Spectroscopic Cities
Project title: Comparative Analysis of Spectroscopic Mixing Spaces: To what Extent Does Fine Particulate Cover Affect Impervious Surface Reflectance in Global Built Environments?
Capturing diagnostic information about the composition of Earth's land surface is a fundamental goal of satellite imaging spectroscopy. But the majority of terrestrial photons observed only interact with the “skin” (approx. surficial 1 micron) of the solid Earth. Fortunately, surficial properties can often be diagnostic of subsurface composition (e.g., in-situ physical weathering of autochthonous rock units).
But important circumstances may also exist where the compositional information of interest might be partially or completely masked by fine particulate coatings (e.g., dust). This problem has been noted in planetary spectroscopy for decades, and has helped motivate the development of iconic optical models. Particulate coatings on plant leaves can drive meaningful changes in physiology. Fine particulate coatings can be especially pervasive in and around human settlements, where both human (vehicle and industrial emissions) and natural (eolian geologic) sources can each be major sources especially in arid regions.
Reflectance spectra from high signal-to-noise spaceborne imaging spectrometers like EMIT can provide diagnostic information on the built environment. In particular, spectroscopic absorption features can clearly identify far more anthropogenic materials and particulate covers than is possible with multispectral imaging. For problems relevant to mapping built environments, e.g. pervious versus impervious surface distinction for hydrological modeling, spectroscopic imagery may enable improvement in current products but this question remains open. Specifically, the potential of fine particulate coatings to mask underlying material composition may (or may not) seriously limit mapping accuracy and precision.
Preliminary analysis of 30 EMIT subscenes from urban cores worldwide reveals geographically distinct and spectrally separable differences in both spectral continuum shape and SWIR absorption features of impervious substrates (https://arxiv.org/abs/2307.04716). Despite compositional heterogeneity of urban impervious substrates at the scale of the EMIT IFOV, this suggests that many combinations of impervious substrates may also be spectrally distinguishable from pervious substrates (e.g. soils) in peripheral non-built environments. Further, this early analysis has found both remarkable geographic consistency in reflectance of geographically proximal cities, and intriguing within-city gradients. But our early results leave many more questions than answers: What component is compositional versus surficial? To what extent do spectral consistencies generalize among biomes, climates and geologic settings? How do intra- versus inter-settlement gradients compare in both spectral amplitude and continuity? Can specific physical drivers of spectral gradients be identified?
The fundamental goals of this grant are thus to:
1. Characterize geographic consistency and spatial gradients in spectral properties within & among a diverse global compilation of ~100 built environments.
2. For a subset of built environments with existing AVIRIS data, use airborne imaging spectroscopy for: a) multiscale comparison of urban dust cover; and b) partition dust among surfaces likely versus unlikely to accumulate (e.g., untrafficked roofs vs high traffic roadways).
3. Investigate specific instances of spatial variations in suspected fine particulate cover across urban development density gradients, and validate with field and lab reflectance spectroscopy.
This research can help provide fundamental physical constraints on important Earth surface properties while offering hands-on, intuition-building experience to a new generation of imaging spectroscopists. Work is also integrated into the West Coast NASA Student Airborne Research Program (SARP West), for which the PI is faculty mentor for terrestrial spectroscopy (the “Land” group).
Project Team
Daniel Sousa
San Diego State University
-
Christopher Small
Columbia University