ion
tified
ators
ly
ous
s, but
equire
at
)OGy
bodies
- the
int
the
‘ea,
licated
iluated
3 data
°lby et
1 be
slant
its.
Lxed
(five)
nd-pass
cted by
nd. c)
RAOBS)
RAN
rough
tics
wave-
The atmospheric correction program finds the value of the true
temperature for a given indicated temperature and zenith angle by iterative
calculation with different target temperatures until the equation is balanced
to a specified residual error. Program output is an array of indicated vs.
corrected temperatures for zenith angles of O to 50 degrees.
Magnitude of the Atmospheric Correction
The magnitude of the atmospheric correction depends mainly on three
factors: 1) the overall humidity of the atmosphere, 2) the temperature of the
moistest layers of the atmosphere, and 3) the temperature difference, or
contrast between the surface target and the moistest layers of the atmosphere.
Atmospheric humidity and temperature are highly variable parameters and
consequently the magnitude of the atmospheric correction also has a
considerable range.
Atmospheric corrections for 102 SRT temperature analyses of the Great
Lakes in 1981 have been examined, in order to ascertain their magnitude and
range over a one year period. The average correction for all analyses taken
together was +1.5 , with a standard deviation of 1.3- The positive average
correction conforms with the fact that on a yearly average the atmosphere as a
whole is colder than the water surfaces. The largest positive corrections
occur in late summer and in the fall when surface water temperature is high,
but the atmosphere can occasionally be quite cold. Negative corrections can
occur in the spring and in early summer, at times when very warm and humic air
masses overlie cold water surfaces.
As can be seen in Table 1, the magnitude of the correction increases from
1.2 at the low end of the indicated temperature range, to 3:59 at the high
end. The change in the correction, AC - vs, is due to the change in surface
versus air temperature difference over the range of indicated temperatures. In
the set of 102 Great Lakes temperature analyses in 1981, the largest AC was
5.906, in a case where the correction was -1.2° for_an indicated temperature of
7.2° and +4.7° for an indicated temperature of 19.1 . The average ^C for all
analvses was 1.89 with a standard deviation of 1.4 .
CONCLUSIONS
A method has been developed for operational retrieval of surface water
temperatures from l'OAA satellite channel 4 data. The method has a demonstrated
capability to provide temperatures with operationally acceptable accuracy under
various atmospheric conditions. Application of the method is limited to inland
water bodies and coastal zones where radiosonde data are regularly available.
REFERENCES
Fortuna, J.J. and L.N. Hambrick, 1974. The operation of the NOAA polar
satellite system. NOAA Techn. Memo., NESS 60, Washington, D.C., 127 pp.
Irbe, J.G. 1972. An operational program for measuring surface water
temperature by airborne radiation thermometer (ART) survey. Proc. lst
Cndn. Symp. Remote Sensing, 183-200, Feb. 1972, Ottawa, Ontario.
Irbe, J.G., R.K. Cross, and A. Saulesleja, 1981. Remote sensing of surface
water temperature on the Great Lakes and off the Canadian Atlantic coast.
NAFO SCR Doc. 81/IX/112, Northwest Atlantic Fisheries Organization,
Scientific Council Studies, in press, 1982.
723
