We present high-precision timing observations spanning up to nine years for 37 millisecond pulsars monitored with the Green Bank and Arecibo radio telescopes as part of the North American Nanohertz Observatory for Gravitational Waves (NANOGrav) project. We describe the observational and instrumental setups used to collect the data, and methodology applied for calculating pulse times of arrival; these include novel methods for measuring instrumental offsets and characterizing low signal-to-noise ratio timing results. The time of arrival data are fit to a physical timing model for each source, including terms that characterize time-variable dispersion measure and frequency-dependent pulse shape evolution. In conjunction with the timing model fit, we have performed a Bayesian analysis of a parameterized timing noise model for each source, and detect evidence for excess lowfrequency, or "red," timing noise in 10 of the pulsars. For 5 of these cases this is likely due to interstellar medium propagation effects rather than intrisic spin variations. Subsequent papers in this series will present further analysis of this data set aimed at detecting or limiting the presence of nanohertz-frequency gravitational wave signals.

We report on an effort to extract and monitor interstellar scintillation parameters in regular timing observations collected for the NANOGrav pulsar timing array. Scattering delays are measured by creating dynamic spectra for each pulsar and observing epoch of wide-band observations centered near 1500 MHz and carried out at the Green Bank Telescope and the Arecibo Observatory. The ∼800-MHz wide frequency bands imply dramatic changes in scintillation bandwidth across the bandpass, and a stretching routine has been included to account for this scaling. For most of the 10 pulsars for which the scaling has been measured, the bandwidths scale with frequency less steeply than expected for a Kolmogorov medium. We find estimated scattering delay values that vary with time by up to an order of magnitude. The mean measured scattering delays are similar to previously published values and slightly higher than predicted by interstellar medium models. We investigate the possibility of increasing the timing precision by mitigating timing errors introduced by the scattering delays. For most of the pulsars, the uncertainty in the time of arrival of a single timing point is much larger than the maximum variation of the scattering delay, suggesting that diffractive scintillation remains only a negligible part of their noise budget.

We present an analysis of regular timing observations of the high magnetic field Rotating Radio Transient (RRAT) J1819−1458 obtained using the 64‐m Parkes and 76‐m Lovell radio telescopes over the past 5 years. During this time, the RRAT has suffered two significant glitches with fractional frequency changes of 0.6 × 10−6 and 0.1 × 10−6. Glitches of this magnitude are a phenomenon displayed by both radio pulsars and magnetars. However, the behaviour of J1819−1458 following these glitches is quite different to that which follows glitches in other neutron stars, since the glitch activity resulted in a significant long‐term net decrease in the slow‐down rate. If such glitches occur every 30 years, the spin‐down rate, and by inference the magnetic dipole moment, will drop to zero on a time‐scale of a few thousand years. There are also significant increases in the rate of pulse detection and in the radio pulse energy immediately following the glitches.

Among efforts to detect gravitational radiation, pulsar timing arrays are uniquely poised to detect "memory" signatures, permanent perturbations in spacetime from highly energetic astrophysical events such as mergers of supermassive black hole binaries. The North American Nanohertz Observatory for Gravitational Waves (NANOGrav) observes dozens of the most stable millisecond pulsars using the Arecibo and Green Bank radio telescopes in an effort to study, among other things, gravitational wave memory. We herein present the results of a search for gravitational wave bursts with memory (BWMs) using the first five years of NANOGrav observations. We develop original methods for dramatically speeding up searches for BWM signals. In the directions of the sky where our sensitivity to BWMs is best, we would detect mergers of binaries with reduced masses of 10 9 M out to distances of 30 Mpc; such massive mergers in the Virgo cluster would be marginally detectable. We find no evidence for BWMs. However, with our non-detection, we set upper limits on the rate at which BWMs of various amplitudes could have occurred during the time spanned by our data-e.g., BWMs with amplitudes greater than 10 −13 must occur at a rate less than 1.5 yr −1 .

We observed the Crab pulsar with the 43 m telescope in Green Bank, WV over a timespan of 15 months. In total we obtained 100 hr of data at 1.2 GHz and seven hours at 330 MHz, resulting in a sample of about 95,000 giant pulses (GPs). This is the largest sample, to date, of GPs from the Crab pulsar taken with the same telescope and backend and analyzed as one data set. We calculated power-law fits to amplitude distributions for main pulse (MP) and interpulse (IP) GPs, resulting in indices in the range of 2.1-3.1 for MP GPs at 1.2 GHz and in the range of 2.5-3.0 and 2.4-3.1 for MP and IP GPs at 330 MHz. We also correlated the GPs at 1.2 GHz with GPs from the Robert C. Byrd Green Bank Telescope (GBT), which were obtained simultaneously at a higher frequency (8.9 GHz) over a span of 26 hr. In total, 7933 GPs from the 43 m telescope at 1.2 GHz and 39,900 GPs from the GBT were recorded during these contemporaneous observations. At 1.2 GHz, 236 (3%) MP GPs and 23 (5%) IP GPs were detected at 8.9 GHz, both with zero chance probability. Another 15 (4%) low-frequency IP GPs were detected within one spin period of high-frequency IP GPs, with a chance probability of 9%. This indicates that the emission processes at high and low radio frequencies are related, despite significant pulse profile shape differences. The 43 m GPs were also correlated with Fermi γ -ray photons to see if increased pair production in the magnetosphere is the mechanism responsible for GP emission. A total of 92,022 GPs and 393 γ -ray photons were used in this correlation analysis. No significant correlations were found between GPs and γ -ray photons. This indicates that increased pair production in the magnetosphere is likely not the dominant cause of GPs. Possible methods of GP production may be increased coherence of synchrotron emission or changes in beaming direction.