oned in FURTHER STUDIES TO BE CONDUCTED 
  
  
1e shot" 
okkaido. The study mentioned above is still ratherin its initial stage. 
nt of the Following steps should be taken for further analysis on the 
N. Only applicability of the DMSP/OLS imagery to estimate 
estimate population: 
(a) Elimination of thin clouds "one shot" imagery of 
Hokkaido, which may have obscured the imager 
in the north westem part of Ilokkaido region, 
using the thermal infrared imagery of the same 
DMSP/OLS data. 
(b) Application of additional data-sets such as road rai! 
network and slope, to find out if these would 
contribute to improve the accuracy of estimating 
population, in particular for small square grids. 
(c) Evaluation of difference in performance, between 
"one shot" imagery and "synthesized" city lights 
data, in estimating population. 
(d) Applicauon of an imagerv taken in the low- 
sensitivity mode, which was conducted 
expenmentally through 1996 and 1997, to avoid 
ast of saturation in urban areas (e.g. around Sapporo city 
in Hokkaido and large cities in the east coast of 
U.S.A). 
"his case 
racy in 
¢ study REFERENCE 
> studies 
s subject Croft, T. A., 1978. Nighttime images of the earth from 
space, Scientific America, 239: pp. 68-79 
Elvidge, C. D. et. al., 1997. Mapping City Lights With 
zs Nighttime Data From the DMSP Operational Linescan 
System, Photogrammetric Engineering and Remote 
Sensing, 63: pp. 727-734. 
Elvidge, C. D. et. al., 1997. Satellite inventory of human 
settlements using nocturnal radiation emissions: A 
contribution for the global toolchest, Global Change 
ov Th Biology 3: pp. 387-395. 
on for 
ty lights 
:nt the 
he one 
, but it 
International Archives of Photogrammetry and Remote Sensing. Vol. XXXII, Part 7. Budapest, 1998 459