The Impacts of Air Quality on Vegetation Health in Dense Urban Environments: A Ground-Based Hyperspectral Imaging Approach

Abstract

We examine the impact of changes in ozone (O3), particulate matter (PM2.5), temperature, and humidity on the health of vegetation in dense urban environments, using a very high-resolution, ground-based Visible and Near-Infrared (VNIR, 0.4–1.0 μm with a spectral resolution of 0.75 nm) hyperspectral camera deployed by the Urban Observatory (UO) in New York City. Images were captured at 15 min intervals from 08h00 to 18h00 for 30 days between 3 May and 6 June 2016 with each image containing a mix of dense built structures, sky, and vegetation. Vegetation pixels were identified using unsupervised k-means clustering of the pixel spectra and the time dependence of the reflection spectrum of a patch of vegetation at roughly 1 km from the sensor that was measured across the study period. To avoid illumination and atmospheric variability, we introduce a method that measures the ratio of vegetation pixel spectra to the spectrum of a nearby building surface at each time step relative to that ratio at a fixed time. This “Compound Ratio” exploits the (assumed) static nature of the building reflectance to isolate the variability of vegetation reflectance. Two approaches are used to quantify the health of vegetation at each time step: (a) a solar-induced fluorescence indicator (SIFi) calculated as the simple ratio of the amplitude of the Compound Ratio at 0.75 μm and 0.9 μm, and (b) Principal Component Analysis (PCA) decomposition designed to capture more global spectral features. The time dependence of these vegetation health indicators is compared to that of O3, PM2.5, temperature, and humidity values from a distributed and publicly available in situ air quality sensor network. Assuming a linear relationship between vegetation health indicators and air quality indicators, we find that changes in both SIF indicator values and PC amplitudes show a strong correlation (r2 value of 40% and 47%, respectively) with changes in air quality, especially in comparison with nearby buildings used as controls (r2 value of 1% and 4%, respectively, and with all molecular correlations consistent with zero to within 3σ uncertainty). Using the SIF indicator, O3 and temperature exhibit a positive correlation with changes in photosynthetic rate in vegetation, while PM2.5 and humidity exhibit a negative correlation. We estimate full covariant uncertainties on the coefficients using a Markov Chain Monte Carlo (MCMC) approach and demonstrate that these correlations remain statistically significant even when controlling for the effects of diurnal sun-sensor geometry and temperature variability. This work highlights the importance of quantifying the effects of various air quality parameters on vegetation health in urban environments in order to uncover the complexity, covariance, and interdependence of the numerous factors involved.

Atmospheric correction of vegetation reflectance with simulation-trained deep learning for ground-based hyperspectral remote sensing

“…Given that we lack the knowledge of the exact solar and atmospheric parameters of each obtained scene, to provide further confidence in the ability of the model to extract the vegetation spectral reflectance from at-sensor spectral radiance, we compare the Enc-Dec CNN predicted output to those produced by another method that does not rely on the synchronous knowledge of such parameters. The Compound Ratio, as presented in more detail in [ 21 ], exploits the temporally static nature of the spectral reflectance of built structures to produce the relative change in the apparent reflectance of adjacent vegetation. Provided the assumptions that buildings have constant reflectivity over moderately short time spans ( ), that the total irradiance incident on the target vegetation is identical to that incident on immediately adjacent buildings ( ), and that atmospheric transmission between the target and sensor is identical to that between the building and sensor ( ), the Compound Ratio of vegetation ( ) at time t is calculated as: Using the spectral radiance of the buildings immediately adjacent to the vegetation, and considering the first obtained scan as having time , we compute the Compound Ratio of the vegetation in each scene.…”

“…Prior to reaching the sensor, the light is also impacted by the atmosphere in the line of sight between the sensor and the target vegetation, where the impact magnitude can vary significantly depending on the platform and its distance to the target. Therefore, while the sensor-obtained spectral radiance contains information regarding the target vegetation, its heavy mixing with atmospheric and solar effects significantly impacts the quality of the extracted information, and its applicability to vegetation health inference in the presence of highly covariant variables such as atmospheric conditions and compositions [ 20 , 21 ].…”

“…Algorithmic methods used for the solar and atmospheric correction of HSI data can be grouped into two broad categories: scene-based empirical approaches, and radiative transfer approaches [ 23 ]. Scene-based empirical approaches, such as the Quick Atmospheric Correction (QUAC) [ 24 ], flat-field (FF) correction [ 25 ], and the Compound Ratio approach [ 21 ], rely on parameters being derived through prior knowledge or assumptions about in-scene elements. Such methods can be computationally efficient and accurate, but carry large uncertainties in the corrected spectral reflectance if insufficient information is obtained from the image, or when assumptions are idyllic [ 26 ].…”

“…Light reflecting off of a target object encounters atmospheres that substantially differ in composition, density, and length if travelling in the azimuth direction through to the top of the atmosphere and beyond, as opposed to travelling horizontally along the troposphere for up to several kilometers. The recent utilization of side-facing HSI platforms for the study of vegetation [ 18 , 21 ], demonstrates that there is interest in the development of methods for the atmospheric correction of spectral reflectance for ground-based platforms, including for cases in which precise atmospheric parameters at time of capture are unknown.…”

Carotenoid Content Estimation in Tea Leaves Using Noisy Reflectance Data

“…However, these approaches are both time-consuming and expensive, making them ill suited for tracing temporal changes in carotenoid content. In contrast, the use of hyperspectral reflectance offers a nondestructive approach for measuring carotenoid content that has been applied in forestry, vegetation, and environmental monitoring [13][14][15][16][17][18][19][20]. However, while field-portable spectroradiometers have been developed to obtain hyperspectral data, their high cost renders them impractical for use at the consumer level [21,22].…”