Trends and variations in global stratospheric temperatures are an integral
part of the changes occurring in the Earth's climate system. Data sets for
analyzing long-term (a decade and more) changes in the stratospheric temper
atures consist of radiosonde, satellite, lidar, and rocketsonde measurement
s; meteorological analyses based on radiosonde and/or satellite data; and p
roducts based on assimilating observations using a general circulation mode
l. Each of these contain varying degrees of uncertainties that influence th
e interpretation and significance of trends, We review the long-term trends
from approximately the mid-1960s to the mid-1990s period. The stratosphere
: has, in general, undergone considerable cooling over the past 3 decades.
At northern midlatitudes the lower stratosphere (similar to 16-21 km) cooli
ng over the 1979-1994 period is strikingly coherent among the various data
sets with regard to magnitude and statistical significance. A substantial c
ooling occurs in the polar lower stratosphere during winter-spring; however
, there is a large dynamical variability in the northern polar region. The
vertical profile of the annual-mean stratospheric temperature change in the
northern midlatitudes over the 1979-1994 period is robust among the differ
ent data sets, with similar to0.75 K/decade cooling in the similar to 20- t
o 35-km region and increasing cooling above (e.g., similar to2.5 K/decade a
t 50 km). Model investigations into the cause or causes of the observed tem
perature trends are also reviewed, Simulations based on the known changes i
n species' concentrations indicate that the depiction of lower stratospheri
c ozone is the major radiative factor in accounting for the 1979-1990 cooli
ng trend in the global, annual-mean lower stratosphere (similar to0.5 to 0.
6 K/decade), with a substantially lesser contribution by the well-mixed gre
enhouse gases. Ozone loss is also an important causal factor in the latitud
e-month pattern of the lower stratospheric cooling trend. Uncertainties ari
se due to incomplete knowledge of the vertical profile of ozone loss near t
he tropopause. In the middle and upper stratosphere, both well-mixed greenh
ouse gases and ozone changes contribute in an important manner to the cooli
ng, but model simulations underestimate the observed decadal-scale trend. W
hile there is a lack of reliable information on water vapor changes over th
e 1980s decade, satellite measurements in the early to middle 1990s indicat
e increases in water vapor that could be a significant contributor to the c
ooling of the global lower stratosphere.