Construction-related soil disturbance, such as road construction, trenching, landstripping, earthmoving and blasting, is a
significant source of fugitive (or airborne) dust, and fugitive dust is a potential health hazard as well as a primary cause
of decreased air quality. Presented here is a remote sensing change detection method using annual Landsat Thematic
Mapper (TM) images spanning 1995 and 2009 over southern Arizona to identify and characterize construction-related
soil disturbance. To guide development of the remote sensing method, spatial coordinates of construction activity permit
inspections performed by a local environmental quality agency to control fugitive dust are obtained and processed in a
GIS. Satellite change detection methods are compared with kernel density plots of the construction activity inspection
points. Band differencing in the mid-infrared spectral region (TM band 5), with a change threshold of four standard
deviations above and below the change image mean, is identified as a simple and effective method to identify
construction-related soil disturbance. As an accuracy assessment, buffers of 920 meter radius were generated around
each dust inspection point and around an equal number of random points in a GIS. The dust inspection point buffers
captured statistically significantly more of the remote sensing change pixels as compared to random point buffers
(P < 0.0001 in a Mann-Whitney rank sum U-test for each year compared; 44.3% change pixel capture rate average as
compared to 16.2% for random points). The remote sensing model is used to estimate location and annual surface area
of construction-related soil disturbance in eastern Pima County, Arizona during the fourteen year study period. With
limited preprocessing and processing requirements, the proposed model is simple to perform and may be suited for
public and other environmental and health agencies to identify and assess fugitive dust sources and inputs to total
ambient dust predictive models.
|