Comparative Evaluation of Fitness of Interpolation Techniques of ArcGIS Using Leave-One-Out Scheme for Air Quality Mapping

Awkash Kumar,Shraddha Dhakhwa,Anil K. Dikshit
DOI: https://doi.org/10.1007/s41651-022-00102-4
2022-03-14
Journal of Geovisualization and Spatial Analysis
Abstract:Air pollution is one of the most critical concerns encountered in urban areas, especially in developing nations. Four out of five cities in the world with the worst air quality are located in India. The concentration levels of air pollutants, including sulfur dioxide (SO2), nitrogen dioxide (NO2), and suspended particulate matter (SPM), are monitored by instituted air monitoring stations. The present study assesses the spatial air quality and the fitness of spatial mapping using data collected by monitoring stations in Mumbai, India. As part of this, air quality was monitored at twelve selected locations in Mumbai city. The collected data were spatially interpolated using the available interpolation tools of ArcGIS, including inverse distance weight (IDW), kriging (spherical and Gaussian), and spline techniques by the leave-one-out scheme. In each case, the interpolated concentrations of SO2, NO2, and SPM at the eleven locations were compared to the observed values of the unmonitored locations in the corresponding region. Percentage error, normal mean bias (NMB), normal root mean square error (NRMSE), and degree of agreement (“d”) were carried out to evaluate the spatial interpolation methods adopted for air quality mapping. The spline method is not preferred for spatial mapping due to negative concentrations in many locations due to either overestimation or underestimation. The minimum average percentage error for SO2 and NO2 mapping is achieved using IDW interpolation, while SPM mapping is achieved by the kriging method. The NMB and “d” were minimum for IDW, while NRMSE was minimum for the kriging Gaussian method in all the cases. The results of this study can be used in low-middle-income countries (LMIC) to obtain an aerial view of air quality to help the regulatory bodies in policy formulation and decision-making. This method needs fewer resources and time when compared to setting up new monitoring stations or using dispersion modeling to obtain an aerial perspective.
What problem does this paper attempt to address?