ESPaDOnS
time allocation statistics


Up to now, ESPaDOnS has been offered to the community for 4 successive semesters (2005a, 2005b, 2006a and 2006b). A quick account on requested/allocated nights is given below in each case. The relatively large number of nights already allocated on ESPaDOnS (116 nights altogether) allows a meaningful statistical study, indicating that the average oversubscription rate for ESPaDOnS is about twice higher than that for other CFHT instruments (and in particular for MEGACAM and WIRCAM). Another equivalent way of saying this is that ESPaDOnS program PIs are, in average, half as successful at getting time as PIs on other instrument programs. This trend is present since the very beginning of the ESPaDOnS operation, and there is no obvious indication that the situation is improving with time.

Semester 2006b

The total ESPaDOnS allocation in 2006b is 27 nights, among which 25 scientific nights (11 programs), and 2 discretionary nights. The 25 scientific nights split into 12 nights on the France/Opticon side (6 programs), 6 nights on the Canada side (2 programs), 2 nights on the Taiwan side (1 program) and 5 nights on the Hawaii side (2 programs). This allocation splits into 2 main observing blocks (despite this being rather damaging for a few programs that cannot address their main science goals or observe their main targets); the first one is in 2006 Aug (1 to 14 included) and the second one in 2006 Nov/Dec (Nov 28 to Dec 10 included).
The number of requested nights per instrument and agency is as follows:
ESPaDOnS WIRCAM MEGACAM Other Total
France+Opticon 57 48 31 21 157
Canada 30 23 45 5 103
Taiwan 6 15 0 1 22
Total 93 86 76 27 282

The average oversubscription factor for ESPaDOnS in 2005b on the France/Opticon/Canada side is 87/18=4.8. This number is 4.7 and 5.0 for France/Opticon and Canada agencies respectively. For a total allocation of 80 nights by France/Opticon/Canada in 2006a (40 nights on each side, CFHTLS excluded), the average oversubscription rates are equal to 260/80=3.2, 157/40=3.9 and 103/40=2.6 on the France/Opticon/Canada, France/Opticon and Canada sides respectively, ie all three significantly lower than that the corresponding values for ESPaDOnS.
The total allocation on WIRCAM, MEGACAM and other instruments on the France/Opticon/Canada side is 62 nights (28 nights on the France/Opticon side, 34 nights on the Canada side), yielding average oversubscription rates on WIRCAM, MEGACAM and other instruments (as a whole) of 173/62=2.8 (100/28=3.6 and 73/34=2.1 on the France/Opticon and Canada sides respectively). The number of nights allocated specifically on WIRCAM, MEGACAM and the other instruments is not known at this stage.
Once again, the average oversubscription rate for ESPaDOnS is significantly larger (x1.7) than that of WIRCAM, MEGACAM and other instruments.

Semester 2006a

The total ESPaDOnS allocation in 2006a is 23 nights, among which 19.3 scientific nights (9 programs), 1.7 discretionary nights and 2 engineering nights. Note that this is the smallest allocation of all semesters (partly reflecting the introduction of WIRCAM at CFHT). The 19.3 scientific nights split into 10.3 nights on the France/Opticon side (5 programs), 3 nights on the Canada side (1 program), 3 nights on the Taiwan side (2 programs) and 3 nights on the Hawaii side (1 program). This allocation splits into 2 main observing blocks, the first one in 2006 Feb (6 to 16 included) and the second one in 2006 Jun (7 to 18 included).
The number of requested nights per instrument and agency is as follows:
ESPaDOnS WIRCAM MEGACAM Other Total
France+Opticon 54 42 19 25 140
Canada 26 30 44 11 111
Taiwan 3 21 1 0 25
Total 83 93 64 36 276

The average oversubscription factor for ESPaDOnS in 2005b on the France/Opticon/Canada side is 80/13.3=6.0. This number is 5.3 and 8.7 for France/Opticon and Canada agencies respectively. For a total allocation of 72 nights by France/Opticon/Canada in 2006a (36 nights on each side, CFHTLS excluded), the average oversubscription rates are equal to 251/72=3.5, 140/36=3.9 and 111/36=3.1 on the France/Opticon/Canada, France/Opticon and Canada sides respectively, ie significantly lower than that the corresponding values for ESPaDOnS.
The total allocation on WIRCAM and MEGACAM on the France/Opticon/Canada side is 55 nights (26 nights on the France/Opticon side, 29 nights on the Canada side), yielding average oversubscription rates on both WIRCAM and MEGACAM of 2.5 (2.3 and 2.7 on the France/Opticon and Canada sides respectively). The number of nights allocated specifically on WIRCAM and MEGACAM is not known at this stage.
As in previous semesters, the average oversubscription rate for ESPaDOnS is much higher (x2.4) than that of WIRCAM and MEGACAM.

