High-precision timing of 42 millisecond pulsars with the European Pulsar Timing Array

International audienceWe report on timing, flux density, and polarimetric observations of the transient magnetar and 5.54 s radio pulsar XTE J1810-197 using the GBT, Nancay, and Parkes radio telescopes beginning in early 2006, until its sudden disappearance as a radio source in late 2008. Repeated observations through 2016 have not detected radio pulsations again. The torque on the neutron star, as inferred from its rotation frequency derivative f-dot, decreased in an unsteady manner by a factor of 3 in the first year of radio monitoring. In contrast, during its final year as a detectable radio source, the torque decreased steadily by only 9%. The period-averaged flux density, after decreasing by a factor of 20 during the first 10 months of radio monitoring, remained steady in the next 22 months, at an average of 0.7+/-0.3 mJy at 1.4 GHz, while still showing day-to-day fluctuations by factors of a few. There is evidence that during this last phase of radio activity the magnetar had a steep radio spectrum, in contrast to earlier behavior. There was no secular decrease that presaged its radio demise. During this time the pulse profile continued to display large variations, and polarimetry indicates that the magnetic geometry remained consistent with that of earlier times. We supplement these results with X-ray timing of the pulsar from its outburst in 2003 up to 2014. For the first 4 years, XTE J1810-197 experienced non-monotonic excursions in f-dot by at least a factor of 8. But since 2007, its f-dot has remained relatively stable near its minimum observed value. The only apparent event in the X-ray record that is possibly contemporaneous with the radio shut-down is a decrease of ~20% in the hot-spot flux in 2008-2009, to a stable, minimum value. However, the permanence of the high-amplitude, thermal X-ray pulse, even after the radio demise, implies continuing magnetar activity