Context. Each solar activity cycle is characterized by a small number of superactive regions (SARs) that produce the most violent of space weather events with the greatest disastrous influence on our living environment. Aims. We aim to re-parameterize the SARs and study the latitudinal and longitudinal distributions of SARs. Methods. We select 45 SARs in solar cycles 21-23, according to the following four parameters: 1) the maximum area of sunspot group, 2) the soft X-ray flare index, 3) the 10.7 cm radio peak flux, and 4) the variation in the total solar irradiance. Another 120 SARs given by previous studies of solar cycles 19-23 are also included. The latitudinal and longitudinal distributions of the 165 SARs in both the Carrington frame and the dynamic reference frame during solar cycles 19-23 are studied statistically. Results. Our results indicate that these 45 SARs produced 44% of all the X class X-ray flares during solar cycles 21-23, and that all the SARs are likely to produce a very fast CME. The latitudinal distributions of SARs display the Maunder butterfly diagrams and SARs occur preferentially in the maximum period of each solar cycle. Northern hemisphere SARs dominated in solar cycles 19 and 20 and southern hemisphere SARs dominated in solar cycles 21 and 22. In solar cycle 23, however, SARs occurred about equally in each hemisphere. There are two active longitudes in both the northern and southern hemispheres, about 160• -200• apart. Applying the improved dynamic reference frame to SARs, we find that SARs rotate faster than the Carrington rate and there is no significant difference between the two hemispheres. The synodic periods are 27.19 days and 27.25 days for the northern and southern hemispheres, respectively. The longitudinal distribution of SARs is significantly non-axisymmetric and about 75% SARs occurred near two active longitudes with half widths of 45• .

Context. The vector magnetic field characteristics of superactive regions (SARs) hold the key for understanding why SARs are extremely active and provide the guidance in space weather prediction. Aims. We aim to quantify the characteristics of SARs using the vector magnetograms taken by the Solar Magnetic Field Telescope at Huairou Solar Observatory Station. Methods. The vector magnetic field characteristics of 14 SARs in solar cycles 22 and 23 were analyzed using the following four parameters: 1) the magnetic flux imbalance between opposite polarities; 2) the total photospheric free magnetic energy; 3) the length of the magnetic neutral line with its steep horizontal magnetic gradient; and 4) the area with strong magnetic shear. Furthermore, we selected another eight large and inactive active regions (ARs), which are called fallow ARs (FARs), to compare them with the SARs. Results. We found that most of the SARs have a net magnetic flux higher than 7.0×10 21 Mx, a total photospheric free magnetic energy higher than 1.0 × 10 24 erg cm −1 , a magnetic neutral line with a steep horizontal magnetic gradient (≥300 G Mm −1 ) longer than 30 Mm, and an area with strong magnetic shear (shear angle ≥80• ) greater than 100 Mm 2 . In contrast, the values of these parameters for the FARs are mostly very low. The Pearson χ 2 test was used to examine the significance of the difference between the SARs and FARs, and the results indicate that these two types of ARs can be fairly distinguished by each of these parameters. The significance levels are 99.55%, 99.98%, 99.98%, and 99.96%, respectively. However, no single parameter can distinguish them perfectly. Therefore we propose a composite index based on these parameters, and find that the distinction between the two types of ARs is also significant with a significance level of 99.96%. These results are useful for a better physical understanding of the SAR and FAR.