JIVE Wiki

This page is read only. You can view the source, but not change it. Ask your administrator if you think this is wrong.
====== Science Operations and Support ======

Biennial Report 2011-12
[as usual, editorial comments, temporary fig/tbl numbers, etc. in square
brackets -- captions for 2 non-repeated figures from the 2009-10 version are enclosed within ####### ]


2. Science Operations and Support

2.1  Production Correlation

2.1.1 Sessions and Their Experiments

The most significant feature of the correlation environment over this
biennial period has been the shift away from the MkIV correlator to 
the EVN software correlator at JIVE (SFXC) as the primary workhorse for
EVN correlation.  Figure [##1] shows the evolution of the fraction of
(disk-based) experiments per session correlated on the MkIV and on SFXC.
We had already begun using SFXC in 2010 for observations requiring pulsar
gating.  The transition continuing into 2011-12 was gradual, as we continued 
to use the MkIV for multi-epoch observations that begun on it.  By session
2/2012, all disk-based observations were correlating on SFXC.  e-EVN 
observations took longer to shift away from the MkIV, because of its
inherent real-time nature and concerns about the ability of SFXC to keep up
with 9-10 stations at 1 Gbps.  We had overcome those concerns by the 
December 2012 regularly scheduled e-EVN day, and all e-EVN observations since
then have also correlated on SFXC.  The section "Astronomical Features"
(2.1.3) will discuss the new kinds of experiments SFXC permits us to handle.

{{:2011-2012:sfxcuse.png?100|}}[figure 1   sfxcuse.png   goes here]
Figure [##1] Fraction of disk-based experiments correlated on the MkIV and
on SFXC, per session.


Session 1/2011 had a total of 16 user experiments correlated at JIVE,
including 3 e-EVN experiments conducted during the session.  Four user
experiments were correlated on SFXC (pulsar gating, cross-pol spectral line).
Kunming had it's first Gbps fringes in the X-band NME.  VERA_Ishigaki-jima 
participated for the first time in an EVN experiment (5cm methanol maser 
observation).

Session 2/2011 had a total of 21 user experiments correlated at JIVE.
Twelve user experiments were correlated on SFXC (cross-pol spectral line, 
pulsar gating, wide-field mapping, global spectral line) -- the first session 
in which SFXC correlated the majority of the experiments.  SFXC also
provided the first-ever correlation of 32MHz subbands at JIVE, in 2 Gbps
tests using the CDAS digital back-end on the Chinese stations.

Session 3/2011 had a total of 25 user experiments correlated at JIVE,
including 5 e-EVN experiments conducted during the session.  Thirteen
user experiments were correlated on SFXC, including the first time
that the capabilities for 8192 frequency points, multiple phase-centers,
and more than 16-station single-pass correlation were used (among different
experiments).

In 2011, there were 38 e-EVN user experiments, eight of which were
conducted during the regular disk sessions.   There were seven target-of-
opportunity experiments, and two triggered observations.


Session 1/2012 had a total of 14 user experiments correlated at JIVE,
13 done on SFXC.  Kunming participated in an X-band user experiment.
The KVAZAR stations began using their R1002 digital back-ends in
all observations.  This session saw what will most likely be the last
disk-based user experiment to correlate on the EVN MkIV data processor
(EM071D).

Session 2/2012 had a total of 18 user experiments correlated at JIVE,
including five e-EVN experiments conducted during the session.
This was the first session in which all disk-based observations
correlated on SFXC.  A record for largest network size (20 stations) was
attained in GF018B.

Session 3/2012 had a total of 31 user experiments correlated at JIVE.
Irbene participated for the first time in regular NMEs.  All three KVN
stations participated in the K-band NME, and Torun participated for the
first time in K-band observations.  The record for largest network size
fell to GM070 (23 stations).  During this session, we re-made a large
number of observing schedules after they had been uploaded by PIs, to
accommodate last-minute casualties requiring Jodrell2 to replace Jodrell1
and preventing Gbps recording at Medicina (schedules shifted to 1-bit
sampling).

