We explore a technique for identifying the highest redshift ( z > 4 ) sources in Herschel /SPIRE and BLAST submillimeter surveys by localizing the position of the far-infrared dust peak .
Just as Spitzer /IRAC was used to identify stellar ‘ bump ’ sources , the far-IR peak is also a redshift indicator ; although , the latter also depends on the average dust temperature .
We demonstrate the wide range of allowable redshifts for a reasonable range of dust temperatures and show that it is impossible to constraint the redshift of individual objects using solely the position of the far-IR peak .
By fitting spectral energy distribution models to simulated Herschel /SPIRE photometry we show the utility of radio and/or far-infrared data in breaking this degeneracy .
With prior knowledge of the dust temperature distribution it is possible to obtain statistical samples of high redshift submillimeter galaxy candidates .
We apply this technique to the BLAST survey of ECDFS to constrain the number of dusty galaxies at z > 4 .
We find 8 \pm 2 galaxies with flux density ratios of S _ { 500 } > S _ { 350 } ; this sets an upper limit of 17 \pm 4 deg ^ { -2 } if we assume all are at z > 4 .
This is < 35 % of all 500 \mu m-selected galaxies down to S _ { 500 } > 45 mJy ( L _ { IR } > 2 \times 10 ^ { 13 } L _ { \odot } for z > 4 ) .
Modeling with conventional temperature and redshift distributions estimates the percentage of these 500 \mu m peak galaxies at z > 4 to be between 10–85 % .
Our results are consistent with other estimates of the number density of very high redshift submillimeter galaxies and follows the decline in the star formation rate density at z > 4 .