Not listed PRECISION PHOTOMETRIC REDSHIFTS FOR COSMOLOGYTHE GROWTH OF STRUCTURE AS MEASURED BY WEAK LENSING HAS BEEN IDENTIFIED AS ONE OF THE MOST SENSITIVE PROBES OF DARK ENERGY AND DARK MATTER AND IS ONE OF THE THREE KEY DARK ENERGY EXPERIMENTS PROPOSED FOR WFIRST. HOWEVER THE WEAK LENSING MEASUREMENT DEPENDS STRONGLY ON ROBUST PHOTOMETRIC REDSHIFTS AND IS HIGHLY SENSITIVE TO SYSTEMATIC BIASES IN THESE REDSHIFT ESTIMATES. SEVERAL METHODS HAVE BEEN PROPOSED TO REMOVE SYSTEMATIC BIASES BASED ON SPECTROSCOPIC SAMPLES AND SPATIAL CLUSTERING BUT NONE HAS BEEN DEMONSTRATED TO PERFORM AT THE LEVEL REQUIRED FOR WFIRST. MAKING THE PROBLEM MORE CHALLENGING AT LEAST TWO INDEPENDENT METHODS MUST BE DEVELOPED: ONE TO CORRECT THE SYSTEMATIC ERRORS AND ANOTHER TO VERIFY THE CORRECTION AND QUANTIFY RESIDUAL ERROR. HERE WE PROPOSE TO DEVELOP AN INFORMED CALIBRATION OF THE COLOR-REDSHIFT RELATION THAT WILL MINIMIZE THE NUMBER OF SPECTROSCOPIC REDSHIFTS NEEDED FOR MACHINE LEARNING ALGORITHMS AND PRODUCE ACCURATE BAYESIAN PRIORS FOR TEMPLATE FITTING ALGORITHMS.THE PROPOSED METHOD USES OUR CURRENT KNOWLEDGE OF GALAXIES AND GALAXY EVOLUTION FROM EXISTING DEEP SURVEYS TO PARAMETERIZE WHERE IN THE WFIRST COLOR SPACE THE PHOTOMETRIC REDSHIFTS ARE WELL UNDERSTOOD AND WHERE THEY ARE NOT. FIRST WE WILL DEVELOP A METHOD TO MAP FROM THE WFIRST N-DIMENSIONAL COLOR SPACE TO REDSHIFT. THIS WILL DETERMINE WHICH REGIONS OF COLOR SPACE MAP TO REDSHIFT IN A WELL-BEHAVED WAY AND WHICH HAVE A MORE COMPLEX BEHAVIOR. WE WILL MAKE USE OF THE FACT THAT HIGHER-DIMENSIONAL DATA(NARROWER BAND PASSES MORE SENSITIVE DATA AND LARGER SPECTRAL COVERAGE) ARE AVAILABLE IN SELECT AREAS OF THE SKY TO DETERMINE HOW MUCH UNCERTAINTY EXISTS IN WFIRST COLOR REGIONS. FINALLY WE WILL DEVELOP A STATISTICAL METHOD TO DETERMINE HOW MANY SPECTROSCOPIC REDSHIFTS ARE NEEDED IN EACH CELL OF WFIRST COLOR SPACE TO ACCURATELY MAP FROM COLOR TO REDSHIFT AND WHICH COLOR SPACE CELLS SHOULDBE EXCISED FROM THE WEAK LENSING ANALYSIS DUE TO REDSHIFT DEGENERACY. IN ADDITION TO PROVIDING AN OPTIMAL TRAINING SET FOR MACHINE LEARNING THIS METHOD WILL BE INVERTED TO PROVIDE BAYESIAN PRIORS FOR TEMPLATE FITTING ALGORITHMS WHICH TYPICALLY USE EITHER NO PRIOR OR AD HOC ONES. THIS WILL PROVIDE AN ALTERNATIVE PATH TO OBTAINING THE REQUIRED PHOTOMETRIC REDSHIFTS AND MAY ALSO SHED LIGHT ON GALAXYEVOLUTION. MOREOVER THE COLOR SPACE MAPPING PROVIDES A MEANS TO AUTOMATICALLY IDENTIFY RARE AND INTERESTING ROSETTA STONE OBJECTS IN THE WFIRST DATA. ($27) 3/6/18 3 PRECISION PHOTOMETRIC REDSHIFTS FOR COSMOLOGY THE GROWTH