In 2012, there were 29 e-EVN user experiments, five of which were
conducted during the regular disk sessions.   There were four target-of-
opportunity experiments, and a triggered observation.


Figure [##2] shows the evolution of annual EVN network hours since 2004, with
the contribution of e-EVN represented by the shaded area (not all of the
disk-based experiments were correlated at JIVE).  Figure [##3] focuses on the
e-EVN experiments, showing a division of annual observing hours into different
categories: ToOs, triggered observations, short (<= 2hr) exploratory
observations not requiring a formal proposal, experiments proposed for disk
recording, but conducted in e-VLBI (after consultation with the PI), and the
standard e-EVN observations in regularly scheduled e-VLBI (or disk) sessions.
By their nature, all e-EVN observations correlate at JIVE, and occupy a single
correlator pass.

{{:2011-2012:ntwkhr2012.png?100|}}[figure 2   ntwkhr2012.png   goes here]
Figure [##2]: Annual EVN network hours, with the contribution by e-EVN
observations shown by the shaded area.

{{:2011-2012:etypes2012.png?100|}}[figure 3   etypes2012.png   goes here]
Figure [##3]: Division of annual e-EVN network hours into categories.



Tables [##1 and ##2] summarize projects observed, correlated, distributed,
and released in 2011 and 2012.  They list the number of experiments as well
as the network hours and correlator hours for both user and test/NME
experiments.  Here, correlator hours are the network hours multiplied by any
multiple correlation passes required (e.g., because of continuum/line,
separate correlation by subband/pol to maximize spectral resolution, etc.).
Note that the instances of multiple correlator passes is largely reduced from
SFXC, since it does not have any explicit maximum spectral capacity as did
the MkIV, and that the maximum number of stations is currently limited by
the number of input Mark5 playback units, which was sufficient for all
observations in this biennial period.  There still remain some experiments
that have separate continuum and line passes, to keep the output FITS file
size more manageable.  Thus the "Ntwk_hr" and "Corr_hr" values have grown
closer together.  The "Corr_hr" statistic for SFXC does not reflect the
fact that, unlike the MkIV, SFXC is not constrained to correlate at a
"UTC rate"; this is reflected in the efficiency plot below.

                       User Experiments          Test & Network Monitoring
                     N   Ntwk_hr  Corr_hr          N   Ntwk_hr   Corr_hr
  Observed           89     701     979            28      91        91
  Correlated         71     535     728            28      92        92
  Distributed        76     585     797            24      80        80
  Released           77     619     789            27      88        88
  e-EVN experiments  38     249     249
  e-EVN ToOs          7      53      53

Table [##1]: Summary of projects observed, correlated, distributed, and
released in 2011.

                       User Experiments          Test & Network Monitoring
                     N   Ntwk_hr  Corr_hr          N   Ntwk_hr   Corr_hr
  Observed           87     711     801            31      96        96
  Correlated         85     728     916            26      82        82
  Distributed        76     673     861            31      94        94
  Released           72     634     828            30      91        91
  e-EVN experiments  29     233     233
  e-EVN ToOs          4      25      25

Table [##2]: Summary of projects observed, correlated, distributed, and
released in 2012.



Figure [##4] presents various measures of correlator efficiency.  The red
line plots the completed correlator hours per during time actively devoted to
production correlation.  The green line shows completed correlator hours over
the total operating time of the correlator -- the red and green lines diverge
more in periods when production takes up a smaller fraction of the total time
available.  The blue line shows completed network hours over total operating
time -- the green and blue lines diverge because some experiments require
multiple passes.  A twelve-week running average is shown to smooth out
spurious peaks caused by periods with no remaining production correlation.

{{:2011-2012:ar1112_f4.png?100|}}[figure 4  ar1112_f4.png   goes here]
Figure [##4]: Various measures of correlator efficiency.


