[Lofarpwg] LBA census paper, submitting tomorrow

[Anna Bilous]

Hi Charlotte,

> {Is it worth also adding a sentence that most pulsars' spectral indices in this work favour a power law with a single break at median frequency 155 MHz? If not in the abstract, perhaps in the results or discussion?}

I know that it is tempting to start looking for pattering in the
results, but in my (strong) opinion it is way too early for that. I
don't really trust a single measurement in any frequency range
(especially LBA measurements, knowing about the problems we had)
unless it was based on multiple observing sessions, with proper
scintillation estimates, etc. Judging from the spread of points in
well-measured spectra, the difference between fluxes reported can
easily reach order of magnitude! For pulsars with less well-known
spectra shifting individual points up and down by a factor of few may
change fitted spectral break substantially (also see below). So... we
need more low-frequency fluxes!

> Results/discussion:
> - Just wondering if/how the uncertainty in the flux measurements, including due to scintillation, affect the (presence of a) spectral break frequency and uncertainty? In the plots in Fig C.1 it seems like the effect of scintillation can be seen in several of the pulsars with many flux density measurements from different epochs, but the overall spectral fits are not greatly affected  e.g. 0809, 0823, 1133, etc.

Flux density uncertainties definitely affect the existence and
location of fitted spectral breaks. Even more, since we fit for a
single value of "extra flux scatter not covered by erorbars" per
pulsar, adding more chaos at low-frequency end affects the whole
fitted shape, not only the low-frequency part. For the pulsars you
have mentioned, adding low-frequency data points indeed does not
change the overall spectral shape as seen by human eye, but the fit
does change for the last two of them (compare with HBA census paper).
In particular, number of frequency breaks and some of their locations
are shifted by more than the error reported. I would say we need
better shape characteristic than a collection of broken PLs... and
also better fluxes.

> - In addition, is it worth mentioning some of the emission mode-changers in the set, e.g. 0823 and 0943 (and relevant references!)? It seems that they were observed in the bright emission mode here, but that may not have been the case in the past. E.g. 0823 has literature upper limits between HBA and LBA frequencies, which is quite surprising based on how bright this pulsar is. Do you know the reference for these flux density upper limits and whether observing during the quiet-mode is responsible for this?

0943 is in Q-mode in LBA census and in late B in HBA :) Mode switching
definitely adds to the spread of points in the spectra, we have
already discussed it in HBA census paper.
The upper limits for 0823 fluxes around 100 MHz all come from Murphy
et al (2017) based on GLEAM data. If I understand them correctly,
their observing time was 2 minutes and they observed at frequencies
between 70 and 200 MHz simultaneously. Interestingly enough, they
detected 0823 at 76 and 115 MHz, but not in between. I guess
non-detection more probably stems from their insufficient sensitivity
there rather than pulsar being in the quiet mode.

Cheers,
Anya

> --
> On 2/9/19, 8:22 pm, "lofarpwg on behalf of Anna Bilous" <lofarpwg-bounces at astron.nl on behalf of hanna.bilous at gmail.com> wrote:
>
>     Dear all,
>
>     thanks to everyone who provided feedback on the draft. Please check
>     your affiliations in the pdf attached and also let me know if you have
>     any acknowledgements. I plan to submit the draft tomorrow (24 hr from
>     now).
>
>     Thanks again,
>     Anya
>
>