Oki et al., Hydrol. Proc., 1995

1
Relative change of total terrestrial water
storage can be estimated from water vapor
convergence and river discharge information based
on atmospheric water balance.
(Oki et al., Hydrol. Proc., 1995)
2
Atmospheric Water Balance -?HQ?P-E
(Oki, et. al, Hydrol. Proc., 1995)
3
Zonal Mean Annual Freshwater Transport
(Oki, et. al, 1995)
4
(Oki, Global Energy and Water Cycles, 1999)
5
Global Soil Wetness ProjectData set production,
comparison, and validation

• GSWP is an ongoing GEWEX project, led from COLA
  (P. Dirmeyer) and IIS/UT (T. Oki), involving over
  a dozen modeling groups on four continents.
• GSWP1 used the ISLSCP I-1 data to examine
  1987-1988.
• GSWP2, a 10-year (1986-1995) global land-surface
  analysis, will begin next year
• Regional studies are investigating issues of
  aggregation, sub-grid variability, and
  assimilation of remote sensing data.

Unlike the ocean, the land surface state
variables (soil moisture, snow depth and
coverage, soil temperature) are not routinely
measured and reported. Land-surface models
driven by observed meteorology give us a means to
generate a surrogate observed data set.
(Dirmeyer, et. al, 1998, BAMS)
6
Goals of GSWP2

• Produce the best global data sets of soil
  moisture, surface fluxes, and related hydrologic
  quantities (including runoff) for 1986-1995 with
  1x1 degree grid spacing considering the
  uncertainties associated with
• Land surface models more than 10 LSMs
• Model parameters 2 kinds of vegetation etc.
• Forcing data 3 kinds of reanalysis data,
  corrected precipitation, etc.
• Temporal and spatial scales to run LSMs

7
Tc Mc gc
Vegetation Canopy
T temperature (K) e vapor pressure (Pa) u wind
speed (m s-1) c CO2 partial pressure (35 Pa) o O2
partial pressure (2090 Pa) z reference height
(m) IS solar radiation (W m-2) IL thermal
infrared radiation (W m-2) P precipitation rate
(mm) p atmospheric surface pressure (101300 Pa)
Tg Mg W1
Surface Layer
W2
Root Zone
Recharge Zone
Td W3
8
(No Transcript)
9
Offline Simulation of LSMs
10
From Runoff by Land Surface Model To River
Discharge through TRIP
11
(No Transcript)
12
Simulated Monthly River Discharge
(Oki, et. al, 1999, JMSJ)
13
Global River Discharge Simulations

• Runoff Estimates by Land Surface Models (LSMs)
• Designed for Providing Lower boundary of GCMs
• Estimates balances of energy, water, carbon,
• Offline simulation with observed forcing data
• River Discharge LSM TRIP Routing
• ?Runoff from offline simulations by LSMs
• ?Estimates by Climate Models
• Validation and the error characteristics

14
Global Runoff Database

• Collaboration with Global Runoff Data Centre
  (GRDC)
• Major gauging stations in major river basins in
  the world
• Non-unified observational period
• Quality Check
• Detection of Anthropogenic Effect
• Correction of human errors

15
(No Transcript)
16
(No Transcript)
17
(Oki and Musiake, 1999)
18
(No Transcript)
19
(No Transcript)
20
Based on Observation Only
(Oki, 1999)
21
Global PUB no tuning global parameters only
(Oki, et. al, 1999, JMSJ)
22
All the models are biased. ?forcing issue?
(Oki, et. al, 1999, JMSJ)
23
Lower raingauge Network density ?low accuracy
Garbage in ?Garbage out!!
(Oki, et. al, 1999, JMSJ)
24
Spatial Coverage of TRMM and GPM
(Oki, et. al, 1999, JMSJ)
25
(No Transcript)
26
Annual Runoff Error by Latitude
(Figure 15 of Oki et al., 1999.)
27
Whats Gauge Correction?

• 1. Precipitation gauge cannot catch 100 of
  particles under the strong wind condition.
• 2. Systematic error depends also on the gauge
  shape, attachment height (see Sevurk and Klemm,
  1989), besides wind speed.
• 3. It becomes a big error especially in case of
  snowfall.

28
Sevurk and Klemm, 1989
There are many kinds of gauges in the World !!
29
Gauge Catch Ratio Versus Wind Speed
30
Sources of Uncertainties

• Runoff
• annual mean
• low flow
• 100 year flood

Environmental Info. Precip., radiation, temp.,
humidity, wind, (Atmospheric Forcing)
Theory, Equation, or Model
Regional Info. Soil, Vegetation,Topography, Basin
area, (parameters)

• Solutes
• quality
• sediment
• isotopic ratio

Initial/Boundary Conditions
Specific Temporal and Spatial Scales
Applicability/Accuracy may differ region by
region.
31
Validation of GSWP-2 1m Column Soil Moisture
(IL, GSMDB)
Total Field (1m, 1986-1995)
Anomaly (1m, 1986-1995)
(Dirmeyer, et. al, 2005)
32
Global Terrestrial Water Budget
Unit mm/year
Legend
103
771
Inter-annual range
108
71
771
727
513
765
89
727
765
388
600
706
Snow
489
706
Rainfall
Average of 12 models (1986-1995)
ET
Inter-model range
111
253
242
21
362
471
96
Surface runoff
98
117
633
209
307
Subsurface runoff
163
Total runoff
51
343
160
Soil water storage
(Oki, et. al, 2005)
33
Global Composition of Evapotranspiration
Unit mm/year
Legend
513
Inter-annual range
771
600
388
489
727
765
706
ET
Average of 12 models (1986-1995)
Inter-model range
Es
Ew
Et
Ei
85
145
31
81
225
344
114
77
219
17
213
28
0
16
188
15
369
27
180
175
(Oki, et. al, 2005)
34
Water Balance in variousClimatic Zones

• Classifying each grid box in climatic zones by
  Budykos Aridity Index (BAI)
• BAI (ET/P) / (Rn/lP)
• Arid 4.0 lt BAI, Semi arid 2.0 lt BAI lt 4.0
• Semi humid 1.2 lt BAI lt 2.0, humid 0.7 lt BAI lt
  1.2
• 0.0 lt BAI lt 0.7
• Tropical humid P gt 2000 mm/y and latitude lt 30
  degree
• Very humid the rest
• Ice cover BAI lt 0.0

35
arid
semi-arid
semi-humid
humid
Tropic h
ice
cold h
(scatter plot for GSWP1 data for example)
36
Climatic Zones delineated by Budykos Aridity
Index (BAI)
(Budyko Aridity IndexRn/lP)
(Oki, et. al, 2005)
37
Mean Annual Precipitation
(Oki, et. al, 2005)
38
Mean Partitioning Ratio ()
(Oki, et. al, 2005)
39
Mean Water Balance (mm/y)
Base Flow
Surface Runoff
ET
(by GSWP2 Multi-model Ensemble)
(Oki, et. al, 2005)
40
Mean Evapotranspiration (mm/y)
Evap fromWater Surface
Interception E
Transpiration
Bare Soil E
(by GSWP2 Multi-model Ensemble)
(Oki, et. al, 2005)