Figure [##5] presents the size of the correlator queue at different stages
in the processing cycle, showing a snapshot of the status at the end of
each week.  The red line plots the number of correlator hours that remain
to be correlated.  The blue line plots the number of correlator hours whose
data remain to be distributed to the PI.  The green line plots the number
of correlator hours associated with recording media that have yet to be
released back to the pool (in practice, release occurs prior to the
following session, leading to a blocky pattern for the green line).

{{:2011-2012:ar1112_f5.png?100|}}[figure 5  ar1112_f5.png   goes here]
Figure [##5]: Size of various correlator queues, measured in correlator hours.



2.1.2  Logistics and Infrastructure

The disk-shipping requirements are derived from the recording capacity
needed by a session (from the EVN scheduler) and the supply on-hand at the
stations (from the TOG chairman).  The EVN and VLBA stations follow different
sets of guidelines:

   a) the EVN policy that stations should buy two sessions' worth of disks,
      hence the disk flux should balance over the same 2-session interval.

   b) the VLBA's need for sub-session turn-around, which essentially requires
      pre-positioning the difference between what NRAO stations will
      observe in globals to be correlated at JIVE and what EVN stations
      will observe in globals to be correlated in Socorro.

Following distribution to the stations for session 3/2012, the cumulative
flux-balance summing over both EVN and NRAO stations was with 10TB of zero.

There were unusual problems receiving packs after session 3/2011 from
Japan and VLBA stations.  Packs from these stations actually were returned
to Japan/the U.S. by the shippers.  The Japanese data were eventually 
e-shipped (they divided the data into 10-second segments and made those 
available to ftp) and we reconstituted packs at JIVE.  For the VLBAs, 
some of the packs had been recycled before we noticed the problem (3 stations 
entirely lost, 3 partially lost, 4 unaffected).   Other e-shipping included
data from parallel-DBBC testing at Hartebeesthoek in NMEs.

In this period, experiments have continued to go to three different
correlators (JIVE, Socorro, Bonn).  planning.  A principal goal in planning
the pre-session disk-pack distribution is to avoid individual packs
containing data for more than one correlator.  Thus in the disk-distribution
plan the load for each target correlator for each station is computed
separately.  Packs on-hand at a station are applied to one of these
individual-correlator loads prior to calculating what replenishment is
required from JIVE.  These loads are computed by
    *) assuming a 100% recording duty-cycle for the duration of each experiment
    *) subtracting 0.09TB from the capacity of each pack (to provide a buffer
         to compensate for unused space at the ends of packs du to the (last + 1) Mark5 scan not having fit onto the pack). 
in order to provide some insurance against loss of capacity due to bad packs,
(but would likely be insufficient if a very large pack fails).  The
disk-distribution plan tabulates the specific set of packs to use per 
target correlator in terms of how many of which capacity.  We try to use
preferentially the largest-available packs for the farthest stations (to cut
down on shipping), but consistent with minimizing unused capacity (e.g., 
4.4TB = 3.2TB + 1.48TB instead of a 6TB or 8TB).

The current play-back line-up is 14 Mark5As, 2 Mark5Bs, 1 Mark5B+, and 
7 Mark5Cs.   The MarkIV correlator is still limited to a maximum of 
16 stations.  Correlation on SFXC bypasses the station units, so is entirely
divorced from any 16-station limitation.  Further, the flavor of Mark5 unit
is immaterial to SFXC correlation, thus setting an effective maximum
array-size for single-pass correlation of 24.  Among the standard EVN
stations, Effelsberg, Westerbork, Yebes, Urumqi, Shanghai, Hartebeesthoek,
Badary, Zelenchukskaya, and Svetloe currently provide Mark5B recordings (as do
typically Irbene, Kunming, the Japanese stations, and KVN stations among the
non-EVN stations correlated in this biennial period).  The need to retain a
number of Mark5A units depends only on the contingency that the MkIV
correlator would be needed (e.g., e-VLBI with more stations than SFXC could
handle at a given time).  Since SFXC can now keep up with 12 stations at 
1 Gbps, this eventuality is not presently pressing.  Playback through the
Mark5B, bypassing the station units, does result in lower statistical noise
in the correlated phases.

