difference vegetation index(Kustas and Daughtry, 1990). 
The empirical equation estimating soil heat tlux can be 
expressed as below: 
M 4-—TM3 
Gz(0325 — 0208 S TMS )- Rn ------- 13) 
TM4*- TM3 
3-4. Compute of instantaneous evapotranspira- 
tion, ET, 
From the above energy balance equation, the 
instantaneous evapotranspiration(mm / hour). £7, can 
be computed as follows: 
  
ET, = 3600 - En-G-H 14) 
L 
where L is the latent heat of vaporization 
(=[2501 - 2.37 T,] - 10°: J/ kg). 
4. COMPUTE OF DAILY EVAPOTRANSPIRATION 
As above described, the instantaneous evapotranspira- 
tion, ET, can be calculated from LANDSAT TM data at 
one time of day and in conjunction with other 
meteorological data. The daily change of radiation and 
evapotranspiration in the field assume a sine 
form(Jackson and Hatfield, 1983). 
ETER CSRILN) ener d$) 
where ET,,. is the maximum evapotranspiration of a 
day, N is the day length between sunrise and sunset, f 
is the time of LANDSAT overpass after sunrise. 
Therefore, the daily evapotranspiration(mm/day) was 
computed as follows: 
e N RN 
EL =], ET = || Ela SPAN) m --16) 
=2N-ET a / 7=2N-ET, / x-sin(xt/N) 
  
  
  
  
  
  
PES 
max 
* 
= 
= 
S 
Lal 
5e 
$ 
= 
> 
© 
sunrise t sunset 
hour 
Figure 4. Graphic representation of the estimation of 
daily evapotranspiration from instantaneous 
evapotranspiration. 
5. RESULTS AND DISCUSSION 
Table 2 summarized daily averaged surface energy 
balance at each land cover and composition ratio of land 
cover at study area from LANDSAT TM. 
Table 2 Daily averaged surface energy balance 
at each land cover 
  
le ratio(%) 
Rn(Wm?)|EcWm?) |LE(Wm2) 
  
  
  
  
  
  
  
tidal flat 20 513 155 333 
vegetated 39 306 290 191 
nonvegetated 41 — 470 145 83 
  
Figure 4 shows the estimated evapotranspiration 
according to method of chapter 3 and 4. The latent 
heat flux which equal to evapotranspiration in water 
balance is greater on the tidal flat where contains lots of 
water in the soil. nonvegetated surfaces such as urban 
and reclamated area shows the half amount of latent 
heat rather than vegetated surfaces(Table 2). In 
nonvegetated areas, there is a little amount of 
evapotranspiration, but vegetated area especially in 
forest. shows the high evapotranspiration rate 
(> 2mm/day). 
According to the national development program. in 
the study area, west coasts of Korea. the land cover/land 
use change is very severe. This situation can be seen in 
Figure 3. The tidal flat(or intertidal zone) was 
reclamated to use of industrial complex, construction of 
new residence section. : 
As a result. those land use/land cover changes will 
induce the potential change of thermal environment. For 
example, when existing tidal flat will reclamated, and 
changes to non vegetated land surface, the effect on 
surface energy balance become: 
ARn - - 43: reclamated surface area 
AH z-10 reclamated surface area 
À LE = -250- reclamated surface area 
As known this results, the change of latent heat flux 
due to change of land use/land cover is greater than that 
of net radiation and sensible heat flux. Decreasing net 
radiation(increase of urban area, decrease of tidal flat or 
vegetated surfaces) due to increase of surface albedo 
works on lowering surface temperature. In the other 
hand, decrease of latent heat which is greater at least 
six times than decreasing net radiation and sensible heat 
flux works on highering surface temperature. 
In this study, the author estimates of the 
evapotranspiration using LANDSAT TM data with 
ancillarv ground-based meteorological data and topolo- 
34 
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B4. Vienna 1996 
Siamo