Subjective mapping of dust-emission sources by using MODIS imagery : reproducibility assessment

Abstract

Dust storms (5 to 100 km across) often originate from multiple dust-emission sources (1 to 10 km across). Remote-sensing-based dust-source identification is a challenge. A previous study developed a subjective approach for mapping dust sources by using enhanced MODIS satellite imagery; therefore, this study conducted mapping exercises to assess the reproducibility of this technique amongst multiple analysts and in different regions. Multiple staff members independently analyzed satellite imagery for mappable dust sources for Southwest Asia and the Southwest United States. Mapped points were considered reproducible if the location of dust emission plumes identified by all participants could be constrained to a 10 km buffer. Using this definition, point-source locations were 28% reproducible in Southwest Asia and 85% reproducible in the Southwest United States. Increasing the allotted buffer to 15 km, however, improved results to 71% and 100%, respectively. Mapped dust sources for Southwest Asia were compared to geomorphic landform maps. At the 1:750,000 map scale, points mapped by all analysts for a single dust plume tended to overlay one landform, while at the 1:100,000 map scale, points were strewn across several landforms. Results suggest that the methodology is reproducible for certain applications but that location-uncertainty tolerance affects perceived conclusions. DISCLAIMER: The contents of this report are not to be used for advertising, publication, or promotional purposes. Citation of trade names does not constitute an official endorsement or approval of the use of such commercial products. All product names and trademarks cited are the property of their respective owners. The findings of this report are not to be construed as an official Department of the Army position unless so designated by other authorized documents. DESTROY THIS REPORT WHEN NO LONGER NEEDED. DO NOT RETURN IT TO THE ORIGINATOR. DISCLAIMER: The contents of this report are not to be used for advertising, publication, or promotional purposes. Citation of trade names does not constitute an official endorsement or approval of the use of such commercial products. All product names and trademarks cited are the property of their respective owners. The findings of this report are not to be construed as an official Department of the Army position unless so designated by other authorized documents. DESTROY THIS REPORT WHEN NO LONGER NEEDED. DO NOT RETURN IT TO THE ORIGINATOR.

The AFWA dust emission scheme for the GOCART aerosol model in WRF-Chem v3.8.1

“…Image dust enhancement was performed using a processing algorithm by Miller (2003), in which atmospheric dust is distinguished from the underlying background terrain using visible, near-infrared, thermal infrared, and water vapor channels. The script used for acquiring MODIS granules and generating imagery in GeoTiff format is available in Sinclair and Jones (2017). We also use the 1 km resolution MODIS MCD19A2 daily AOD product (Lyapustin and Wang, 2018) provided by the NASA Land Processes Distributed Active Archive Center (LP DAAC), USGS/Earth Resources Observation and Science (EROS) Center, Sioux Falls, South Dakota (https://lpdaac.usgs.gov/data_access/data_pool, last access: June 2017) to evaluate the simulated AOD.…”

A comparison of simulated dust produced by three dust-emission schemes in WRF-Chem : case-study assessment

“…However, these observations provide almost no information regarding atmospheric dust concentrations (only surface visibility measurements used as a proxy for dust concentrations provide this information). Therefore, to qualitatively assess spatial dust concentrations, dust-enhanced satelliteimagery derived from Moderate Resolution Imaging Spectroradiometer (MODIS) data using a processing algorithm by Miller (2003) was used (see Sinclair and Jones [2017] for the python script we used to generate dust-enhanced imagery in GeoTiff format). In the Miller (2003) technique, atmospheric dust is distinguished from the underlying background terrain using visible, near infrared, thermal infrared, and water vapor channels.…”

“…It was unclear from the MODIS image if this error was due to issues in dustsource representation or a deficiency in how the WRF-Chem model handled dust deposition as the MODIS image only provides qualitative information about atmospheric dust conditions at a single date and time. Furthermore, it is unlikely that subjective dust-source identification methods (e.g., Walker et al 2009;Sinclair and Jones 2017) would help for this case study due to the presence of clouds.…”

Reproducibility assessment and uncertainty quantification in subjective dust source mapping

“…To examine the role of user subjectivity, four analysts independently mapped dust plume-head point sources from MODIS imagery following the procedures outlined in Sinclair and Jones (2017). The methodology used here differs from the original W09 technique in that 1) analysts place a single point at the location of an unobscured plume head, rather than outlining the plume head with a curve, and 2) analysts assign each mapped point a quality score (confidence level in marker placement) on a qualitative scale of 1-3, where 3 is the most confident and 1 is the least confident.…”