Wildland Fire Data: Issues and Proposed Solutions
Ball, Kendall Carothers
MetadataShow full item record
Wildfires occur all over the world. Many countries collect data on such fires. In an effort prevent and study wild fires, the sharing of this information is imperative. Information can best be shared through a wildland fire database. In order to create a wildland fire database in the United States, data integrity issues must be dealt with first. Data integrity issues became apparent during a wildland fire data warehousing project conducted at Texas A&M University for the National Association of State Foresters (NASF) from 2003-2009. Two main issues emerged from this project. One was the lack of consistency in fire cause and the other was the location information provided was not accurate at the county level. In an effort to propose a solution to the lack of consistency in wildland fire cause, raw data from the NASF project was analyzed as to how to modify the nine statistical causes currently used by the United States Forest Service (USFS) to provide a more useful representation of the current state of wildland fires. In addition, it was hypothesized that location information in the United States Public Land Survey format was on average more accurate than other submitted location data (e.g. latitude/longitude). It was found that the cause information could benefit by sub-dividing the current USFS especially for the causes of debris burning and miscellaneous in order to help determine the actually cause of a wildland fire. Using a ?^2 statistical test it was conclusively determined that USPLS information was the most accurate location information submitted during the NASF project. As a result of these two analyses, a new proof of concept data submittal system was also developed. The development of this system was based on the system that was utilized during the NASF project. However the new system incorporated autonomous data integrity checking at the front end of data entry instead of after the data had been submitted. The new submittal system was costly in terms of data processing time due to the lack of consistency in the data being submitted. The overall issue in wildland fire data is that there is a lack of consistent reporting methodologies and the data that is being submitted.