Gnuplot
Produced by GNUPLOT 5.4 patchlevel 3
-5000
-4000
-3000
-2000
-1000
0
500
1000
1500
2000
2500
3000
Difference from (proleptic) TT [s]
Besselian epoch [yr]
gnuplot_plot_1
gnuplot_plot_2
gnuplot_plot_3
gnuplot_plot_4
gnuplot_plot_5
gnuplot_plot_6
gnuplot_plot_7
gnuplot_plot_8
gnuplot_plot_9
gnuplot_plot_10
gnuplot_plot_11
gnuplot_plot_12
gnuplot_plot_13
gnuplot_plot_14
gnuplot_plot_15
gnuplot_plot_16
gnuplot_plot_17
gnuplot_plot_18
gnuplot_plot_19
observed UT (time and date according to the sun)
extrapolated UT (time and date according to the sun)
1 hour
broadcast time signals
navigation/computer time -- TI?
or civil time -- UTC?
the ITU-R will choose
The Torino Colloquium in 2003
recommended a new name, TI
for broadcasts without leaps.
With a new name the "right" zoneinfo
files for computer time already demonstrate
how civil time could retain leap seconds while
broadcast and POSIX time could be uniform.
Civil time already has leaps of an hour,
and POSIX time can also handle leaps of seconds.
The technical issues have already been addressed.
The question the ITU-R will decide is whether all
civil clocks, and all civil calendars,
will drift away from the sun.
Steve Allen https://www.ucolick.org/~sla/leapsecs/
Future Time Scales -- Will Civil Time follow the Sun?