“ The Jansky–Very Large Array Sky Survey ( VLASS ) ” ( 28 ) comprises two distinct S-band ( 2 < \nu < 4 \mathrm { ~ { } GHz } ) surveys : ( 1 ) The shallow ( rms noise \sigma _ { \mathrm { n } } \approx 69 \mu \mathrm { Jy~ { } beam } ^ { -1 } \approx 1.5 \mathrm { % ~ { } K } at \theta \approx 2 \farcs 5 resolution ) but wide ( covering all 33 , 885 \mathrm { ~ { } deg } ^ { 2 } north of \delta = -40 ^ { \circ } ) “ All-sky ” and ( 2 ) the sensitive ( \sigma _ { \mathrm { n } } \approx 1.5 \mu \mathrm { Jy~ { } beam } ^ { -1 } \approx 0.32 % \mathrm { ~ { } K } at 0 \farcs 8 resolution or \approx 0.13 \mathrm { ~ { } K } at 2 \farcs 0 \times 0 \farcs 8 resolution ) but narrow ( 10 \mathrm { ~ { } deg } ^ { 2 } in three patches ) “ Deep . ”

All-Sky is intended to be a community resource , the JVLA update of the high-impact FIRST and NVSS VLA surveys made at 1.4 GHz .
FIRST and NVSS succeeded for two reasons : ( 1 ) they are > 10 \times better in number of sources detected , sensitivity , resolution , position accuracy , etc .
than prior “ all-sky ” radio surveys , and ( 2 ) they have not been surpassed for almost two decades , so their high citation rates have not diminished .

In contrast , ( 1 ) the proposed All-Sky is only about 1.5 \times more sensitive than FIRST for point-source populations whose effective spectral index is \langle \alpha \rangle \approx - 0.7 , and its sensitivity to extended sources ( e.g. , many radio galaxies and quasars , diffuse sources in galaxy clusters , low-redshift star-forming galaxies ) is about 3 \times worse than FIRST and about 60 \times worse than NVSS because its angular resolution is so high .
The rms noise \sigma _ { \mathrm { n } } on survey images has units of apparent brightness ( \mu \mathrm { Jy~ { } beam } ^ { -1 } or K ) , not flux density ( \mu \mathrm { Jy } ) , so the high-resolution All-Sky images will miss > 25 % of all sources with flux densities S > 5 \sigma _ { \mathrm { n } } \approx 350 \mu \mathrm { Jy } , including entire low-brightness populations such as normal spiral galaxies , whose median S-band surface brightness is only \langle T _ { \mathrm { b } } \rangle \approx 0.17 \mathrm { ~ { } K } \ll 5 \times 1.5 \mathrm { ~% { } K } .
( 2 ) All-Sky will probably be surpassed before its earliest completion date ( 2023 ) by the contemporary ASKAP EMU survey ( 26 ) covering the whole sky south of \delta = +30 ^ { \circ } , and by the complementary Westerbork WODAN survey north of \delta = +30 ^ { \circ } .
EMU intends to reach \sigma _ { \mathrm { n } } = 10 \mu \mathrm { Jy~ { } beam } ^ { -1 } \approx 0.06 \mathrm { ~ { } K } at \theta = 10 ^ { \prime \prime } resolution at L band ( 1.1 < \nu < 1.4 GHz ) , which is equivalent to \sigma _ { \mathrm { n } } \approx 6 \mu \mathrm { Jy~ { } beam } ^ { -1 } \approx 0.01 \mathrm { % ~ { } K } at S band .
EMU will detect spiral galaxies and generate complete samples of all sources ~ { } 10 \times below the All-Sky detection limit for point sources .
Nearly all multi-wavelength astronomers asking the question “ Are my favorite objects radio sources ? ” will get better answers from EMU than from All-Sky .

The uniquely high angular resolution of All-Sky , and hence its poor surface-brightness sensitivity , was forced by a straw-man optical identification method that fails to distinguish between extended radio sources ( e.g. , core plus lobes and jets ) and radio source components ( brightness peaks on images ) , so it tries to identify individual clearly resolved radio lobes with unrelated galaxies .
The high resolution of All-Sky is not necessary for identification reliability ( except for radio stars ) , and it is counterproductive because it lowers identification completeness—All-Sky will miss > 25 % of all sources with flux densities above its 5 \sigma _ { \mathrm { n } } detection limit .
No radio detection , no optical identification .
Incompleteness is actually worse than unreliability because incompleteness can not be corrected , while sources with marginal identifications can be reobserved .

The high sensitivity and angular resolution of Deep is likely to remain unique until SKA Phase I is operating .
Deep is qualitatively like the large PI VLA survey COSMOS , but bigger .
Its top science goals are ( 1 ) a pilot survey for constraining dark energy by detecting the effect of weak gravitational lensing on distant radio sources and ( 2 ) studying the evolution of complete ( flux-limited ) galaxy samples .
These goals conflict because resolving most faint sources for goal ( 1 ) implies sample incompleteness for goal ( 2 ) .
The proposal suggests that Deep will resolve most sources so it can detect weak lensing , but Deep won ’ t resolve most sources so its completeness limit in \mu \mathrm { Jy~ { } beam } ^ { -1 } can be conflated with source flux densities in \mu \mathrm { Jy } .
The Deep proposal must fully address the angular-size distributions of \mu Jy sources before it can be reviewed responsibly .

The VLASS will cost about 9,000 hours of JVLA observing time plus 20+ FTE-years of Socorro scientific staff effort for development .
User feedback about the VLA time taken by FIRST and NVSS led to the Bridle et al .
( 7 ) Report , which was ignored by the proposal but should be required reading .
The front-loaded VLASS support requirements could divert most of the already-overloaded Socorro scientific staff from JVLA commissioning and helping observers for the next few years .