Sir

We should make use of new technology to improve on the poor state of knowledge of the constancy of the Sun's output. In 1994, the National Research Council of the US National Academy of Sciences called for improved accuracy in measurements that monitor the Sun's output. In its report1, the council stated that, between 1980 and 1986, an estimated 0.1% decrease in the output of the Sun as part of the 11-year cycle just cancelled the effect on the climate produced by anthropogenic greenhouse-gas emission in the same period. This highlights the accuracy required to separate long-term anthropogenic effects from natural ones.

In the past, accurate absolute radiometry, even on the ground, was difficult, and the spread of the results from different satellite radiometers over the 20 years that total solar irradiance has been monitored from space is nearly 0.5% (Fig. 1a). These differences are consistent with an uncertainty of about 0.4% associated with each radiometer2. No useful measurements were obtained from the one set of radiometers recovered after flight (from the SOVA satellite). In trying to interpret these data, the experimenters took the only route possible and applied adjustments to the various data sets to normalize their absolute levels. Figure 1b shows the resulting self-consistent data set, which appears to show that the underlying output of the Sun has remained stable to within 0.02% over the past 20 years. But how reliable is this conclusion?

Figure 1
figure 1

a, Data from measurements of total solar irradiance made from satellites. b, Composite total solar irradiance obtained from the data shown in a by making adjustments to normalize the absolute levels of successive data sets. For details see ref. 2.

Full size image

It is well known in metrology that it is not possible to say by how much a measured quantity is drifting over a long period unless measurements are made at the beginning and end of the period relative to a standard that is demonstrably stable, that is, one directly linked to the fundamental constants of physics. From the data in Fig. 1, taking account of the uncertainties in the calibration of the radiometers and those inherent in the normalization procedures, we believe that an uncertainty of not less than 0.3% must be associated with the apparent null underlying variation in the output of the Sun shown in Fig. 1b. In view of the sensitivity of climate change to systematic variations in solar output, such an uncertainty is not good enough.

Although national metrology institutes now have absolute radiometers using new technology with accuracies below 0.01%, they have not been used for these important measurements. We call on the solar physics and Earth resources communities to collaborate with these institutes to draw up a long-term programme of absolute radiometry in space, using the best technology, with the aim of improving the accuracy of these crucial measurements by at least an order of magnitude. We owe this to future generations, especially to climatologists and those who will use the results of climate studies.

We must be able to rely on the results of all measurements related to climate studies. They provide the essential basis for advice to governments in formulating policy, the financial and human consequences of which will be enormous. Such measurements must be made in SI units, which are firmly linked to the constants of physics.