Multiperiodicity, modulations, and flip-flops in variable star light curves III. Carrier fit analysis of LQ Hydrae photometry for 1982-2014

Aims. We study LQ Hya photometry for 1982-2014 with the carrier fit (CF) method and compare our results to earlier photometric analysis and recent Doppler imaging maps. Methods. As the rotation period of the object is not known a priori, we utilize different types of statistical methods first (least-squares fit of harmonics, phase dispersion statistics) to estimate various candidates for the carrier period for the CF method. Secondly, a global fit to the whole data set and local fits to shorter segments are computed with the period that is found to be optimal. Results. The harmonic least-squares analysis of all the available data reveals a short period, of close to 1.6 days, as a limiting value for a set of significant frequencies. We interpret this as the rotation period of the spots near the equatorial region. In addition, the distribution of the significant periods is found to be bimodal, hinting of a longer-term modulating period, which we set out to study with a two-harmonic CF model. A weak modulation signal is, indeed retrieved, with a period of roughly 6.9 yr. The phase dispersion analysis gives a clear symmetric minimum for coherence times lower than and around 100 days. We interpret this as the mean rotation pattern of the spots. Of these periods, the most significant and physically most plausible period statistically is the mean spot rotation period 1.(d)60514, which is chosen to be used as the carrier period for the CF analysis. With the CF method, we seek any systematic trends in the spot distribution in the global time frame, and locally look for previously reported abrupt phase changes in rapidly rotating objects. During 2003-2009, the global CF reveals a coherent structure rotating with a period of 1.(d)6037, while during most other times the spot distribution appears somewhat random in phase. Conclusions. The evolution of the spot distribution of the object is found to be very chaotic, with no clear signs of an azimuthal dynamo wave that would persist over longer timescales, although the short-lived coherent structures occasionally observed do not rotate with the same speed as the mean spot distribution. The most likely explanation of the bimodal period distribution is attributed to the high-and low-latitude spot formation regions confirmed from Doppler imaging and Zeeman Doppler imaging.