Sparse paleointensity data from 10–50 ka suggest that the average dipole moment (DM) was 50–75% of the average of 8.67 μ 1022 A m2 for the past 5 Ma, and 8.75 μ 1022 for the past 12 ka. A linear ramp function, increasing the DM from 4 to 8.75 μ 1022 A m2 between 50–10 ka BP, generates a total 14C inventory of 126 dpm/cme2, agreeing very well with an inventory assay of 128 dpm/cme2, which includes 14C in sediments. With the Lingenfelter and Ramaty (1970) production function and a model DC gain of about 100, this DM function would give a Δ14C of 500‰ at 20 ka BP, consistent with the Barbados coral record, and also gives a good match to the Holocene record. A Laschamp geomagnetic event at about 45 ka BP, with a DM of 25% of its average value and lasting 5 ka, would only increase the present inventory by 0.3–1.2 dpm/cme2, and would probably have only a small effect on Δ14C at 20 ka BP, but could produce a short-lived 14C spike of over 500‰.

Solar activity, as expressed by interplanetary solar wind magnetic field fluctuations, modulates the atmospheric production of 14C. Variations of atmospheric 14C can be precisely established from the cellulose within annual tree rings, an independently dated conservative archive of atmospheric carbon isotopes. Δ14C time series interpretation shows that solar activity has varied with a recurrence period of 2115 ± 15 (95% confidence) yr (Hallstattzeit) (Damon & Sonett 1991) over the past 7160 yr. From a non-stationary oscillation solar activity hypothesis, 52 possible spectral harmonics may result from this period. Damon and Sonett (1991) identify powerful harmonics such as the 211.5-yr (Suess) and the 88.1-yr (Gleissberg) cycles as independent fundamental periods. These stronger harmonics appear to modulate the 11-yr (Schwabe) sunspot cycle. Variations in the solar magnetic field, thus, may respond to longer period variations of the solar diameter envelope (Ribes et al. 1989). Such variations would affect solar radiative energy output and, consequently, change total solar irradiance (Sofia 1984).

Variability of solar magnetic activity manifested within sunspot cycles demonstrates features of chaotic behavior. We have analyzed cosmogenic nuclide proxy records for the presence of the solar activity signals. We have applied numerical methods of nonlinear dynamics to the data showing the contribution of the chaotic component. We have also formulated what kind of cosmogenic nuclide data sets are needed for investigations on solar activity.

We present here annual data on 14C abundance in tree rings during the Maunder minimum of solar activity (ad 1645–1715). We show that the solar modulation persisted during the minimum. We also compare these data with measurements of 10Be concentration in dated polar ice cores and with records of aurorae recurrence during this time interval.

Atmospheric 14C concentrations vary with time and latitude. These variations, measured directly on atmospheric samples, or in independently-dated organic material such as tree rings, supply data essential for the calibration of dynamic models of the global carbon cycle. Short variations in the production rate of atmospheric 14C are strongly attenuated in the relatively large atmospheric CO2 reservoir. In-situ production of 14C should be negligible for ages up to 80 ka bp. Background problems in AMS dating are more likely attributable to contamination of very small samples.

We discuss the nature of the twofold record of cosmogenic nuclides in ice sheets, of nuclei produced in the atmosphere, and of nuclei produced in situ due to interactions of cosmic-ray particles with oxygen nuclei in ice. We show that a wealth of geophysical information, in principle, can be derived from a suitable combination of nuclides in ice deposited at different latitudes. Such knowledge includes temporal changes in the cosmic-ray flux, in the geomagnetic field and in climate. The rate of deposition of cosmogenic atmospheric nuclei in ice depends on the global cosmic-ray flux and a host of climatic factors. The global cosmic-ray flux, in turn, depends on the level of solar activity, and of the geomagnetic dipole field. Thus, the task of deconvolution of the record of cosmogenic nuclides is difficult, but can be facilitated by considering the recently discovered record of in-situ-produced cosmogenic 14C, whose production rate at high latitudes is independent of the geomagnetic dipole field (Lal 1992b). We also present a brief review of work done to date and new prospects for deciphering geophysical records using ice sheets.

We find no evidence for an anomalously intense 11-yr cycle in Δ14C at high latitudes during the period, AD 1870–1885, as reported by Fan et al. (1983, 1986). However, there does appear to be a regional effect within the MacKenzie River region (67°N, 130°W), with atmospheric 14C depressed by relative to the Olympic Peninsula. Such an effect would require only 5% of CO2 in the air mass to have been derived from 5% 14C-depleted soil gas CO2. This small but apparently significant regional effect could be caused by accumulation of CO2 within the frozen earth followed by outgassing during the spring thaw. The short growing season would enhance the effect by allowing insufficient time for global atmospheric equilibration.

Analysis of the Vostok ice-core record of 10Be (Raisbeck et al. 1987) suggests that the sharply resolved increases in 10Be at 35 ka (kyr) and 60 ka are due to cosmic-ray (CR) increases. As an alternate to long-term solar modulation or strong decreases in the Earth's magnetic field, supernova (SN) forcing is qualitatively consistent with the generation of a forward-reverse shock ensemble from a spherical blast wave of age very approximately at 75 ka. This age agrees with Davelaar, Bleeker and Deerenberg's (1980) identification of 75 ka for the age of a North Polar Spur SN remnant. Confirmation would be the first geochemical detection of supernova forcing of spallogenic and perhaps cosmogenic isotope production in the atmosphere. The three 10Be increases can be satisfied by a modification of the Sonett, Morfill, and Jokipii (1987) model. This consists of 2 or 3 shock waves from a single SN event, which includes the first stage in the expansion, leading to a forward shock, S1+, and a pair of reverse waves, S1− and S2−. One reverse wave arises from the spherical expansion, itself, and the other is a reflected wave from a remnant precursor shell boundary from a more ancient SN. The model requires the solar system to be immersed in the ‘bubble’ of the earlier post-SN evolution, possibly affecting estimates of heliospheric boundary distance. However more recent analysis of Camp Century ice core data discloses only the 35 ka 10Be peak. This recent result compounds the difficulty of constructing a completely consistent model for the source of the Vostok spikes. This paper is written in the spirit of suggesting only one of possibly several different models, even within the subclass of SN models.

We describe our method of determining solar cosmic-ray flux and spectrum in the past, based on the comparison of different cosmogenic isotopes. For the period, AD 1781–1950, we have detected several intervals with a high probability of powerful solar flares.

A study of the solar proton event data between 1954 and 1986 indicates that the large fluence events at the Earth are usually associated with a sequence of solar activity and related geomagnetic storms. This association appears to be useful to infer the occurrence of major fluence proton events extending back to 1934, albeit in a non-homogeneous manner. We discuss the possibility of identifying major solar proton events prior to 1934, using geomagnetic records as a proxy.

To understand better the Earth's climate, we need to know precisely how much radiation the Sun generates. We present here a simple physical mechanism describing the convective processes at the time of low sunspot activity. According to this model, the kinetic energy increased during the Maunder Minimum, causing a decrease of the solar radiation that was sufficient to produce a little Ice Age.

We propose an original method for analysis of low-concentrations of stable and long-lived radioactive nuclides. We discuss the parameters of the main features of the “Trace” spectrometer (a multicharged-ion laser source, a highly sensitive time-of-flight mass spectrometer, a charge-exchange chamber and a mass spectrometer for positive and negative single-charged ion analysis). We also compare these features with conventional AMS devices.