1 Introduction

BL Lacertae objects, together with quasars with flat radio spectrum, belong to the class of active galactic nuclei called blazars. These are objects characterized by extreme emission variability at all wavelengths, noticeable emission at the $\gamma$-ray energies, important polarization, superluminal motion, brightness temperatures by far exceeding the Compton limit (see e.g. Urry 1999 for a review).

The general scenario for the blazar emissivity has long been set: it foresees a plasma jet coming out from a supermassive, rotating black hole surrounded by an accretion disk. In the jet, which is closely aligned with our line of sight, relativistic electrons produce soft photons (from radio to the UV-X-rays) through synchrotron emission, and high-energy photons (up to the TeV energies) via inverse Compton scattering. However, theoretical models differ in the jet structure and physics, and in the origin of the seed photons that are energized to the $\gamma$-ray energies. Constraints on the theoretical models can come from dense optical and radio monitoring, and simultaneous observations throughout all the electromagnetic spectrum, from the radio to the $\gamma$-ray band. Indeed, multiwavelength variability studies (see Ulrich et al. 1997 for a review) can give information on the compactness of the emitting regions, and verify the existence of correlations and time delays among the emissions in different bands. This in turn can shed light on the location of the emitting regions in the jet and on the nature of the seed photons that are comptonized. International collaborations among optical astronomers such as the OJ-94 Project and the Whole Earth Blazar Telescope (WEBT) were born in the last years to make the observational effort more efficient.

In this paper we present the results of a wide collaboration aimed to study the optical and radio variability of the BL Lac object AO 0235+16. Partial and very preliminary results were presented in Villata et al. (1999), where the light curve in the R band from January 1993 to January 1999 was shown and radio-optical correlations during that period analyzed by means of the Discrete Correlation Function (DCF).

The source AO 0235+16 is a well known blazar at z=0.94 that exhibits spectacular emission variations on both short ($\sim$day) and long (months, years) time scales.

It was classified as a BL Lac object by Spinrad & Smith (1975), who observed a $\sim$2 mag variation. Noticeable optical outbursts were described by Rieke et al. (1976), Pica et al. (1980), Webb et al. (1988), Webb & Smith (1989), Webb et al. (2000). From the data reported in the literature, an overall brightness variation of more than $5\rm ~mag$ can be seen, which makes AO 0235+16 one of the most interesting sources for variability studies.

As for the short-term optical behaviour, intraday variability was observed in a number of occasions (Heidt & Wagner 1996; Noble & Miller 1996; Romero et al. 2000). In particular, changes of $0.5\rm ~mag$ were measured by Romero et al. (2000) within a single night, and variations of more than $1\rm ~mag$ in about 24 hours.

Many optical data were taken in the period 1993-1996 by the astronomers participating in the international collaboration called OJ-94 Project, who observed AO 0235+16 as a "complementary" object, in addition to OJ 287 and 3C 66A. The results of their monitoring were published in Takalo et al. (1998). Some of the groups belonging to the OJ-94 Project have then continued to observe AO 0235+16.

A huge monitoring effort has been pursued in the radio band by the University of Michigan Radio Observatory (UMRAO), in the USA, and by the Metsähovi Radio Observatory, in Finland. Indeed, this object is of extreme interest also in the radio band. Long-term radio light curves at frequencies above $1\rm ~GHz$ have shown that it exhibits frequent, relatively well-resolved, high-amplitude events (O'Dell et al. 1988; Clements et al. 1995). This makes AO 0235+16 particularly suitable for studies of correlated activity. Moreover, several recent papers have argued that unusually high Doppler factors/large Lorentz factors may be present: Jorstad et al. (2001) infer a Lorentz factor greater than 45 based on $43\rm ~GHz$ proper motions, using VLBA observations during 1995 and 1996; a Doppler factor for the bulk flow of order 100 was inferred from radio short-term variability in an earlier paper by Kraus et al. (1999). Doppler factors larger than 80 (and Lorentz factors $\Gamma > 60$) were found by Fujisawa et al. (1999) by applying the inverse-Compton effect and equipartition models to VLBI observations at $22\rm ~GHz$. Recent VSOP observations of AO 0235+16 by Frey et al. (2000) have led to an estimate of $T_{\rm b} > 5.8 \times 10^{13} \rm ~ K$ for the rest-frame brightness temperature of the core, which is the highest value measured with VSOP to date and implies a Doppler factor of $\sim$100.

Optical and radio data on AO 0235+16 taken in the last four years (1996-2000) by a wide international collaboration are presented in Sect. 2, and inserted in the source long-term light curves, dating back from the mid seventies. The results of the Whole Earth Blazar Telescope (WEBT) first-light campaign on AO 0235+16 in autumn 1997 are also included. Optical spectra are derived and spectral changes discussed in the same section. Radio light curves are shown in Sect. 3, where the behaviour of the radio flux is analyzed side by side with the optical one. Statistical analysis is presented in Sect. 4: autocorrelation function and Discrete Fourier Transform (DFT) methods are applied to search for characteristic time scales of variability in both the radio and optical domains; DCF analysis is then performed to check the existence of radio-optical correlations. A final discussion is contained in Sect. 5.