Focal Ratio Degradation Measurements of BigBOSS Optical Fiber Bundles

Abstract

BigBOSS (Big Baryon Oscillation Spectroscopic Survey) will target luminous red galaxies, emission line galaxies, and QSOs. It will survey the universe over 80% of the way back to the Big Bang, and make a map of 50 million galaxy positions. This project will be placed on the NOAO 4-m Mayall telescope at Kitt Peak, Arizona, USA, measuring 5000 galaxies, simultaneously through optical fibers. Focal Ratio Degradation (FRD) is the non-conservation of incident cone angle in an optical fiber, resulting in a spreading of the beam at the output. This increase without a gain in information content is undesirable due to reduction of signal to noise ratio (SNR) in the telescope. For BigBOSS, in order to maintain the same total signal count, the signal has to be integrated over a larger area of the CCD. The background CCD thermal and other noise background being constant is integrated over a larger area. Thus SNR becomes poor. If the FRD is large enough, the output image on CCD can be larger than the CCD resulting in signal loss. We want FRD to select fibers with small FRD and reject the ones with large FRD to overcome the outcome resulting poor SNR and possible loss of signal. The optical system has to deal with this is consequently more complex and expensive. Most spectroscopic techniques applied in astronomy rely on as small a FRD as possible so that the longitudinal input solid angle is as close to the output solid angle. Apart from the beam spreading caused by the imperfections of the fiber, there are other mechanisms that contribute to increase in FRD，such as fiber end preparation procedures including polishing and cleaving. It is important to be able to predict and measure the performance of optical fiber in order to develop criteria for fiber selection and rejection. We have compared fiber FRD performance with three methods- using total counts, intensity and 80% flux passing fiber radius and concluded with a matrix for fiber selection and rejection. This allows us to selectively choose the 5000 fibers with good FRD performance by replacing the bad ones. We can now account for FRD loss along the fiber as well as the front and back surface preparation effects. We have algorithms developed to correct for fiber tip misalignment. For BigBOSS, in order to maintain the same total signal count, the signal has to be integrated over a larger area of the CCD. The background CCD thermal and other noise background being constant is integrated over a larger area. Thus SNR becomes poor. If the FRD is large enough, the output image on CCD can be larger than the CCD result-ing in signal loss. We want FRD to select fibers with small FRD and reject the ones with large FRD to overcome the outcome resulting poor SNR and possible loss of signal. The optical system has to deal with this is consequently more complex and expensive. Most spectroscopic techniques applied in astronomy rely on as small a FRD as possible so that the longitudinal input solid angle is as close to the output solid angle. Apart from the beam spreading caused by the imperfections of the fiber, there are other mechanisms that contribute to increase in FRD，such as fiber end preparation procedures including polish-ing and cleaving. It is important to be able to predict and measure the performance of opt-ical fiber in order to develop criteria for fiber selection and rejection. We have compared fiber FRD performance with three methods - using total counts, intensity and 80% flux passing fiber radius and concluded with a matrix for fiber selection and rejection. This allows us to selectively choose the 5000 fibers with good FRD performance by replacing the bad ones. We can now account for FRD loss along the fiber as well as the front and back surface preparation effects. We have algorithms devel-oped to correct for fiber tip misalignment.;BigBOSS는 빛는 발하는 red galaxies, 발사하는 line galaxies와 QSOs를 목표로 한다. BigBOSS는 빅뱅 당시의 80%을 넘는 우주를 관측할 수 있다. 그리고 50 million galaxy 위치의 지도를 그려낼 수 있다. 이 프로젝트는 NOAO 4-m Mayall telescope at Kitt Peak, Arizona, USA에서 진행될 것이며 optical fibers를 통해 5000 galaxies를 측량할 수 있다. Focal Ratio Degradation (FRD) 은 incident cone angle이 fiber의 전파과정에서 원래의 각도를 보존하지 못하여, output에 확장된 beam를 형성한다. Telescope의 signal to noise ratio (SNR)을 놓고 보면, FRD의 확장을 원하지 않는다. BigBOSS에서 동일한 total signal count를 보존하기 위하여 signal을 큰 범위의 CCD에 집중시켰다. CCD thermal 배경과 소음 배경은 상수로 변하여 하나의 큰 공간에 집중된다. 그리하여 SNR은 약하게 변한다. 만약 FRD가 충분히 크다면, CCD위의 output 이미지는 signal loss로 인한 CCD 이미지보다 클 수 있다. 우리는 작은 FRD fibers를 선택하고, 약한 SNR과 signal loss를 초래하는 큰 FRD fibers를 선택하기를 희망한다. 대부분의 spectroscopic techniques은 될수록 작은 FRD에 근거하여 천문학에 적용되는데 이는 longitudinal input solid angle을 output solid angle에 될수록 근접하게 한다. Fiber의 불완전함으로 인해 beam의 전파과정에서 FRD를 증가 시킬 수 있는 것 외에, 또 다른 원인이 FRD의 증가를 초래할 수 있다. 예를 들면, fiber end의 준비과정인 polishing 와 cleaving이 있다. Fiber의 선택을 위하여, fiber의 성능을 예측, 측량할 수 있는 것은 아주 중요하다. 우리는 3가지 방법으로 fiber의 성능-- total counts, intensity 과 80% flux passing fiber radius을 비교하고 마지막에 fiber을 선택했다. 이 3가지 방법을 통하여 우리는 5000 fibers에서 좋은 성능의 FRD를 선택과 교환을 할 수 있다. 우리는 fiber로 인한 FRD loss를 앞뒤표면의 preparation effects로 해석할 수 있다. 우리는 이미 fiber tip misalignment의 산법에까지 발정하였다.