Semester 2005b

The total ESPaDOnS allocation in 2005b is 32 nights, among which 30 scientific nights (14 programs), 1.5 discretionary nights and 0.5 technical night. The 30 scientific nights split into 11.5 nights on the France/Opticon side (5 programs), 13.5 nights on the Canada side (6 programs), 2 nights on the Taiwan side (2 programs) and 3 nights on the Hawaii side (1 program). This allocation splits in 4 main observing blocks, one in 2005 Aug (18 to 25), one in 2005 Sep (14 to 24), one in 2005 Dec (13 to 20) and one in 2006 Jan (8 to 12).
The number of requested nights per instrument and agency is as follows:
ESPaDOnS WIRCAM MEGACAM Other Total
France+Opticon 93 23 34 29 179
Canada 45 12 31 12 100
Total 138 35 65 41 279

The average oversubscription factor for ESPaDOnS in 2005b on the France/Opticon/Canada side is 138/25=5.5, and equal to 8.1 and 3.3 for France/Opticon and Canada agencies respectively. Since 73 nights are allocated by France/Opticon/Canada in 2005a (36.5 nights on each side, CFHTLS excluded), the average oversubscription rates are equal to 279/73=3.8, 179/36.5=4.9 and 100/36.5=2.7 on the France/Opticon/Canada, France/Opticon and Canada sides respectively, ie significantly lower than that the corresponding values for ESPaDOnS.
Out of the 73 allocated nights, 6.5 nights were allocated to WIRCAM (all on the France/Opticon side), 24 nights to MEGACAM (7 on the France/Opticon side) and 17.5 to other instruments (11.5 on the France/Opticon side) in addition to the 25 nights allocated to ESPaDOnS. The resulting average oversubscription factors are 5.4, 2.7 and 2.3 for WIRCAM, MEGACAM and other instruments respectively (5.5 for ESPaDOnS), and 3.5, 4.9 and 2.5 on the France/Opticon side (8.1 for ESPaDOnS). The number for WIRCAM is weakly significant since WIRCAM was not fully operational in 2005b (hence the small number of requested/allocated nights).
The oversubscription rate for ESPaDOnS is thus the highest of all CFHT instruments - this difference is highest on the France/Opticon side, where the oversubscription rate for ESPaDOnS is about twice higher than for all other CFHT instruments.

Semester 2005a

The total ESPaDOnS allocation in 2005a is 34 nights, among which 30.5 scientific nights (11 programs) and 3.5 discretionary nights. The 30.5 scientific nights split into 19.5 nights on the France/Opticon side (6 programs), 10 nights on the Canada side (4 programs) and 1 night on the Korean side (1 program). This allocation splits in 5 main observing blocks, one in 2005 Feb (19 to 22), one in 2005 Mar (21 to 26), one in 2005 May (18 to 24), one in 2005 Jun (15 to 25) and one in 2005 Jul (14 to 19).
The number of requested nights on the French/Opticon side is as follows (the number of requested nights on the other sides are not known):
ESPaDOnS MEGACAM+Others Total
France+Opticon 95 27 122

The average oversubscription factor for ESPaDOnS in 2005a on the France/Opticon side is 95/19.5=4.9. Since 33 nights are allocated by France/Opticon in 2005a (CFHTLS excluded), the average oversubscription rate is 33/122=3.7, significantly lower than that for ESPaDOnS alone.
Out of the 33 allocated nights, 4.5 nights were allocated to MEGACAM and 9 to other instruments, in addition to the 20.5 nights allocated to ESPaDOnS. The resulting average oversubscription factors MEGACAM and other instruments is 27/13.5=2.0 (4.9 for ESPaDOnS).
The oversubscription rate for ESPaDOnS is thus more than twice higher than for other CFHT instruments.

Additional notes

The reader should note that only the number of allocated nights on ESPaDOnS and on instruments other than MEGACAM and WIRCAM is publically known (through the schedules, available from the CFHT web site). In particular,
the total number of nights to be allocated (CFHTLS excluded) is changing each semester and is not officially available (eg on the CFHT website);
the total number of nights allocated on MEGACAM (CFHTLS excluded) and WIRCAM is not officially available;
the total number of requested nights on all instruments are not officially available.
The non-official numbers mentioned above were obtained through various indirect channels, and may thus be slightly off, by no more than a few nights typically. The associated uncertainties are however small enough to ensure that the results and conclusions outlined above remain unaffected. Access to the true numbers has been requested to the director of CFHT (no answer yet).

Another information collected during this statistical study is that time allocated in queue scheduling (MEGACAM and WIRCAM) is counted differently than time allocated in normal scheduling. For time allocated in queue scheduling (usually counted in hr), nights are counted as lasting 6 hr in average (and the rest as overheads, including in particular a provision for bad weather). For time allocated in normal scheduling (and in particular ESPaDOnS time), nights are counted as real nights, lasting for 10-12 hr typically, depending on the season, with no provision made for bad weather. This difference further accentuates the oversubscription rate contrasts mentioned above.



© Jean-François Donati, last update Jun 27 2006.