In-operation field-of-view retrieval (IFR) for satellite and ground-based DOAS-type instruments applying coincident high-resolution imager data

Abstract

Knowledge of the field of view (FOV) of a remote sensing instrument is particularly important when interpreting their data and merging them with other spatially referenced data. Especially for instruments in space, information on the actual FOV, which may change during operation, may be difficult to obtain. Also, the FOV of ground-based devices may change during transportation to the field site, where appropriate equipment for the FOV determination may be unavailable. This paper presents an independent, simple and robust method to retrieve the FOV of an instrument during operation, i.e. the two-dimensional sensitivity distribution, sampled on a discrete grid. The method relies on correlated measurements featuring a significantly higher spatial resolution, e.g. by an imaging instrument accompanying a spectrometer. The method was applied to two satellite instruments, GOME-2 and OMI, and a ground-based differential optical absorption spectroscopy (DOAS) instrument integrated in an SO2 camera. For GOME-2, quadrangular FOVs could be retrieved, which almost perfectly match the provided FOV edges after applying a correction for spatial aliasing inherent to GOME-type instruments. More complex sensitivity distributions were found at certain scanner angles, which are probably caused by degradation of the moving parts within the instrument. For OMI, which does not feature any moving parts, retrieved sensitivity distributions were much smoother compared to GOME-2. A 2-D super-Gaussian with six parameters was found to be an appropriate model to describe the retrieved OMI FOV. The comparison with operationally provided FOV dimensions revealed small differences, which could be mostly explained by the limitations of our IFR implementation. For the ground-based DOAS instrument, the FOV retrieved using SO2-camera data was slightly smaller than the flat-disc distribution, which is assumed by the state-of-the-art correlation technique. Differences between both methods may be attributed to spatial inhomogeneities. In general, our results confirm the already deduced FOV distributions of OMI, GOME-2, and the ground-based DOAS. It is certainly applicable for degradation monitoring and verification exercises. For satellite instruments, the gained information is expected to increase the accuracy of combined products, where measurements of different instruments are integrated, e.g. mapping of high-resolution cloud information, incorporation of surface climatologies. For the SO2-camera community, the method presents a new and efficient tool to monitor the DOAS FOV in the field

<em>Pyplis</em> - A Python Software Toolbox for the Analysis of SO₂ Camera Data

“…220 2. IFR routine: this method is based on [25] and uses an inversion algorithm to retrieve the FOV.…”

“…[3], [4], [5], [6], and references therein). 25 The gas composition of the emission plumes can, inter alia, be studied using ground-based passive remote sensing techniques. The column-densities (CDs) of the gases in the plumes are quantified based on the absorption signature carried by scattered sunlight that has penetrated the plume.…”

“…Therefore, a parametrised FOV is fitted to the IFR results, which is more applicable to ground-based instruments than the parametrisation proposed by [25]. We propose the following elliptical Super-Gaussian FOV parametrisation g of the IFR result depending on pixel coordinates i, j in horizontal and vertical direction, respectively:…”

“…The FOV can either be measured 80 experimentally (e.g. in the lab) or can be retrieved directly from the field data ([18], [25]). The filter transmission curves shift towards lower wavelengths for non-perpendicular illumination.…”

“…[18]), Pyplis incorporates more recent developments. These 310 include an implementation of the DOAS FOV retrieval algorithm proposed by [25], a routine to correct for the signal dilution effect based on [19], or pixel based gas velocity retrievals using an optical flow algorithm (e.g. [20]).…”