Kea : A New Tool to Obtain Stellar Parameters from Low to Moderate Signal-to-noise and High-resolution Echelle Spectra

In this paper, we describe Kea a new spectroscopic fitting method to derive stellar parameters from moderate to low signal-to-noise, high-resolution spectra. We developed this new tool to analyze the massive data set of the Kepler mission reconnaissance spectra that we have obtained at McDonald Observatory. We use Kea to determine effective temperatures (T sub(eff)), metallicity ([Fe/H]), surface gravity (log g), and projected rotational velocity (v sin i). Kea compares the observations to a large library of synthetic spectra that covers a wide range of different T sub(eff), [Fe/H], and log g values. We calibrated Kea on observations of well-characterized standard stars (the Kepler field "platinum" sample) that range in T sub(eff) from 5000 to 6500 K, in [Fe/H] from -0.5 to +0.4 dex, and in log g from 3.2 to 4.6 dex. We then compared the Kea results from reconnaissance spectra of 45 Kepler objects of interest (KOIs) to stellar parameters derived from higher signal-to-noise spectra obtained with Keck/HIRES. We find typical uncertainties of 100 K in T sub(eff), 0.12 dex in [Fe/H], and 0.18 dex in log g.