OF STRUCTURE AS MEASURED BY WEAK LENSING HAS BEEN IDENTIFIED AS ONE OF THE MOST SENSITIVE PROBES OF DARK ENERGY AND DARK MATTER, AND IS ONE OF THE THREE KEY DARK ENERGY EXPERIMENTS PROPOSED FOR WFIRST. HOWEVER, THE WEAK LENSING MEASUREMENT DEPENDS STRONGLY ON ROBUST PHOTOMETRIC REDSHIFTS, AND IS HIGHLY SENSITIVE TO SYSTEMATIC BIASES IN THESE REDSHIFT ESTIMATES. SEVERAL METHODS HAVE BEEN PROPOSED TO REMOVE SYSTEMATIC BIASES BASED ON SPECTROSCOPIC SAMPLES AND SPATIAL CLUSTERING, BUT NONE HAS BEEN DEMONSTRATED TO PERFORM AT THE LEVEL REQUIRED FOR WFIRST. MAKING THE PROBLEM MORE CHALLENGING, AT LEAST TWO INDEPENDENT METHODS MUST BE DEVELOPED: ONE TO CORRECT THE SYSTEMATIC ERRORS, AND ANOTHER TO VERIFY THE CORRECTION AND QUANTIFY RESIDUAL ERROR. HERE WE PROPOSE TO DEVELOP AN INFORMED CALIBRATION OF THE COLOR-REDSHIFT RELATION THAT WILL MINIMIZE THE NUMBER OF SPECTROSCOPIC REDSHIFTS NEEDED FOR MACHINE LEARNING ALGORITHMS AND PRODUCE ACCURATE BAYESIAN PRIORS FOR TEMPLATE FITTING ALGORITHMS. THE PROPOSED METHOD USES OUR CURRENT KNOWLEDGE OF GALAXIES AND GALAXY EVOLUTION FROM EXISTING DEEP SURVEYS TO PARAMETERIZE WHERE IN THE WFIRST COLOR SPACE THE PHOTOMETRIC REDSHIFTS ARE WELL UNDERSTOOD, AND WHERE THEY ARE NOT. FIRST, WE WILL DEVELOP A METHOD TO MAP FROM THE WFIRST N-DIMENSIONAL COLOR SPACE TO REDSHIFT. THIS WILL DETERMINE WHICH REGIONS OF COLOR SPACE MAP TO REDSHIFT IN A WELL-BEHAVED WAY, AND WHICH HAVE A MORE COMPLEX BEHAVIOR. WE WILL MAKE USE OF THE FACT THAT HIGHER-DIMENSIONAL DATA (NARROWER BAND PASSES, MORE SENSITIVE DATA, AND LARGER SPECTRAL COVERAGE) ARE AVAILABLE IN SELECT AREAS OF THE SKY TO DETERMINE HOW MUCH UNCERTAINTY EXISTS IN WFIRST COLOR REGIONS. FINALLY, WE WILL DEVELOP A STATISTICAL METHOD TO DETERMINE HOW MANY SPECTROSCOPIC REDSHIFTS ARE NEEDED IN EACH CELL OF WFIRST COLOR SPACE TO ACCURATELY MAP FROM COLOR TO REDSHIFT, AND WHICH COLOR SPACE CELLS SHOULD BE EXCISED FROM THE WEAK LENSING ANALYSIS DUE TO REDSHIFT DEGENERACY. IN ADDITION TO PROVIDING AN OPTIMAL TRAINING SET FOR MACHINE LEARNING, THIS METHOD WILL BE INVERTED TO PROVIDE BAYESIAN PRIORS FOR TEMPLATE FITTING ALGORITHMS, WHICH TYPICALLY USE EITHER NO PRIOR, OR AD HOC ONES. THIS WILL PROVIDE AN ALTERNATIVE PATH TO OBTAINING THE REQUIRED PHOTOMETRIC REDSHIFTS, AND MAY ALSO SHED LIGHT ON GALAXY EVOLUTION. MOREOVER, THE COLOR SPACE MAPPING PROVIDES A MEANS TO AUTOMATICALLY IDENTIFY RARE AND INTERESTING ROSETTA STONE OBJECTS IN THE WFIRST DATA. Funding Only Action ($27) 3/6/18 Not listed PRECISION PHOTOMETRIC REDSHIFTS FOR COSMOLOGYTHE GROWTH OF STRUCTURE AS MEASURED BY WEAK LENSING HAS BEEN IDENTIFIED AS ONE OF THE MOST SENSITIVE PROBES OF DARK ENERGY AND DARK MATTER AND IS ONE OF THE THREE KEY DARK ENERGY EXPERIMENTS PROPOSED FOR WFIRST. HOWEVER THE WEAK LENSING MEASUREMENT DEPENDS STRONGLY ON ROBUST PHOTOMETRIC REDSHIFTS AND IS HIGHLY SENSITIVE TO SYSTEMATIC BIASES IN THESE REDSHIFT ESTIMATES. SEVERAL METHODS HAVE BEEN PROPOSED TO REMOVE SYSTEMATIC BIASES BASED ON SPECTROSCOPIC SAMPLES AND SPATIAL CLUSTERING BUT NONE HAS BEEN DEMONSTRATED TO PERFORM AT THE LEVEL REQUIRED FOR WFIRST. MAKING THE PROBLEM MORE CHALLENGING AT LEAST TWO INDEPENDENT METHODS MUST BE DEVELOPED: ONE TO CORRECT THE SYSTEMATIC ERRORS AND ANOTHER TO VERIFY THE CORRECTION AND QUANTIFY RESIDUAL ERROR. HERE WE PROPOSE TO DEVELOP AN INFORMED CALIBRATION OF THE COLOR-REDSHIFT RELATION THAT WILL MINIMIZE THE NUMBER OF SPECTROSCOPIC REDSHIFTS NEEDED FOR MACHINE LEARNING ALGORITHMS AND PRODUCE ACCURATE BAYESIAN PRIORS FOR TEMPLATE FITTING ALGORITHMS.THE PROPOSED METHOD USES OUR CURRENT KNOWLEDGE OF GALAXIES AND GALAXY EVOLUTION FROM EXISTING DEEP SURVEYS TO PARAMETERIZE WHERE IN THE WFIRST COLOR SPACE THE PHOTOMETRIC REDSHIFTS ARE WELL UNDERSTOOD AND WHERE THEY ARE NOT. FIRST WE WILL DEVELOP A METHOD TO MAP FROM THE WFIRST N-DIMENSIONAL COLOR SPACE TO REDSHIFT. THIS WILL DETERMINE WHICH REGIONS OF COLOR SPACE MAP TO REDSHIFT IN A WELL-BEHAVED WAY AND WHICH HAVE A MORE COMPLEX BEHAVIOR. WE WILL MAKE USE OF THE FACT THAT HIGHER-DIMENSIONAL DATA(NARROWER BAND PASSES MORE SENSITIVE DATA AND LARGER SPECTRAL COVERAGE) ARE AVAILABLE IN SELECT AREAS OF THE SKY TO DETERMINE HOW MUCH UNCERTAINTY EXISTS IN WFIRST COLOR REGIONS. FINALLY WE WILL DEVELOP A STATISTICAL METHOD TO DETERMINE HOW MANY SPECTROSCOPIC REDSHIFTS ARE NEEDED IN EACH CELL OF WFIRST COLOR SPACE TO ACCURATELY MAP FROM COLOR TO REDSHIFT AND WHICH COLOR SPACE CELLS SHOULDBE EXCISED FROM THE WEAK LENSING ANALYSIS DUE TO REDSHIFT DEGENERACY. IN ADDITION TO PROVIDING AN OPTIMAL TRAINING SET FOR MACHINE LEARNING THIS METHOD WILL BE INVERTED TO PROVIDE BAYESIAN PRIORS FOR TEMPLATE FITTING ALGORITHMS WHICH TYPICALLY USE EITHER NO PRIOR OR AD HOC ONES. THIS WILL PROVIDE AN ALTERNATIVE PATH TO OBTAINING THE REQUIRED PHOTOMETRIC REDSHIFTS AND MAY ALSO SHED LIGHT ON GALAXYEVOLUTION. MOREOVER THE COLOR SPACE MAPPING PROVIDES A MEANS TO AUTOMATICALLY IDENTIFY RARE AND INTERESTING ROSETTA STONE OBJECTS IN THE WFIRST DATA. $0 5/2/17 2 PRECISION PHOTOMETRIC REDSHIFTS FOR COSMOLOGY THE GROWTH OF STRUCTURE AS MEASURED BY WEAK LENSING HAS BEEN IDENTIFIED AS ONE OF THE MOST SENSITIVE PROBES OF DARK ENERGY AND DARK MATTER, AND IS ONE OF THE THREE KEY DARK ENERGY EXPERIMENTS PROPOSED FOR WFIRST. HOWEVER, THE WEAK LENSING MEASUREMENT DEPENDS STRONGLY ON ROBUST PHOTOMETRIC REDSHIFTS, AND IS HIGHLY SENSITIVE TO SYSTEMATIC BIASES IN THESE REDSHIFT ESTIMATES. SEVERAL METHODS HAVE BEEN PROPOSED TO REMOVE SYSTEMATIC BIASES BASED ON SPECTROSCOPIC SAMPLES AND SPATIAL CLUSTERING, BUT NONE HAS BEEN DEMONSTRATED TO PERFORM AT THE LEVEL REQUIRED FOR WFIRST. MAKING THE PROBLEM MORE CHALLENGING, AT LEAST TWO INDEPENDENT METHODS MUST BE DEVELOPED: ONE TO CORRECT THE SYSTEMATIC ERRORS, AND ANOTHER TO VERIFY THE CORRECTION AND QUANTIFY RESIDUAL ERROR. HERE WE PROPOSE TO DEVELOP AN INFORMED CALIBRATION OF THE COLOR-REDSHIFT RELATION THAT WILL MINIMIZE THE NUMBER OF SPECTROSCOPIC REDSHIFTS NEEDED FOR MACHINE LEARNING ALGORITHMS AND PRODUCE ACCURATE BAYESIAN PRIORS FOR TEMPLATE FITTING ALGORITHMS. THE PROPOSED METHOD USES OUR CURRENT KNOWLEDGE OF GALAXIES AND GALAXY EVOLUTION FROM EXISTING DEEP SURVEYS TO PARAMETERIZE WHERE IN THE WFIRST COLOR SPACE THE PHOTOMETRIC REDSHIFTS ARE WELL UNDERSTOOD, AND WHERE THEY ARE NOT. FIRST, WE WILL DEVELOP A METHOD TO MAP FROM THE WFIRST N-DIMENSIONAL COLOR SPACE TO REDSHIFT. THIS WILL DETERMINE WHICH REGIONS OF COLOR SPACE MAP TO REDSHIFT IN A WELL-BEHAVED WAY, AND WHICH HAVE A MORE COMPLEX BEHAVIOR. WE WILL MAKE USE OF THE FACT THAT HIGHER-DIMENSIONAL DATA (NARROWER BAND PASSES, MORE SENSITIVE DATA, AND LARGER SPECTRAL COVERAGE) ARE AVAILABLE IN SELECT AREAS OF THE SKY TO DETERMINE HOW MUCH UNCERTAINTY EXISTS IN WFIRST COLOR REGIONS. FINALLY, WE WILL DEVELOP A STATISTICAL METHOD TO DETERMINE HOW MANY SPECTROSCOPIC REDSHIFTS ARE NEEDED IN EACH CELL OF WFIRST COLOR SPACE TO ACCURATELY MAP FROM COLOR TO REDSHIFT, AND WHICH COLOR SPACE CELLS SHOULD BE EXCISED FROM THE WEAK LENSING ANALYSIS DUE TO REDSHIFT DEGENERACY. IN ADDITION TO PROVIDING AN OPTIMAL TRAINING SET FOR MACHINE LEARNING, THIS METHOD WILL BE INVERTED TO PROVIDE BAYESIAN PRIORS FOR TEMPLATE FITTING ALGORITHMS, WHICH TYPICALLY USE EITHER NO PRIOR, OR AD HOC ONES. THIS WILL PROVIDE AN ALTERNATIVE PATH TO OBTAINING THE REQUIRED PHOTOMETRIC REDSHIFTS, AND MAY ALSO SHED LIGHT ON GALAXY EVOLUTION. MOREOVER, THE COLOR SPACE MAPPING PROVIDES A MEANS TO AUTOMATICALLY IDENTIFY RARE AND INTERESTING ROSETTA STONE OBJECTS IN THE WFIRST DATA. Funding Only Action $0 5/2/17 Not listed PRECISION PHOTOMETRIC REDSHIFTS FOR COSMOLOGYTHE GROWTH OF STRUCTURE AS MEASURED BY WEAK LENSING HAS BEEN IDENTIFIED AS ONE OF THE MOST SENSITIVE PRO $212.5k 4/28/16