We continue to encounter the occasional individual bad disk (or two)
in an incoming pack.  We maintain a small bench stock of disks of
various sizes so that we could replace a bad disk locally if that
is the most appropriate course of action (in light of warranty status,
urgency of recycling, etc.), and then we would get a new disk from the
pack's "owner" to replenish our bench stock.


2.1.3  Astronomical Features

e-VLBI capabilities have remained a cornerstone aspect of the EVN.
The total e-EVN hours are down somewhat from their high in 2010, arising
mostly from fewer Target-of-opportunity (ToO) observations using e-VLBI
(11, but there were also 11 disk-based ToO observations -- ones requiring 
multiple correlator passes or stations not having e-VLBI connections (e.g.,
KVAZAR stations, Urumqi)).  Still, over the biennial period, 34% of the
observed EVN network hours correlated at JIVE were e-EVN observations.
There were a growing number of e-EVN observations conducted during regular
EVN (disk) sessions.  This can provide the opportunity to get longer e-EVN
observations than would be possible in the regularly-scheduled e-EVN days.
A record for the longest e-EVN observation at 48hr was attained in session
1/2011.  In terms of e-EVN network improvements, Noto joined for the 
first time in June 2012 at 512 Mbps, and by September 2012 could sustain
896 Mbps (i.e., channel-dropping one of eight 16MHz subbands in a Gbps mode).
Hartebeesthoek, Medicina, and Yebes all improved to being able to sustain
a full Gbps (no channel dropping required any longer).

SFXC has now correlated many user experiments that would have been
impossible or at best much less efficient on the MkIV:
   *) 4 spectral-line experiments having more than 2048 frequency points
     per subband/polarization  (record so far = 8192)
   *) 17 spectral-line experiments with cross-polarizations
   *) 7 pulsar gating experiments  (record minimum period so far = 16.45ms)
   *) 15 experiments with multiple phase centers  (spanned fields range from
     25" to 10';  record number of multiple phase centers so far = 50)
   *) 4 experiments having more than 16 stations  (record so far = 23)
There is some overlap among the above list (e.g., an experiment used both 
pulsar gating and multiple phase centers).  There have been 22 other user 
experiments that SFXC was able to correlate in a single pass, but would have 
required multiple MkIV passes, even though they exceeded no individual 
MkIV limitation in terms of only number of stations or frequency points.

SFXC avoids the physical limit present in the MkIV, by which a single
interferometer (baseline/subband/polarization) could not exceed the
capacity of a single correlator board.  In local validity, this meant
no more than 2048 frequency points per interferometer.  The selection of
observing/correlation parameters is greatly simplified for the PI:  instead 
of having to navigate through a series of non-intuitive formulas, one now
can set the subband bandwidth and number of frequency points directly from
a desired velocity spacing and continuum sensitivity.  Besides the possibility
of increased spectral resolution, SFXC also offers spectral-line observations
the advantages of station-based fringe-rotation (no need for CVEL corrections)
and the ability to select the spectral-windowing function (default = Hanning,
but uniform, Hamming, and cosine are currently available -- the MkIV provided
only uniform).  A more esoteric improvement pertains to cross-pol spectral-line
observations (which are growing in popularity, as it has been demonstrated that
methanol and OH masers provide the ability to map out the magnetic fields
around massive protostars).  The MkIV applied (baseline-based) fringe-rotation
entirely to one station (always the first station as fed to the correlator 
from the station units), but for the fourth polarization, it would swap 
the first/second order of the stations in the baseline (e.g., in Ef-Wb,
polarization LR would be Ef(L)-Wb(R) with fringe rotation done to Ef; but 
RL would be Wb(L)-Ef(R) with fringe rotation done to Wb).  This asymmetry 
between the fringe-rotation zero-point for the two cross-hands polarizations 
was never repairable in AIPS.  It is avoided altogether in SFXC.

