85th AMS Annual Meeting

Tuesday, 11 January 2005: 4:45 PM
Global Runoff Dataset Produced by GSWP2
Naota Hanasaki, The university of Tokyo, Tokyo, Japan; and S. Kanae and T. Oki
Poster PDF (791.0 kB)

The second Global Soil Wetness Project (GSWP2) is an ongoing scientific activity that aims at primary producing state-of-the-art global datasets of land surface fluxes, state variables and related hydrologic quantities by integrating uncoupled land surface schemes. As of 31st July 2004, 12 institutes have submitted their baseline simulation (Bucket, CLM2-TOP, ISBA, LaD, MOSES2, NOAH, NSIPP-CATCH, ORCHIDEE, SSiB, SiBUC, SWAP and VISA). In this report, preliminary result of their intercomparison is shown using global, 10-year average fluxes and state variables components.

Prior to the analysis, the outputs are closely examined whether energy and water is balanced. First, energy components are well balanced for all models. Water components, however, are not fully balanced for several models. It may be caused by recharge of soil moisture from below the bottom soil layer since most of them increases water amount in their system.

The energy and water balance is shown in Table 1 and Table 2. Even using global, 10-year average, energy and water partitioning has large discrepancy between models. Bowen Ratio (= Sensible Heat / Latent Heat) ranges from 0.473 (SWAP) to 2.767 (NOAH), the average of 12 schemes is 0.948. Runoff Ratio (= Total Runoff / Total Precipitation) ranges from 0.32 (SWAP, NSIPP-CATCH) to 0.70 (NOAH). Table 1 Global energy balance components (unit: W/m3, global, 10-year average)

SWnet

Lwnet

lE

H

G

Qfusion

Qsublim

Qrain

Balance

BUCK

144.7

-77.2

41.0

24.2

1.4

1.3

-0.4

0.0

-0.1

CLM2

128.8

-66.1

35.9

26.4

0.3

1.1

0.2

0.0

-1.3

ISBA

142.8

-57.6

37.4

45.6

0.2

0.0

1.1

0.0

1.0

LaD_

141.0

-64.9

42.0

31.5

0.0

1.2

1.3

0.0

-0.1

MOSE

142.4

-68.2

37.5

35.5

0.0

1.2

0.0

0.0

0.0

NOAH

143.0

-66.7

19.0

52.6

-0.3

4.8

-3.2

0.0

3.4

NSIP

143.3

-60.4

44.1

39.6

-0.2

1.4

-0.6

-1.2

-0.2

ORCH

133.5

-64.3

30.3

37.6

0.1

0.9

-1.3

0.0

1.5

SSiB

143.7

-63.4

31.7

47.4

1.3

0.9

0.1

0.0

-1.0

SWAP

141.5

-74.4

45.5

21.5

0.0

0.8

-0.7

0.0

0.0

SiBU

139.5

-67.1

36.4

25.2

8.6

0.8

0.0

0.0

1.4

VISA

151.2

-69.9

44.0

34.2

2.1

0.9

0.1

0.0

0.0

Mean

141.3

-66.7

37.1

35.1

1.1

1.3

-0.3

-0.1

0.4

Table 2 Global water balance components (unit: km3/year, global, 10-year average)

Rainf

Snowf

Evap

Qs

Qsb

Balance

BUCK

97774

14654

70865

43610

0

-2047

CLM2

93766

8688

62165

3842

36588

-142

ISBA

99921

13902

66344

15060

32561

-142

LaD_

99722

13909

74729

0

38716

186

MOSE

99923

13902

65066

16306

32376

76

NOAH

97774

14654

31847

2590

75840

2152

NSIP

96730

13244

74352

20152

15339

131

ORCH

96865

13514

51031

3746

71168

-15566

SSiB

105979

7967

55700

37446

20926

-126

SWAP

104603

9220

77158

10938

25654

73

SiBU

102621

11182

65540

27761

25648

-5146

VISA

99921

13902

76116

12380

30110

-4782

Mean

99633

12395

64243

16153

33744

-2111

To validate outputs, total runoff (= Surface Runoff + Subsurface Runoff) of each model is integrated using a global runoff routing model, namely TRIP, and compared with 270 river gauging stations globally for 1987-88. Most of models tend to overestimate discharge in North America, Europe to western Siberia and Africa around Gulf of Guinea, in other hand; slightly underestimate in eastern Siberia to Southeast Asia. It may be attributed to the characteristics of forcing precipitation data. Root mean square error (RMSE) is smaller for models with smaller runoff ratio, since the simulation has generally overestimation tendency.

In the annual meeting, both regionally and temporally detailed analyses and their background mechanism will be presented.

Supplementary URL: