International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B7. Istanbul 2004
vegetation cover method of Valor and Caselles (1995). Albedo,
land surface temperature, emissivity, and vegetation index
(NDVI), in conjunction with land surface in-situ observed data
and atmospheric PBL data, are applied within the surface
energy balance system model (SEBS) to calculate the actual
evapotranspiration. The needed PBL (Planetary Boundary
Layer) data for calculating sensible heat flux (see Section 2.1.3)
include PBL height (m), PBL wind speed (m/s), PBL relative
humidity, PBL potential air temperature (K), PBL air pressure
(Pa) and surface air pressure (Pa) and were obtained by tether
balloon sounding or radiosonde.
4. VERIFICATION
In order to check whether the SEBS model is acceptable for
estimating daily actual evapotranspiration or not, we compared
ground station-measured evaporation with SEBS mode-
estimated evaporation for non-vegetative (i.e., no transpiration)
land surface. Two places, Yongjing weather station near
Liujiaxia reservoir and Qianyang weather station near
Fengjiashan reservoir, are selected for ET verification.
Liujiaxia Reservoir, about 5 km away from Yongjing weather
station, covers an area of 130 km?. Hongjiashan reservoir, 3 km
away from Qianyang station covers an area of 10 km”.
According to the observational data from these two weather
stations, conversion factor of evaporation from 0.2-m diameter
pan to 5-m diameter pan is 0.469. We assume that the 5-m
diameter evaporation pan reasonably represents the reservoir
evaporation condition. That is
E, = E, =0.469*E,, (22)
where E. is reservoir actual ET, E; ET of 5-m pan, Fos ET
of 0.2-m pan. The data from the tow stations are available for
the entire year of 2002 and one day per month for the year
(totally 12 days from 12 months corresponding the remote
sensing data dates) was selected for the purpose of ET
verification. Table 1 shows that the absolute errors between the
estimated values and observation values of these two places are
relatively small, varying from —0.43 to 0.88. The relative errors
are mostly less than 20%.
Table 1, Figs.4 and 5 demonstrate that the SEBS model-
estimated values for the Liujiaxia and Fengjiashan Reservoira
are in reasonably good agreement with the station-measured
values at the nearby weather stations, suggesting that the remote
sensing-aided SEBS modeling can achieve an acceptable level
of accuracy in estimating actual evapotranspiration in semiarid
regions.
5. CONCLUSIONS AND DISCUSSIONS
Our experimental study to estimate ET in the Western Chinese
Loess Plateau shows that an acceptable estimation of ET in
semi-arid area can be achieved at regional scales by using
NOAA/AVHRR data-derived parameters as the input variables
in simulating the surface energy balance. The spatial
distribution of the estimated ET shows that ET is the highest in
the well-vegetated high elevations and major water bodies
(white and pink areas in Fig 3), intermediate in hilly plateau
areas (green and blue areas in Fig. 3), and the lowest in low-
elevation valleys and basins and also in northern desert areas
(black areas in Fig. 3). The regional-scale distribution shows
that ET in the western Chinese Loess Plateau is primarily
determined by vegetation coverage, which is in turn determined
by soil water availability. This study suggests that the spatial
distribution in ET is primary modulated by the temperature
difference and humidity difference between land surface and
nearly surface atmosphere. The results show that in the areas
where soil moisture is higher and vegetation coverage is better,
the energy transfer between land surface and nearly surface
atmosphere is accomplished through latent heat flux, while in
the lowlands where soil moisture is lower and vegetation
coverage is poorer, the energy transfer is primary through
sensible heat flux.
ACKNOWLEDGEMENT
This research is financially supported by a grant from “The
Project of Retrieving Remote Sensing Key Land Parameters” of
“Hi-tech Research and development program of China, 863
program"(Grant No.:2001AA135110) and "Key Teacher
Program" of the Chinese Education Ministry to Feng Zhaodong
(Lanzhou University, 2000) on his project: "Watershed
Hydrology and Landscape Ecology in the Gansu Loess Plateau:
GIS and RS-assisted Modeling Approach"(Grand No.:ETD-
2000-65)
REFERENCES
I. Albert Rango and Ahlam I. Shalaby, 1998a, Operational
applications of remote sensing in Hydrology :success,
prospects and peoblems, Hydrological Sciences,
43(6):947-969
2. Bastiassnssen, W.G.M.,. Pelgrum, H., Wang, J., Ma, Y.,
Moreno, J.F., Roerink, G.J., and Van der Wal, 1998a, A
remote sensing surface energy balance algorithm for land
(SEBAL) 2. Validation, Journal of Hydrology | 213:213-
229 :
3. BeckerF. & Z-L.Li , 1990a, Surface temperature and
Emissivity at various scales: definition,measurements and
related problems, Remote Sens. Rev. 12::225-253
4. Bob Z. Su, 2000, Remote sensing of land use and
vegetation for mesoscale hydrological studies, /nr. J.
Remote Sensing, 21:213-233
5. Bob Z. Su, A surface Energy Balance System (SEBS) for
estimation of turbulent heat fluxes from point to
continental scale. In Bob Z. Su and C. Jacobs (eds.), 2001b,
Advanced Earth Observation-Land Surface Climate,
Publication of the National Remote Sensing Board, USP-2,
01-01, pp 183 7
6. Boni, G., Entekhabi, D., and Castelli, F, 2001a, Land Data
Assimilation with Satellite Measurements for the
Estimation of Surface Energy Balance Components and
Surface Control on Evaporation;, Water Resour. Res. 37:
1713-1722
7. Coll, C. and. V. Caselles, 1997 b, A split-window
algorithm for land surface temperature from AVHRR data,
validation and algorithm comparison,JGR,
1317