The combination of an essentially arbitrarily large number of frequency
points and an arbitrarily small integration time in SFXC makes it a much
more powerful wide-field mapping correlator (the integration time in the
MkIV was limited by the time required to read out the correlator output
board -- 0.25s for the whole correlator or 0.125s for half the correlator),
one that could map an area on the sky on the order of the single-dish
beams without appreciable bandwidth- or time-smearing.  The price of course
is huge data sets (one can see in the growth of the archived FITS files
in fig [##9] that there are a higher number of very large experiments once
the transition to SFXC has been completed).  Multiple phase-center correlation
performs an "internal" correlation with a very large number of frequency
points and a very small integration time (current records are 32k frequency
points and 4.864ms), but then outputs only subsets of this initial wide
field using more traditional values for frequency points and integration times.
The most popular applications for multiple phase-center correlation have
been following an in-beam phase-referencing calibrator (this sometimes 
requires different schedules for the small and large telescopes, the latter 
ones still having to slew between the two close sources) and investigating
a population of sources (e.g., from FIRST, NVSS, or GB6) that happen to
lie in the field of the principal VLBI target.

SFXC provides pulsar-gated correlation, which never attained operational
availability on the MkIV.  A number of independent bins can be placed within 
a single gate, defined by a start/stop phase with respect to the pulsar period.
Each bin would produce a separate FITS file.  Traditional gating in the MarkIII
sense corresponds to 1 bin.  Figures 2.7 and 2.8 in the 2009-10 JIVE Biennial 
Report show the correspondence between pulse profiles constructed from 
gated SFXC correlation using multiple bins and observed single-dish pulse 
profiles.

##### OLD FIGURES CAPTIONS FROM 2009-10#######
#[ Figure [##7] shows an example of a correlation with
#  100 bins spread over a gate of a tenth of the period for PSR 0329+54 at 
#  1.4 GHz.  Figure [##8] shows the Effelsberg single-dish pulse profile,
#  illustrating that the SFXC pulse profile built up from the independent 
#  bins within the gate reproduces the pulse profile well. 
#
#    [figure 7   profile-onpulse.png   goes here]
#Figure [##7]  Pulse profile for PSR 0329+54 built up from 100 separate bins
#within a gate of one-tenth of the pulsar's period, from a test observation
#at 1.4 GHz.
#
#    [figure 8   sieber2.png   goes here; side-by-side with fig##7, or
#        preferably on top of each other, with a horizontal scaling to
#        make the pulses appear the same width (i.e., two side-pulses
#        line up with each other in the two plots)]
#Figure [##8]  Effelsberg single-dish pulse profile from Sieber et al. (1975,
#A&A, 38, 169). ]
###############################3


With the transition towards e-MERLIN and the removal of the microwave
links connecting the out-stations to Jodrell Bank, we have temporarily
lost the ability to include out-stations in the EVN correlation. 
Jodrell Bank and JIVE personnel are working to develop the ability to 
include the fiber-connected out-stations (after having passed through 
the e-MERLIN correlator) anew in an EVN correlation.


2.2  EVN Support

Automatic-ftp fringe tests are included in all network monitoring experiments
(NMEs) at the beginning of each new frequency sub-session within EVN
sessions, or as a separate fringe-test observation when the NME does not
appear first in the schedule or falls well outside working hours.  Under the
control of sched and the field system, a specified portion of a scan is sent
directly to the SFXC cluster at JIVE.  Multiple ftp transfers per experiment
provide the opportunity to iterate with the stations in investigating any
problems identified.  Use of ftp transfer and near-real-time correlation
permits stations that don't have a full e-VLBI connection to participate.
A skype chat session during the ftp fringe-test observations provides even
more immediate feedback between the station friends and the JIVE support
scientists.  Correlation results go to a web page available to all the
stations, showing baseline amplitude and phase across the band as well as
autocorrelations, and each plot is accessible by moving the cursor over
color-coded baseline/subband/polarization cells.  The web-based results
from the first and probably the second ftp transfer would be available to the
stations before the end of the NME.  These ftp fringe tests continue to be
very successful in identifying telescope problems and helping to safeguard
user experiments by allowing the station friends to take care of any such
problems before the actual astronomical observations begin.

