Evaluation of Rapid Conversion Zones and Atmospheric Nitrate Deposition Interactions using Multi-Criterion Modeling Approach and Remotely-Sensed Data

Abstract:

This study evaluates the impacts of the atmospheric deposition of nitrates (ADN) on rapid conversion zones (RCZ) in the Southeastern United States based on a multi-criterion modeling approach. The ADN estimates were obtained by spatial downscaling of the satellite remotely-sensed data products from MODIS (moderate resolution imaging spectroradiometer instrument on the Terra and Aqua satellites). The RCZ, where annual land cover changes occur between 2004 and 2014, were estimated using trained unsupervised classification, spatial superposition, and conditional objective functions. A considerable number of RCZ was observed, increase in developed spaces (7.1 %) and vegetation (4.6 %), and decrease of forested lands (13.9 %) during the ten years. The impacts of RCZ on atmospheric nitrate fluxes were identified by developing the RCZ-ADN model with the aid of geographically-weighted statistical modeling. The modeled spatial trends exhibited diverse linkages of RCZ and ADN spanning from Nash Sutcliffe Efficiency (NS): 0.19 to 0.46 and coefficient of determination (R²): 0.24 to 0.52 for six different land covers across the Southeastern United States. The RCZ-ADN model results portrayed strong positive spatiotemporal correlations for the deciduous forests (NS = 0.41; R² = 0.50), evergreen forests (NS = 0.39; R² = 0.47), and developed spaces (NS = 0.36; R² = 0.47). Therefore, rapid changes in land covers within multiple land uses act as a potential carrier of atmospheric nitrate deposition, and can be regarded as hotspots for atmospheric exposure and land cover monitoring. The study findings can be utilized in hydrologic assessment tools to represent the interactive impacts of atmosphere and surface processes in a broader space-time context.