GOFER

GOFER: GOES-Observed Fire Event Representation

The GOFER algorithm uses geostationary satellite observations of active fires from GOES-East and GOES-West to map the hourly progression of large wildfires (over 50,000 acres or 202 sq. km). GOES observes North and South America with a spatial resolution of 2 km at the equator and at a frequency of 10-15 minutes for the full disk view. Along with the fire perimeter, we derive the active fire lines and fire spread rates. We tested the GOFER algorithm on a set of 28 wildfires in California from 2019-2021 and produced three versions of the product: GOFER-Combined, GOFER-East, and GOFER-West. GOFER-Combined uses both GOES-East and GOES-West observations, while GOFER-East and GOFER-West use only GOES-East and only GOES-West observations, respectively. We find that GOFER performs reasonably well compared to final perimeters from California's Fire and Resource Assessment Program (FRAP) and 12-hourly perimeters from the Fire Event Data Suite (FEDS), derived from 375-m active fire observations. See our ESSD paper for more details. The GOFER Product Visualization app on Earth Engine Apps provides an overview of the product, alongside other products and datasets, such as FEDS and FRAP perimeters and 30-m burn severity from Monitoring Trends in Burn Severity (MTBS). The product of 28 large wildfires in California from 2019-2021 is available on Zenodo. The current version of GOFER is version 0.2.

For more information on all of our fire tracking algorithms and datasets, please visit the following webpages: UCI/ESS, NASA.

Code Structure

Code	Description
largeFires_metadata.js	dictionary of key metadata for all fires in EE
fireData.csv	dictionary of key metadata for all fires in R
0a - Calc_staging.js	set temporal and spatial constraints for each fire by manually inspecting GOES active fire pixels and timeseries
0b - Calc_kernelRes.js	calculate the kernel radius for smoothing based on the GOES spatial resolution
0c - Export_fireData.js	export fire metadata and save as fireData.csv
1a - Export_FireConf.js	export GOES fire detection confidence
1b - Export_Parallax.js	export GOES parallax displacement in x and y-directions
2 - Export_ParamSens.js, eePro_ParamSens.R	export optimization metric for confidence threshold and parallax adjustment factor
4 - Export_FireProg.js, eePro_FireProg.R	export fire perimeters
5 - Export_FireProgQA.js	quality control post-processing for fire perimeters
6a - Export_cFireLine.js, eePro_cFireLine.R	export concurrent active fire lines
6b - Export_rFireLine.js, eePro_rFireLine.R	export retrospective active fire lines
6c - Export_FireIg.js	export fire ignitions
7 - Export_FireProgStats.js	export fire spread rate, growth in fire-wide area
8 - table_GOFERstats.R	export GOFER summary stats for each fire
9 - make_GOFERfinal.R	export final GOFER data files, combined for all fires

File Structure

Earth Engine Assets

GOFER/
	GOFERC_fireProg/
	GOFERC_fireIg/
	GOFERC_cfireLine/
	GOFERC_rfireLine/
	GOFERC_fireProgStats/
	GOFERC_scaleVal/
	...
	GOESEast_MaxConf/
	GOESWest_MaxConf/
	GOESEast_Parallax/
	GOESWest_Parallax/

Local

GOES/
	fireData.csv
	ee_fireProg_chunks/
	ee_fireProg_temp/
	ee_rfireLine/
	ee_cfireLine_chunks/
	ee_fireIg/
	ee_fireStats/
	ee_paramSens_chunks/
	gofer_combined/
		fireProg/
		fireIg/
		cfireLine/
		rfireLine/
		paramSens/
		fireStats/
		summary/
	gofer_east/
		...
	gofer_west/
		...

Data Structure

GOFER/
	fireData.csv
	GOFER-Combined/
		GOFERC_fireProg.shp
		GOFERC_cfireLine.shp
		GOFERC_rfireLine.shp
		GOFERC_fireIg.shp
		GOFERC_summary.csv
	GOFER-East/
		...
	GOFER-West/
		...

Earth Engine Repository

https://code.earthengine.google.com/?accept_repo=users/embrslab/GOFER

Product Description

Name	Short Name	Units
Global variables
Fire name	fname
Fire year	fyear
End-of-hour variables (t=1,2,3…)
Hours after ignition, end of hour	timestep	hours
UTC time	tUTC
Local time, with daylight savings	tLocal
Local time, without daylight savings	tLocalGMT
Area within fire perimeter	farea	km2
Area within fire perimeter, as a percentage of the final area	fareaPer	%
Active fire line length (concurrent)	cflinelen	km
Active fire line length (retrospective)	rflinelen	km
Length of the perimeter	fperim	km
State of the fire	fstate	0 = dormant, 1 = active
Half-hour variables (t=0.5,1.5,2.5…)
Hours after ignition, half hour	timestep_hh	hours
Growth in area	dfarea	km2
Fire spread rate (MAE)	maefspread	km/h
Fire spread rate (AWE)	awefspread	km/h

• For cflinelen, e.g. cflinelen5 = the portion of the perimeter that intersects with concurrent active fires with fire detection confidence > 0.05; cflinelen5 should be used as the default fire line as it incorporates most active fire pixels along the perimeter and most closely matches FEDS; for cfline, fconf similarly denotes the fire detection confidence of its definition
• For fstate, both rfline and cfline have fstate columns, e.g. rfline_fstate; if the fire line is dormant at the timestep (fstate = 0), then that timestep is filled with the most recent cfline or the most immediate rfline after the timestep; to get the original flinelen, multiply by the corresponding fstate

Versions

• 0.0 - initial version for ESSD Discussions (Liu et al., 2023)
• 0.1 - published version in ESSD (Liu et al., 2024)
• 0.11 - bug fixes: fix mismatched ignition time (GOES_UTC) in EE and fireData.csv for the 2019 Walker Fire, fix scaleVal.csv that previously duplicated the summary CSVs due to an error in the R code
• 0.12 - bug fixes: fix rfireline.shp that previously duplicated the fireProg.shp due to an error in the R code, fix headers of scaleVal.csv
• 0.2 (current) - several minor updates including data pipeline streamlining: revise fire confidence for saturated fire pixels from 0.9 to 1.0, fix rfline in instances for multipolygons where a dormant segment of rfline persists too far in advance, use ICS-209 IDs to crosswalk GOFER, FEDS, FRAP, and MTBS, add EE script (Export_fireData.js) to download fireData information directly from EE to avoid future mismatches in metadata, fix instances of artifacts in GOFER-East and GOFER-West caused clipped parallax terrain correction inputs, revise colnames in scaleVal.csv to be consistent with other files; EE Apps now calls the Zenodo version of GOFER
• Future updates: CA fires to be added - 2022 (Mosquito, McKinney), 2023 (Smith River Complex, York, SRF Lightning Complex), 2024 (Park, Borel, Bridge); manuscript on ICS-209 and crosswalking to geospatial datasets is in progress

Datasets

Liu, T., J.T. Randerson, Y. Chen, D.C. Morton, E.B. Wiggins, P. Smyth, E. Foufoula-Georgiou, R. Nadler, and O. Nevo (2024). GOES-Observed Fire Event Representation (GOFER) product for 28 California wildfires from 2019-2021. https://doi.org/10.5281/zenodo.8327264

Publications

Liu, T., J.T. Randerson, Y. Chen, D.C. Morton, E.B. Wiggins, P. Smyth, E. Foufoula-Georgiou, R. Nadler, and O. Nevo (2024). Systematically tracking the hourly progression of large wildfires using GOES satellite observations. Earth Sys. Sci. Data, 16, 1395-1424. https://doi.org/10.5194/essd-16-1395-2024