The EVN pipeline runs under ParselTongue (a Python interface to classic
AIPS), providing greater scope for future development due to the improved
coding environment.  The pipeline scripts are available from the ParselTongue
wiki (RadioNet) and should provide a good basis for other (semi-)automated
VLBI reduction efforts.  We continue to process all experiments, including
NMEs via the pipeline, with results being posted to the EVN Archive.  The
pipeline also provides stations with feedback on their general performance
and in particular on their gain corrections, and identifies
stations/frequency bands with particular problems.

The transition from the analog mark4/vlba4 formatters to digital back-ends
has gathered pace in this biennial period.  Effelsberg recorded session 1/2011
in parallel onto the DBBC, and has used the DBBC for all observations starting
in session 2/2011.  Further parallel-DBBC testing has taken place in Onsala,
Hartebeesthoek, Noto, Yebes, and Metsahovi.  This has been in the "Digital
Down-Converter" personality, which can mimic the BBC-tuning and subband-
bandwidths available on the existing back-ends.  Figure [##6] shows a
comparison of Ef-On and Ef-Od baselines (comparing the mark4 formatter and
DBBC back-end at Onsala, while Ef is using the DBBC).  The passband is
flatter on the DBBC-DBBC baseline, and the phase across the entire range
of BBCs is much flatter, with no phase shifts between BBCs apriori (no
phase-cal alignments applied in this plot).  Extracting calibration 
information remains one of the last stumbling blocks for more stations
permanently moving over to the DBBC.  2 Gbps fringes on the Chinese digital
back-end CDAS were achieved in October 2011.  The KVAZAR stations shifted
to their R1002 digital back-ends in session 1/2012.  Previously, they each
had a unique configuration, so this transition improves consistency --
espeically avoiding Gbps C-band quirks such as the Svetloe cut-off at 5000 MHz
and internal down-converter interference costing one of eight subbands at 
Zelenchukskaya.

{{:2011-2012:ondbbc.png?100|}}[figure 6  ondbbc.png   goes here]
Figure [##6]:  Amplitude and phase vs. frequency on the baselines Ef-On 
(Onsala with a mark4 formatter) and Ef-Od (Onsala with a DBBC) for the session 
1/2012 L-band fringe-test experiment F12L1.


NEW-STA
There have been quite a few new stations participating in astronomical
observations.   VERA_Ishigaki_jima participated for the first time in
some methanol-maser astrometric observations, starting in session 1/2011
and continuing throughout all of 2011.  Like the other VERA stations,
this is not under field-system control, and provides Mark5B-format disk-packs
they generate by translating from their native VERA recording tapes.
Without a field-system log to control the antennas or associate bytes on
the pack with scan start times, they record continuously, moving the antenna
under local control to match the schedule, and we dead reckon the byte/scan
associations from knowing the begin/end times of the recordings (per
individual original VERA tape).  Kunming obtained Gbps fringes in the X-
and S/X-band NMEs in session 2/2011, enabled by the Mark4 back-end that
was originally at Wb.  The first K-band fringes from the KVAZAR stations
Svetloe and Zelenchukskaya came in a ToO in September 2011.  Two Korean
VLBI Network stations (Yonsei, Tamna) participated in their first test
with EVN stations in October 2011, an e-VLBI observation at 512 Mbps.
All three KVN stations (also Ulsan) got fringes in the K-band NME in
session 3/2012.  Figure [##7] shows the fringes on the baselines formed
among the three KVN stations and Shanghai (fringes were also visible on
Korean-European baselines, but were weaker due to the length of these
baselines).  The KVN stations use their own digital back-ends, and here
they also exhibit linear phases across the entire band, with no offsets
apriori between adjacent subbands (Shanghai here was using their VLBA4
back-end).  Irbene obtained its first fringes during a test
observation in April 2012 (C-band, 512 Mbps), and got fringes in both
the C- and L-band NMEs in session 3/2012.  Figure [##8] shows the fringes
on the baseline Effelsberg-Irbene, with both stations having a DBBC
back-end.

