A SPATIAL-TEMPORAL POINT PROCESS MODEL FOR ESTIMATING PROBABILITY OF WILDFIRES IN LOS ANGELES COUNTY

Loading...
Thumbnail Image
Authors
Yi, Wook
Subjects
point process models
wildfire probabilities
Los Angeles County Fire Department
random forests
generalized additive models.
Advisors
Koyak, Robert A.
Date of Issue
2022-03
Date
Publisher
Monterey, CA; Naval Postgraduate School
Language
Abstract
In Los Angeles County, wildfires are among the most catastrophic environmental events caused by regional characteristics and climate change. In this study, we develop a point process model to estimate the probability of wildfires based on historical weather data and past wildfires data from Los Angeles County from 2004 to 2018. First, we partition Los Angeles County into small rectangular regions, called voxels, with daily temporal resolution. Then, we use random forests and generalized additive models to obtain estimated probabilities on a training data set. In addition to daily weather and fuel-condition measurements, our models incorporate seasonal and geographical effects. Because measurements on weather and fuel conditions are available only from a fixed set of remote automated weather stations, their data must be averaged to relate them to the voxel level, and the way this is done is a factor in modeling. Through the developed model, it is possible to obtain localized, estimated probabilities of wildfires. Ultimately, this tool can aid Los Angeles County Fire Department in improving its capability and effectiveness.
Type
Thesis
Description
Series/Report No
Department
Operations Research (OR)
Organization
Identifiers
NPS Report Number
Sponsors
Funder
Format
Citation
Distribution Statement
Approved for public release. Distribution is unlimited.
Rights
Copyright is reserved by the copyright owner.
Collections