IAPRS & SIS, Vol.34, Part 7, "Resource and Environmental Monitoring", Hyderabad, India,2002 
  
    
2.1 Approach 
A physical retrieval scheme (Appendix I) developed by 
Tanahashi et al. (2001) for deriving surface insolation on 
operational basis from VISSR-GMS-5 in Japan is adopted to 
Meteosat data for present study. The different steps are given 
below. 
1).Cloud detection and characterization was found out by 
bispectral threshold technique using visible and thermal band 
data from a time series of images of 30 days in December 1998. 
2) Latitude and longitudes corresponding to METEOSAT lines 
and pixels were generated using algorithm given by 
EUMETSAT. 
3) Solar Zenith angle 0) layer was generated from latitude, 
. longitude, Julian day of the year and time of the day. 
4) Atmospheric correction was carried out with respect to 
atmospheric costitituents such as: ozone, water vapor, aerosol 
which are contributing to atmospheric turbidity. 
5) The daily total surface insolation was computed by Simpson 
integration of hourly insolation at five different times in a day. 
The inputs to this model are listed below: 
i) A time series of images of visible, thermal IR bands for 
particular acquisition time 
ii) Day of the year and time in a day 
iii) Solar zenith angle 
iv) Cloud attenuation coefficients for different cloud classes 
V) Average ozone content 
vi) Total atmospheric precipitable water 
vii) Average Angstrom turbidity parameters to determine 
aerosol transmittance in a season 
The surface insolation (Sg) is derived by the following 
equation. 
Ss = (Sı+Sp+Sa)(1-2.A) <<) 
Where, cloud attenuation coefficients (a) to insolation were 
derived from look-up table of cloud top albedo (A) and 
brightness temperature (TBB) given by (Tanahashi et al., 2001). 
Direct (S;) and diffuse irradiation components due to Rayleigh 
(Sg) and aerosol (SA) scattering were found out. 
Se=S;tSptSa tone BUS Sy. 0 0 n. or. ror 77 (2) 
SESTO OT dn A er où CT (3) 
SE = ST (0. TR 1 5; 1t: vp ae vui (4) 
S. zSlolmi-a.Eowa(- TA). fon nali mad L^ (5) 
S-Lecos0 |... :,, , —, T  . -— (6) 
Where, S is the total surface insolation. 
e = correction for sun-earth distance , which is 0.9674 on July 5 
and 1.0344 on January 3. 
Io7solar constant (1367 Wm?) 
Transmittance due to absorption (To) by ozone was estimated 
by the following equation 
To71-0.02118(um)/(1--0.042(um)43.23x10 *(um)?)-1.082u/ 
(1+138.6(um))°*°*- 0.0658(um)/(1+(103.6(um))) ^ — -— (7) 
Where, m is air mass= 1/ (cos0+0.15(93.885-6)"!2**) 
u = Ozone content in atm-cm.The mean value of ozone is 0.3 
atm-cm during December over india 
Transmittance due to Rayleigh (Tg ) scattering was computed 
as function of central wavelength of visible band. 
Transmittance due to attenuation by aerosols(TA ) was 
estimated by the equation below. 
T,=0.10445a-0.0162+(1.003-0.125)xexp(-Bm(1.0890 
505123) 0° fémote sensing, China, 15-2! Noveraa: (8) 
a,=2.9wm/((1+141.5wm)0.635 + 5.925 wm) | . -—— (9) 
where a, « and p are water vapor absorption coefficient and 
Angstrom's turbidity parameters. Values are available from 
Iqbal's table (Iqbal, 1983) 
tw *' is precipitable water in cm. The NOAA TOVS (Tiros 
operational vertical sounder) derived total precipitable water 
(19x19) was used under present study. 
3. RESULTS AND DISCUSSION 
3.1 Sensitivity analysis 
Since the different atmospheric parameters play major role to 
influence the absorption and scattering processes of shortwave 
incoming radiation, the sensitivity of surface insolation output 
and direct to diffuse insolation ratio to a given change within 
theoretical ranges of total precipitable water (w), angstrom 
turbidity parameters for aerosol transmittance,o and p, was 
studied. The results are presented in figures 1a,1b and 1c. 
Sensitivity : atmospheric precipitable water (w) 
1000 
  
  
  
  
  
  
  
  
(a) 
w=0.0 
15 
SE 900 | e 
z € 
g =0.5 2 
$ ais 3 
© me EST Po aime FFT 1Œ 
$ = 
z ren CICR E 
© 800 4 a 
E 
w=0.0,0.5,1.0,2.0,3.0 L5 
total insolation 
* * * ? direct/diffuse 
700 0 
330 335 340 345 350 355 360 365 370 
Julian day 
Fig 1. (a) Sensitivity analysis of insolation retrieval scheme to 
different atmospheric inputs 
    
au eus. a 43 A SEAT SEA 
TI 
pr 
ur 
ab 
de 
tu 
re 
be 
re 
de 
CO 
in 
Fc 
av 
D 
or 
hc