##### FIGS
{{:2011-2012:kvn.png?100|}}[figure 7   kvn.png   goes here]
Figure [##7]:  Amplitude and phase vs. frequency on the baselines among
the three KVN stations and Shanghai for the session 3/2012 K-band NME N12K4.

{{:2011-2012:irbene.png?100|}}[figure 8  irbene.png   goes here]
Figure [##8]:  Amplitude and phase vs. frequency on the baseline Effelsberg -
Irbene for the session 3/2012 C-band NME N12C4.  Irbene had only RCP available,
a known feature of these observations.


2.3  PI Support

The EVN Archive at JIVE provides web access to the station feedback,
standard plots, pipeline results, and FITS files for experiments correlated
at JIVE.  Public access to the FITS files themselves and derived
source-specific pipeline results is governed by the EVN Archive Policy --
the complete raw FITS files and pipeline results for sources identified by
the PI as "private" have a one-year proprietary period, starting from
distribution of the last experiment resulting from a proposal.  PIs can
access proprietary data via a password they arrange with JIVE.  PIs receive
a one-month warning prior to the expiration of their proprietary period.

We moved the Archive onto a bigger, more powerful machine (the EVN 
pipeline also runs on this machine).  It has 33 TB of available disk space, 
which it shares with the pipeline work area (currently using around 2.3 TB). 
The total size of the FITS files in the archive at the end of 2012 was 14.84 TB
(a 5.87 TB gain in the two-year period); figure [##9] shows the growth 
of the FITS-file size in the EVN archive size over time.  A pick-up in
the number of very large experiments can be seen following completion
of the transfer to SFXC.

[figure 9  archvgro2012.png   goes here]
Figure [##9]:  Growth in the size of FITS files in the EVN archive.
Experiments archived in this biennial period are plotted in red.
Vertical blue lines demark the date of archiving the FITS files from
the first SFXC correlation and that from the last MkIV correlation of 
a disk-based observation, bounding the transition period in populating
the archive from the two correlators.

The science operations and support group continues to contact all PIs once
the block schedule is made public to ensure they know how to obtain help
with their scheduling.  There were 12 first-time EVN PIs in 2011 and another
9 in 2012.  We also checked schedules PIs posted to VLBEER prior to
stations downloading them, with safeguards in place to minimize the chance
that different stations use different versions of an experiment's schedule.
There were a couple instances in which last-minute casualties at the
stations caused us to re-make schedules centrally after the PIs had already
uploaded them, most notably in session 3/2012.  Jodrell1 suffered an
azimuth wheel casualty, so we shifted all of its schedules (24) to Jodrell2,
including inserting it into scans that Jodrell1 intentionally missed in
fast cycle-time phase-reference observations.  Medicina saw in the L-band
sub-session that they were not able to record Gbps observations, so we
remade schedules for their remaining C-band and X-band Gbps observations 
using 1-bit sampling (hence, bit-rate falling to 512 Mbps).

The preferred patchings for stations using digital back ends can fail
the various checking rules in sched, so we have continued to provide
PIs of experiments with plug-ins for their sched-input files that 
properly specify the patchings while allowing sched to run without
complaint.  We have provided new code to sched to handle the KVAZAR
R1002 digital back ends, and are close to being to do similarly for
the DBBC/DDC personality (here, there stations separate into two camps
in terms of their preferred patchings).

We continued to provide maintenance for the EVN-specific portion of the
NorthStar Proposal Tool.  Further, we modified the the organization of the
available "facilities" within the EVN portion:  merging the separate 
EVN+MERLIN and e-EVN facilities into one.  Now e-EVN observing can be 
requested on an "observation" basis within a single EVN+MERLIN proposal, 
allowing proposals that contain some parts using e-VLBI and some parts 
using disk-based observations, presenting the correct choice of frequencies
and telescopes for each such observation.