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

Marijn Siemons, Ralf Kohlhaas · 2024

Read Theoretical performance limitations and filter selection based on Fisher information of a computational photonic crystal spectrometer for trace-gas retrieval by Marijn Siemons, Ralf Kohlhaas (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2411-02048.txt.

Description

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.

Catalog record

CREATORMarijn Siemons, Ralf Kohlhaas
YEAR2024
COLLECTIONPAPERS
FORMATTXT
FILE SIZE28 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2411-02048

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Theoretical performance limitations and filter selection based on Fisher information of a computational photonic crystal spectrometer for trace-gas retrieval Marijn Siemons, Ralf Kohlhaas 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 Cramér-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 Cramér-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 300 × \times 300 m 2 ground resolution element and realistic instrument parameters. a]SRON Netherlands Institute for Space Research, 2333 CA Leiden, The Netherlands b]Department of Precision and Microsystems Engineering, Delft University of Technology, 2628 CD Delft, The Netherlands keywords Remote Sensing, compressive sensing, computational spectroscopy, trace gas retrieval, photonic crystals † † authorinfo: Further author information: (Send correspondence to M.S.) M.S.: E-mail: [email protected] 1 INTRODUCTION The need for the monitoring of greenhouse gas emission sources motivates the development of earth observation systems with a higher spatial and temporal resolution. For this complete distributed systems with many satellites could be considered, which would require small, low-weight optical imaging spectrometers. Conventional grating spectrometers have intrinsic physical limits on the system size due to the need of large free-space optical path length for the separation of wavelengths. Other solutions, as for example static Fourier-transform spectrometers as developed with NanoCarb [ 1 ] , suffer less from such limitations but have reduced spectral resolution. Recently, a spectrometer concept was invented [ 2 , 3 ] which uses an array of photonic crystal filters with quasi-random spectral transmission functions together with computational inversion (for example compressive sensing) to reconstruct the input spectrum. Here we introduce a novel instrument concept, using compressive sensing enabled by photonic crystals, see Figure 1 . The instrument orbits the earth and consists of an earth observing optical telescope, where photonic crystals are positioned on the sensor. These photonic crystals consists of a wafer with a thin dielectric material. In our case this is a glass wafer with a 700 nm thin amorphous Si-film. It contains 2D photonic crystals in a column wise arrangement in the across-track (ACT) direction. These photonic crystals consist of sub-wavelength features etched in the Si-film and have unique spectral transmission profiles. These transmission profiles arise from resonances in the Si-film and depend on the thickness of the film as well as the specific shape and size of the etched structure and the lattice size. By choosing a specific hole shape and lattice size, different spectral transmission profiles can be achieved. Using the push-broom concept, the transmitted intensity for each ground instantaneous field of view pixel can be measured for each spectral filter. From the measured detector values, and the known transmission profiles the trace gas concentration can be estimated using a retrieval algorithm and atmosphere model. This concept allows a far larger field-of-view in along-track (ALT) direction than with an imaging spectrometer with a slit. This concept has similarities with a linear gradient filter [ 4 ] , but the photonic crystal transmission allow for a more diverse transmission profiles and a highly tuned and selective optical system. Moreover, different spectral ranges can be combined on one detector (for example the O2A band and methane band, or several relevant methane spectral ranges). Figure 1: Illustration of the instrument concept. A) The instrument consists of a telescope and sensor. A photonic crystal filter array is placed directly on the sensor. B) As the instrument flies over the earth, ground pixels are observed with different flilters. C) The photonic crystals filter array consists of strips in the across track direction. D) The photonic crystal filters itself is made with a small layer of amorphous silicon (aSi) on top of a quartz substrate. In the aSi-layer 2D patterns are etched, giving rise to distinct and spectrally rich transmission profiles (E). A particular interesting advantage is that the photonic crystal filters can be chosen to optimize the information measured by the instrument to tailor trace gas retrieval. This is however very challenging. The photonic crystals can can consist of any different shapes, shape size and lattice spacing, each with their own unique transmission profile. Selecting the best filter set is a non-trivial and difficult optimization problem. Assuming a filter library of around 4000 filters and a maximum of 64 filters to be used, the total amount of combination from the library is 4000 choose 64 = 10^{21} . This number is so large, all combinations can never be evaluated in linear computational time. Therefore, an algorithm or procedure is needed to find a good performing filter set. Fisher information, a concept from mathematical statistics, is a way of quantifying the amount of information a random variable (in this case the measured detector counts) carries about an unknown (fit) parameter. Photonic crystal filters which carry more Fisher information will result in more precise trace gas retrieval. In estimation theory, the Cramér-Rao Lower Bound (CRLB) is measure of the best achievable precision for a fit parameter, which can be obtained with an unbiased estimator, i.e. fit algorithm, based on the available Fisher information. In practice, fit algorithms are often biased to some extend and therefore not strictly bounded by the CRLB. In this work we derive a lower bound for trace-gas retrieval for this computational photonic crystal spectrometer using the Fisher information and compare this with the achieved performance. We furthermore set up a frame work how to select photonic crystal filters based on maximizing the Fisher information carried by the filters and how to use the CRLB to find good filter sets. A filter set with a low CRLB is however to guarantee that the retrieval algorithm will achieve this precision. Therefore, the CRLB can only be used to find good candidate filter sets, but actual achieved trace gas retrieval precision is used for final selection of a filter set. The outline of this proceeding is as follows. First we describe the instrument and atmosphere model and the trace gas retrieval algorithm. Next, the photonic crystal simulations are discussed in section 3. Then the Fisher information is derived for the instrument and the filter selection procedure is described in section 4. In

Read the complete text