Spatial-Temporal Analysis of Wildfires in Washington State

Authors: David Szpakowski*, Texas State University - San Marcos
Topics: Geographic Information Science and Systems, Hazards, Risks, and Disasters, Earth Science
Keywords: Spatial-Temporal Analysis, Fire, Landcover/Land use
Session Type: Poster
Day: 4/4/2019
Start / End Time: 9:55 AM / 11:35 AM
Room: Lincoln 2, Marriott, Exhibition Level
Presentation File: No File Uploaded

Wildfire occurrences can drastically alter landcover, damage property and infrastructure, and cause deadly harm to people caught in the fire’s path. The state of Washington has experienced many fires on federal lands over the past few decades. This research aims to characterize the patterns of these fire events over space and time. Fire data for federal lands belonging to the NPS, USFS, BLM, BIA, and FWS and located in the state of Washington is used. Space Time Pattern Mining is implemented in ArcMap to identify areas with a higher proportion of fire events over time. Only wildfires which affected 4.05 or more hectares are considered in the analysis, with year of occurrence used as the temporal component. A Space-time Cube (STC) was generated with a bin size based on a distance interval of 5km and a temporal interval of 1 year. The emerging hot spot analysis tool was then used to identify clusters and trends in fire occurrence. Getis-Ord G* statistic is performed on each STC bin resulting in a p-value, z-score, and hotspot bin classification for that bin. A trend analysis is then used to identify upward and downward trends in occurrence of hotspots over time with the Mann-Kendall test. NLCD data was then examined to determine which landcovers are experiencing increasing trends in hotspots of fire occurrences. In the areas identified as exhibiting upward trends the shrub/scrub (40%), evergreen forest (33%) and cultivated crops (12%) were the dominate landcovers.

Abstract Information

This abstract is already part of a session. View the session here.

To access contact information login