Drylands Home | Human Health | Native Lands | Public Lands

The Inventory of North American Dust Sources: General Approach for Data to be Collected

This document is the result of meetings at the U.S. Geological Survey office in Denver, CO on August 31, 2005. The goal of these discussions was to prepare a draft framework for an inventory of sources of windblown dust with a focus on North America. This report will review the purpose of the inventory, the approach for data management, and a schedule for discussions with the community of scientists actively involved in dust-source research. We expect that the larger dust-research community will contribute to the design of the inventory and that the architecture for the North American dust-source inventory will serve as a template for describing dust sources and events elsewhere.

Purpose
At a workshop in Boulder, CO held in October 2004, a group of researchers gathered to discuss emissions of dust (http://esp.cr.usgs.gov/info/dust/). One of the recommendations emerging from this workshop was the construction of a central database of North American dust events and the characteristics of their sources. Following on this recommendation, we propose that this database focus on North America where a substantial history of data gathering exists but that the database also be expandable to include dust sources and events from other parts of the globe where the relevant data exist. It was recognized that prior to that time, dust emission studies had been performed piecemeal, from the perspective of a number of disciplines from geomorphology to agricultural science to atmospheric modeling. The participants acknowledged that the mechanisms allowing distributed sources to feed into large dust events are poorly understood, in spite of a long history of plot and laboratory-scale erosion studies and a more recent record of remote sensing observations that can characterize the dust storms themselves.

An inventory of dust sources and events will gather essential information available on sources and events in a variety of settings and conditions in one location for comparison. In some cases, this will enable archival data (meteorological, satellite, etc.) to be obtained to further investigate the meteorological, vegetative, and physical surface conditions associated with a given event. The database will illuminate consistent source regions and individual hot spots, as well as the conditions under which they are active. By collecting information on the landforms, vegetation cover, sediment availability, land use, hydrologic conditions, and meteorological conditions associated with dust storms, an unprecedented database will be established for exploring hypotheses leading to better understanding of contemporary and future dust-storm production. From such understanding, predictive models of dust production can be developed for use in atmospheric models or as a risk assessment tool for regions that may become dust sources under changing land use conditions. Although housed at the USGS, this will be a community database in that the data are provided by the community and available to the scientific community for research purposes.

Data Management
A flexible, web-based input framework will be designed to provide researchers with a method for adding data on events to the database. Although the focus is on the characteristics of events, the user will also be able to enter sources that were known to have been active in recorded history. Ultimately, the goal will be to have the data entry and retrieval completely automated such that a minimum of labor is necessary for database management by the USGS.

The data collection would have 3 levels of priority for information on dust events and their sources.

Level 1 Data (essential for any entry):

• Location (Lat./Long, UTM datum. With a code for accuracy of location) and geographic region from a list (northern Great Plains, Mojave Desert, etc.)
• Date and local/UTC time of event (the web interface will convert between local and UTC)
• Reference in literature or press

• Duration of event
• Magnitude of event (visibility, relative size, or other qualitative measure)
• Wind speed during event (time series during event if possible)
• Wind direction during event
• Remote sensing of event and location (satellite, airborne, ground photography)
• Area (size) of source
• Landforms at the source
• Land use/cover at source (e.g., Anderson classes)
• Land ownership at source (e.g., public agency (federal or state), military, private, etc.)
• Antecedent precipitation

• Magnitude of event (estimates of amount of dust mobilized, AOD, etc.)
• Sediment availability and soils and at the source
• Presence and fractional cover of physical or biotic crusts at the source
• Vegetation % cover at and near the source
• Vegetation composition/structure at and near the source (at least physiognomic type, preferably categories or quantification of structural parameters)
• Vegetation response to antecedent precipitation
• Land use history/human disturbance of vegetation or surface at the source
• Subsurface hydrology at the source
• Antecedent surface hydrology at the source
• Transport paths from the dust source
• Regional wind fields associated with the event and period leading up to event
• Amount of dust deposited by event and locations where deposited
• Chemistry of dust and source sediments (nutrients of biogeochemical importance, toxic constituents, and elements useful in dating and provenance identification)
• Vegetation composition/structure in deposition zone (trapping efficiency)

A web input interface will be easier to implement for some of these data than others. A standard set of units will need to be chosen or the web interface must be able to read the units and convert to a standard within the overall database. Once complete, the database will be searchable by a range of key variables such as date of event, location, geographic region, likely source landform, etc. Each source type will be classified (e.g., playa, loess deposit, outwash plain, etc.) to help identify the role of geomorphology and soils in dust emissions at the regional scale. The type of weather event generating the dust storm will also be identified to aid in the prediction of dust events based on their atmospheric drivers.

Each contributor or retrieving researcher would have a user ID to track usage. Researchers retrieving data would be asked to cite the USGS and the originating researcher who gathered the data. The user ID would provide a method for tracking this activity as well.