We simulate the performance of a new type of instrument , a Superconducting Multi-Object Spectrograph ( SuperMOS ) , that uses Microwave Kinetic Inductance Detectors ( MKIDs ) .
MKIDs , a new detector technology , feature good QE in the UVOIR , can count individual photons with microsecond timing accuracy and , like X-ray calorimeters , determine their energy to several percent .
The performance of Giga- z , a SuperMOS designed for wide field imaging follow-up observations , is evaluated using simulated observations of the COSMOS mock catalog with an array of 100,000 R _ { 423 nm } = E/ \Delta E = 30 MKID pixels .
We compare our results against a simultaneous simulation of LSST observations .
In three years on a dedicated 4 m-class telescope , Giga- z could observe \approx 2 billion galaxies , yielding a low resolution spectral energy distribution ( SED ) spanning 350 - 1350 nm for each ; 1000 times the number measured with any currently proposed LSST spectroscopic follow-up , at a fraction of the cost and time .
Giga- z would provide redshifts for galaxies up to z \approx 6 with magnitudes m _ { i } \lesssim 25 , with accuracy \sigma _ { \Delta z / ( 1 + z ) } \approx 0.03 for the whole sample , and \sigma _ { \Delta z / ( 1 + z ) } \approx 0.007 for a select subset .
We also find catastrophic failure rates and biases that are consistently lower than for LSST .
The added constraint on Dark Energy parameters for WL + CMB by Giga- z using the FoMSWG default model is equivalent to multiplying the LSST Fisher matrix by a factor of \alpha = 1.27 ( w _ { p } ) , 1.53 ( w _ { a } ) , or 1.98 ( \Delta \gamma ) .
This is equivalent to multiplying both the LSST coverage area and the training sets by \alpha , and reducing all systematics by a factor of 1/ \sqrt { \alpha } , advantages that are robust to even more extreme models of intrinsic alignment .