factor and the increased atmospheric attentuation. In this case, a light 
meter reads higher than exposure would indicate and more exposure is re- 
quired. Several methods, such as the use of the Kodak Aerial Exposure 
Computer and exposure meter calculations, are discussed in detail by 
Harris (1968) in the Manual of Color Aerial Photography. 
Despite the success in using the empirical approach discussed above, 
a more quantitative instrument still needs to be developed. It would be 
particularly helpful to have an infrared sensitive detector (such as 
silicon, fig. 2) which would measure the depth of the atmosphere from 
the air. Such a detector could be used to calibrate each waveband re- 
sponse of LANDSAT data if high altitude flights were made to coincide 
with LANDSAT overpasses. Use of such a detector should also improve the 
consistency of exposing CIR film. 
Improvements in satellite sensors 
Now that the user community has had an opportunity to work with sat- 
ellite data, there are some obvious improvements which could be made. 
1. One of the principal factors impeding the use of LANDSAT data 
in operational resource applications is the delay from data 
acquisition to delivery to the user. Currently, when ordering 
from the EROS Data Center, Sioux Falls, S.D., which is the pri- 
mary distribution center for LANDSAT data, depending on the 
type of product desired, the user can expect to receive data 
from one to two months after acquisition. For many applications 
such as water condition monitoring, crop inventories, or disaster 
assessment, such a delay is clearly unacceptable. This is well 
understood by those responsible for LANDSAT data distribution, 
and efforts are underway to develop a process which will provide 
data on a more timely basis. A system for the production of 
photographic negatives and computer compatible tapes at the EROS 
Data Center directly from high density digital tapes provided by 
NASA is already planned and designed. Hopefully, this will ex- 
pedite the availability of the finished products. In all fair- 
ness, however, it should be remembered that the LANDSAT program 
was begun only four years ago as an experiment, not as an oper- 
ational program, and that we are now in the transition.  Un- 
doubtedly, the next few years will witness a definite improve- 
ment in data "turnaround" time. 
Somewhat of an impasse in the development of an operational 
system does exist, however. On the one hand, user organizations 
are reluctant to commit money and manpower to the full utili- 
zation of LANDSAT data until they can be reasonably certain that 
data will be continually available well into the future and on a 
timely basis. On the other hand, some difficulty exists in ob- 
taining funds for the complete development of a completely oper- 
ational data acquisition and distribution system until significant 
operational use of the data has been demonstrated. Obviously, 
to resolve this problem, greater understanding by both groups is 
necessary. 
Based on airborne MSS data, narrower wavebands (about 0.044m) on