The purpose of the present communication is to identify the short-term (few tens of months) periodicities of several solar indices (sunspot number, CaII area and K index, Lyman α, 2800 MHz radio emission, coronal green-line index, solar magnetic field). The procedure used was: from the 3-month running means (3m) the 37-month running means (37m) were subtracted, and the factor (3m − 37m) was examined for several parameters. For solar indices, considerable fluctuations were seen during the ± 4 years around sunspot maxima of cycles 18−23, and virtually no fluctuations were seen in the ± 2 years around sunspot minima. The spacings between successive peaks were irregular but common for various solar indices. Assuming that there are stationary periodicities, a spectral analysis was carried out which indicated periodicities of months: 5.(QTO). The periodicities of 1.3 year (15.6 months) and 1.7 years (20.4 months) often mentioned in the literature were seen neither often nor prominently. Other periodicities occurred more often and more prominently. For the open magnetic flux estimated by Wang, Lean, and Sheeley (2000) and Wang and Sheeley (2002), it was noticed that the variations were radically different at different solar latitudes. The open flux for <45 • solar latitudes had variations very similar (parallel) to the sunspot cycle, while open flux for >45 • solar latitudes had variations anti-parallel to the sunspot cycle. The open fluxes, interplanetary magnetic field and cosmic rays, all showed periodicities similar to those of solar indices. Many peaks (but not all) matched, indicating that the open flux for <45 • solar latitudes was at least partially an adequate carrier of the solar characteristics to the interplanetary space and thence for galactic cosmic ray modulation.

The relationship between El Niño and droughts in north‐east Brazil is poor. From 46 El Niño events (strong and moderate) during 1849–1992, only 21 (ca. 46 per cent) were associated with droughts at Fortaleza, Ceara (north‐east Brazil) and only 26 (ca. 57 per cent) with negative deviations (of any size). Thus, forecasts of droughts based on the appearance of El Niño alone would be wrong half the time. Instead, predictions based on significant periodicities (ca. 13 and ca. 26 years) give reasonably good results. Data up to 1978 could predict the drought that occurred during 1979– 1983, although the severe drought of 1983 was attributed to a strong El Niño. Mild droughts during 1991–1993 were also foreseen. In future, severe droughts are predicted for 2000–2010 in north‐east Brazil. However, the prediction skill is not very good and predictions should be checked by analysis of as much updated data as possible. Predictions beyond about 5 10 years are not reliable. The significant periodicities in North‐east Brazil rainfall series might be related to periodicities in Atlantic parameters (Sea‐surface temperatures, etc.) that are well correlated with north‐east Brazil rainfall. © 1997 by the Royal Metrological Society.

For many purposes (e.g., satellite drag, operation of power grids on Earth, and satellite communication systems), predictions of the strength of a solar cycle are needed. Predictions are made by using different methods, depending upon the characteristics of sunspot cycles. However, the method most successful seems to be the precursor method by Ohl and his group, in which the geomagnetic activity in the declining phase of a sunspot cycle is found to be well correlated with the sunspot maximum of the next cycle. In the present communication, the method is illustrated by plotting the 12-month running means aa(min) of the geomagnetic disturbance index aa near sunspot minimum versus the 12-month running means of the sunspot number Rz near sunspot maximum [aa(min) versus Rz(max)], using data for sunspot cycles 9 -18 to predict the Rz(max) of cycle 19, using data for cycles 9 -19 to predict Rz(max) of cycle 20, and so on, and finally using data for cycles 9 -23 to predict Rz(max) of cycle 24, which is expected to occur in 2011 -2012. The correlations were good (∼+0.90) and our preliminary predicted Rz(max) for cycle 24 is 142 ± 24, though this can be regarded as an upper limit, since there are indications that solar minimum may occur as late as March 2008. (Some workers have reported that the aa values before 1957 would have an error of 3 nT; if true, the revised estimate would be 124 ± 26.) This result of the precursor method is compared with several other predictions of cycle 24, which are in a very wide range (50 -200), so that whatever may be the final observed value, some method or other will be discredited, as happened in the case of cycle 23.

[1] Since the work of Snyder et al. (1963), who showed a possible link between interplanetary solar wind speed V and geomagnetic index Kp, such a relationship has been examined by many workers and found to be rather loose. In the present communication this relationship is rechecked for all data during . It was noted that moderate or strong geomagnetic storms occurred only when solar wind speed V was above $350 km/s. However, above this limit, the plots of Dst versus V showed a large scatter, and any value of V could be associated with any value of Dst in a wide range of a factor of $2, or any Dst value could be associated with any value of V in a wide range of a factor of $2, indicating a poor relationship between V and Dst. The scatter could be partly because not V but VB s (product of V and the southward component B s of interplanetary field B) is the appropriate variable relevant for Dst changes. This was checked. In the plot of Dst versus VB s it was noticed that the scatter was smaller and correlation better than that in the plot of Dst versus V. Since the relationship between V and Dst is poor, an estimate of V with some antecedence, as is done in a present-day prediction scheme ) V estimated from the lateral extension speed of side halo coronal mass ejections (CMEs) and assuming V 2 proportional to Dst magnitudes) is not likely to give reliable estimates of Dst magnitudes. However, estimate of V could certainly be useful to estimate the time when the storm would hit the Earth but remembering that 15% of the halo CMEs may miss the Earth.

Abstract. The global evolutions of f oF2 anomalies were examined for three very intense geomagnetic storms, namely the Halloween events of October