The Moon is - by far - the closest important celestial body to the Earth. People have suspected for centuries that the Moon influences the climate here on Earth. In more recent times, however, attention has switched to understanding how human activity influences weather and climate, and interest in the Moon's effects has waned.

In a report in the 1 June issue of Geophysical Research Letters, Randall S. Cerveny and Robert C. Balling Jr, of Arizona State University, Tempe, Arizona, seek to redress the balance. They demonstrate a strong relationship between the phase of the Moon and the range of temperatures experienced throughout a 24-hour day (the Diurnal Temperature Range, or DTR.)

Over the past fifty years, the DTR has decreased by about half a degree. Conventional wisdom blames this on the greenhouse effect. But this decrease is just a trend observed in data that vary over shorter timescales. Cerveny and Balling show that for the period between 1950 and 1995, the DTR fluctuates with the phases of the Moon. It tends to increase towards Full Moon, and tends to be lowest at New Moon. Simple monthly differences in DTR between New Moon and Full Moon may be as much as 0.309 °C - in other words, 60% of the entire 50-year decrease. The message should be clear: all possible sources of variation should be investigated before blaming human activity alone for observed changes in climatic parameters.

The researchers look at two possible causes for the observed increase in DTR close to the Full Moon. One is that moonlight - the solar radiation reflected by the Moon - could actually warm the Earth at this time of the month. This effect, which will be greatest at Full Moon, is small, but appreciable when calculated over the course of a month. However, it will only have the effect of increasing night-time temperatures.

More significant is a connection between the phase of the Moon and the overall position of the Earth-Moon system with respect to the Sun. It seems that the Earth is slightly closer to the Sun at Full Moon than at New Moon, and will therefore receive slightly more solar radiation during daylight hours, increasing maximum temperatures and thus DTR as a whole.

The Moon exerts an even more subtle influence, according to a separate report from Charles P. Sonett of the University of Arizona, Tucson, and Leonard A. Smith of the University of Oxford, UK, also published in Geophysical Research Letters. Sonett and Smith's story starts far from the Moon - indeed, it begins by looking downwards, at the problems of archaeologists, rather than upwards at the Moon.

Archaeologists date organic remains, such as wood and bones, by a technique called carbon-14 dating. Carbon-14 is a radioactive form, or 'isotope', of ordinary carbon. It is made by the interaction of high-energy cosmic rays from space with atoms in the Earth's upper atmosphere. Once on the ground, it becomes incorporated into living things, but this incorporation ceases when organisms die. The steady rate at which carbon-14 decays radioactively thereafter is used to estimate the time when the organism concerned was last alive. Archaeologists, therefore, are naturally interested in factors that might modulate the steady 'rain' of carbon-14 onto the Earth's surface, as uncorrected variation distorts dating estimates. The fact is that carbon dating does not reflect a simple one-on-one match with the calendar - archaeologists need to use special calibration tables, which in turn are based on understanding the physical and historical factors that influence carbon-14 flux.

One such influence is the 'solar wind', a stream of particles from the Sun that deflects cosmic rays. In principle, this should lead to a net decrease in carbon-14 flux. Sonett and Smith show that this flux could be additionally modulated if the flux of solar wind is blocked by the Moon - in other words, if the Earth falls under the Moon's solar-wind 'shadow'.

The effect of this is likely to be small; however, it could explain a pronounced peak, seen every 17.5 years, in the geophysicists' standard record of carbon-14 flux. The only cycle that fits the bill is the so-called 'Saros', a cyclic variation in the tilt of the Moon's orbit around the Earth, relative to the Earth's orbit around the Sun. At two points in the cycle, the so-called 'nodes', the planes coincide. Solar eclipses - in which the Moon blots out the Sun - are common at such points. But as Sonett and Smith explain, the Moon could also eclipse the Earth in another way, by shadowing it from the solar wind just enough to affect carbon-14 flux.