.
Clim.,
24,
3624–3648,
Roads, J. O., and Coauthors (2003), GCIP water and energy budget synthesis (WEBS),
J. Geophys. Res., 108, 8609, doi:10.1029/2002JD002583.
Rouse, W., and Coauthors (2003), Energy and water cycles in a high-latitude, northflowing river system, Bull. Am. Meteorol. Soc., 84, 73–87.
Ruprecht, E., and T. Kahl (2003), Investigation of the atmospheric water budget of the
447
BALTEX area using NCEP/NCAR reanalysis data, Tellus, 55A, 426–437.
448
Shie, C.-L. (2010). Science background for the reprocessing and Goddard Satellite-
449
based Surface Turbulent Fluxes (GSSTF2b) Data Set for Global Water and Energy
450
Cycle Research, In: NASA GES DISC, 18 pp, October 12, 2010, Available from <
451
http://disc.sci.gsfc.nasa.gov/measures/documentation/Science-of-the-data.pdf>
452
Stewart, R., Leighton, H., Marsh, P., Moore, G., Ritchie, H., Rouse, W., Soulis, E.,
453
Strong, G., Crawford, R. and Kochtubajda, B (1998), The Mackenzie GEWEX
454
Study: The water and energy cycles of a major North American river basin, Bull.
455
Am. Meteorol. Soc., 79(12), 2665 – 2684.
456
457
Turato, B., O. Reale, and F. Siccardi (2004), Water vapor sources of the October 2000
Piedmont flood, J. Hydrometeorol., 5, 693–712.
458
Xie, P. P., and P. A. Arkin (1997), Global precipitation: A 17-year monthly analysis
459
based on gauge observations, satellite estimates, and numerical model outputs, Bull.
460
Am. Meteorol. Soc., 78(11), 2539-2558.
461
Xie, X., W. T. Liu, and B. Tang (2008), Spacebased estimation of moisture transport in
462
marine atmosphere using support vector regression, Remote Sens. Environ., 112,
463
1846–1855.
464
Yin, X. G., A. Gruber, and P. Arkin (2004), Comparison of the GPCP and CMAP
465
merged gauge-satellite monthly precipitation products for the period 1979-2001, J.
466
Hydrometeorol., 5(6), 1207-1222.
467
468
Yu, L., and R. A.Weller (2007), Objectively analyzed air-sea heat fluxes for the global
ice-free oceans (1981– 2005), Bull. Am. Meteorol. Soc., 88, 527– 539.
469
Yu, L., X. Jin, and R. A. Weller (2008), Multidecade global flux datasets from the
470
objectively analyzed air‐sea fluxes (OAFlux) project: Latent and sensible heat
471
fluxes, ocean evaporation, and related surface meteorological variables, OAFlux
472
Tech. Rep. OA‐2008‐01, 64 pp., WHOI, Woods Hole, Mass.
473
Zangvil, A., D. Portis, and P. Lamb (2004), Investigation of the large-scale atmospheric
474
moisture field over the midwestern United States in relation to summer
475
precipitation. Part II: Recycling of local evapotranspiration and association with
476
soil moisture and crop yields, J. Clim., 17, 3283–3301.
477
478
479
480
481
482
483
484
485
486
Table 1. Characteristics of the datasets for the atmospheric water balance components.
487
The components are evaporation (E), precipitation (P), water vapor flux convergence
488
(WVFC) and water vapor tendency (WVT).
489
Datasets
Version
Variables
Spatial Resol.
Temporal Resol.
