Theoretical performance limitations and filter selection based on Fisher information of a computational photonic crystal spectrometer for trace-gas retrieval

Abstract

As global climate change severely impacts our world, there is an increasing demand to monitor trace gases with a high spatial resolution and accuracy. At the same time, these instruments need to be compact in order have constellations for short revisit times. Here we present a new spectrometer instrument concept for trace gas detection, where photonic crystals filters replace traditional diffraction based optical elements. In this concept, 2D photonic crystal slabs with unique transmission profiles are bonded on a detector inside a regular telescope. As the instrument flies over the earth, different integrated intensities for each filter are measured for a single ground resolution element with a regular telescope. From this detector data, trace gas concentrations are retrieved. As an initial test case we focused on methane and carbon dioxide retrieval and estimated the performance of such an instrument. We derive the Cram\'er-Rao lower bound for trace-gas retrieval for such a spectrometer using Fisher information and compare this with the achieved performance. We furthermore set up a framework how to select photonic crystal filters based on maximizing the Fisher information carried by the filters and how to use the Cram\'er-Rao lower bound to find good filter sets. The retrieval performance of such an instrument is found to be between 0.4% to 0.9% for methane and 0.2% to 0.5% for carbon dioxide detection for a 300x300 m2 ground resolution element and realistic instrument parameters.