164
does vary on very short time scales (of no particular interest to us here
since the atmosphere cannot respond, especially to ultraviolet radiation,
on these short time scales) and varies also with the 11 year sunspot number
(Williams and Hudson, 1988). The magnitude of this decadal variation is
on the order of lor 2 Wjm 2 compared to the solar constant of 1367 Wjm 2 ,
a variation of order 0.1% .
It should be noted that the direction of output variability with sunspot
number is opposite to what intuition tells us: when the sun's disc is clear
(no sunspots) the output is reduced, not increased. This occurs because
the so-called bright features (networks, plages, and faculae) covary with
sunspots and dominate the output variability. Due to interactions in the
sun's upper atmosphere, of order 20% of this variability with sunspot cycle
lies in the ultraviolet and therefore gets absorbed by ozone relatively high in
the earth's atmosphere. The basic question therefore is whether or not the
troposphere is sensitive to the remaining one or so Wjm 2 : this is a general
question about the sensitivity of the climate system to small changes of
radiation. Because the radiation reaching the intercepted disc at the outer
edge of the earth's orbit must be spread over the entire area of the earth,
this number must be divided by 4 and because the earth's albedo in the
visible is .3 (mostly due to clouds) the number must further be multiplied
by 0.7. Thus a solar output variability of 1 Wjm 2 translates to a variability
of 0.2 Wjm 2 at the top of the global atmosphere.
On the basis of CO2 doubling experiments with numerical GCMs, it has
been found that 4 Wjm 2 radiative input (in this case long wave) at the top
of the atmosphere will lead to a warming of global surface temperature from
1.5 to 4.5°C (e.g. IPCC 90), giving a sensitivity of 0.4 to 1.FCW- 1 m 2 to
radiation at the top of the atmosphere. If we accept this model estimate of
climate sensitivity, it is clear that the solar output variability over decadal
time scales could be responsible for a few tenths of a degree response in
globally averaged surface temperature.
The existence of long periods when sunspots were absent from the face of
the sun (for example, the so called "Maunder Minimum," in the late 1600's,
see Eddy, 1976) are well established. To estimate the reduction of solar
output implied by sunspot minima, and analysis of the observed variability
of other main sequence stars similar to the sun has been made (Lean,
Skumanich, and White, 1992) and implies a reduction of solar output as
much as 5Wjm 2 (with a more probable value of half that) which at the
top of the atmosphere would be less than 1Wjm 2 . These estimates are
does vary on very short time scales (of no particular interest to us here
since the atmosphere cannot respond, especially to ultraviolet radiation,
on these short time scales) and varies also with the 11 year sunspot number
(Williams and Hudson, 1988). The magnitude of this decadal variation is
on the order of lor 2 Wjm 2 compared to the solar constant of 1367 Wjm 2 ,
a variation of order 0.1% .
It should be noted that the direction of output variability with sunspot
number is opposite to what intuition tells us: when the sun's disc is clear
(no sunspots) the output is reduced, not increased. This occurs because
the so-called bright features (networks, plages, and faculae) covary with
sunspots and dominate the output variability. Due to interactions in the
sun's upper atmosphere, of order 20% of this variability with sunspot cycle
lies in the ultraviolet and therefore gets absorbed by ozone relatively high in
the earth's atmosphere. The basic question therefore is whether or not the
troposphere is sensitive to the remaining one or so Wjm 2 : this is a general
question about the sensitivity of the climate system to small changes of
radiation. Because the radiation reaching the intercepted disc at the outer
edge of the earth's orbit must be spread over the entire area of the earth,
this number must be divided by 4 and because the earth's albedo in the
visible is .3 (mostly due to clouds) the number must further be multiplied
by 0.7. Thus a solar output variability of 1 Wjm 2 translates to a variability
of 0.2 Wjm 2 at the top of the global atmosphere.
On the basis of CO2 doubling experiments with numerical GCMs, it has
been found that 4 Wjm 2 radiative input (in this case long wave) at the top
of the atmosphere will lead to a warming of global surface temperature from
1.5 to 4.5°C (e.g. IPCC 90), giving a sensitivity of 0.4 to 1.FCW- 1 m 2 to
radiation at the top of the atmosphere. If we accept this model estimate of
climate sensitivity, it is clear that the solar output variability over decadal
time scales could be responsible for a few tenths of a degree response in
globally averaged surface temperature.
The existence of long periods when sunspots were absent from the face of
the sun (for example, the so called "Maunder Minimum," in the late 1600's,
see Eddy, 1976) are well established. To estimate the reduction of solar
output implied by sunspot minima, and analysis of the observed variability
of other main sequence stars similar to the sun has been made (Lean,
Skumanich, and White, 1992) and implies a reduction of solar output as
much as 5Wjm 2 (with a more probable value of half that) which at the
top of the atmosphere would be less than 1Wjm 2 . These estimates are