Available Period
OAFlux
v3
E
1.0˚ⅹ1.0˚
month
1958-2010
GSSTF
v2c
E
1.0˚ⅹ1.0˚
month
1987-2008
GPCP
v2.1
P
2.5˚ⅹ2.5˚
month
1979-2010
CMAP
v1001
P
2.5˚ⅹ2.5˚
month
1979-2009
HOAPS
v3
E, P, WVT
0.5˚ⅹ0.5˚
month
1987-2005
XLT08
v3
WVFC
0.5˚ⅹ0.5˚
month
1999-2008
MERRA
(GEOS-5.2.0)
0.5˚ⅹ2/3˚
month
1979-2012
E, P, WVT,
WVFC
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
Table 2. Period-domain mean values for each component of the atmospheric water cycle
505
and the corresponding residual. Averages of the satellite-based and merged datasets and
506
their standard deviations are denoted as AVGsat and STDsat, respectively. AVGall and
507
STDall also indicate averages and standard deviations for all the datasets including the
508
MERRA dataset.
E
P
WVFC
WVT
Residuals
OGXH
3.450
3.008
-0.456
-0.000
-0.014
OCXH
3.450
3.135
-0.456
-0.000
-0.141
OHXH
3.450
2.910
-0.456
-0.000
0.084
HGXH
3.822
3.008
-0.456
-0.000
0.358
HCXH
3.822
3.135
-0.456
-0.000
0.231
HHXH
3.822
2.910
-0.456
-0.000
0.456
GGXH
3.877
3.008
-0.456
-0.000
0.413
GCXH
3.877
3.135
-0.456
-0.000
0.286
GHXH
3.877
2.910
-0.456
-0.000
0.511
MERRA
3.499
3.235
-0.550
-0.000 (*0.291)
**0.005
3.499
3.235
-0.550
-0.000
-0.286
3.716
3.018
-0.456
-0.000
MERRA
(ECANA)
AVGsat
(STDsat)
(0.232)
(0.113)
AVGall
3.662
3.072
0.243
(0.224)
0.219
-0.503
(STDall)
(0.219)
(0.142)
509
*MERRA water vapor tendency analysis increment (ANA)
510
**MERRA residual including ANA term
511
512
513
514
515
516
-0.000
(0.224)
517
518
519
Figure 1. A diagram of the nine residuals estimated from possible combinations of the
520
datasets for the atmospheric water cycle components.
521
522
523
Figure 2. Time series of the monthly domain averaged evaporation rates (a),
524
precipitation rates (b), water vapor flux convergences (c), and water vapor tendencies (d)
525
for the period from 2000 to 2005. Units are mm d-1.
526
527
Figure 3. Spatial distributions of the standard deviations between three evaporation
528
datasets (OAFlux, HOAPS, and GSSTF) (a) and three precipitation datasets (GPCP,
529
CMAP, and HOAPS) (b). The values in the upper right corners of each panel indicate
530
domain averages. The minimum and maximum are also indicated in parentheses.
531
532
533
534
535
Figure 4. Mean seasonal distributions of WVFC for XLT08 in winter (DJF) (a) and in
536
summer (JJA) (c) and the mean seasonal distributions for MERRA in winter (b) and in
537
summer (d) for the period from 2000 to 2005.
538
539
540
541
542
543
544
545
546
547
548
Figure 5. Mean seasonal distributions of WVT for HOAPS in winter (DJF) (a) and in
549
summer (JJA) (c) and mean seasonal distributions for MERRA in winter (b) and in
550
summer (d) for the period from 2000 to 2005.
551
552
553
554
555
556
557
558
559
560
561
Figure 6. Time series of nine monthly domain averaged residuals (a), residual anomalies
562
(b), and the averages of the nine residuals and their standard deviations (c) for the
563
period from 2000 to 2005. Residuals from MERRA are also illustrated for comparison.
564
565
566
567
568
Figure 7. Mean spatial distributions of the nine residuals for the period between 2000
569
and 2005.
570
571
572
573
574
575
576
577
578
579
580
Figure 8. Mean spatial distributions of the residual excluding the ANA term (a) and the
581
ANA for water vapor (b) as well as the residual including the ANA term (c) for
582
MERRA.
583
584
585
586
Figure 9. Spatial distributions of the standard deviations (a) and the mean absolute
587
errors (b) from the nine estimated residuals.