Identifying building locations in the wildland–urban interface before and after fires with convolutional neural networks
Neda K. Kasraee A * , Todd J. Hawbaker
B and Volker C. Radeloff A
A SILVIS Lab, Department of Forest and Wildlife Ecology, University of Wisconsin-Madison, 1630 Linden Drive, Madison, WI 53706, USA.
B US Geological Survey, Geosciences and Environmental Change Science Center, Denver, CO 80225, USA.
Abstract
Wildland–urban interface (WUI) maps identify areas with wildfire risk, but they are often outdated owing to the lack of building data. Convolutional neural networks (CNNs) can extract building locations from remote sensing data, but their accuracy in WUI areas is unknown. Additionally, CNNs are computationally intensive and technically complex, making them challenging for end-users, such as those who use or create WUI maps, to apply.
We identified buildings pre- and post-wildfire and estimated building destruction for three California wildfires: Camp, Tubbs and Woolsey.
We evaluated a CNN-based building dataset and a CNN model from a separate commercial vendor to detect buildings from high-resolution imagery. This dataset and model represent to end-users the state of the art of what is readily available for potential WUI mapping.
We found moderate accuracies for the building dataset and the CNN model and a severe underestimation of buildings and their destruction rates where trees occluded buildings. The CNN model performed best post-fire with accuracies ≥73%.
Existing CNNs may be used with moderate accuracy for identifying individual buildings post-fire and mapping the extent of the WUI. The implications are, however, that CNNs are too inaccurate for post-fire damage assessments or building counts in the WUI.
Keywords: aerial photography, building detection, housing growth, machine learning, urbanisation, wildfire destruction, wildfire hazard, wildland fire.
References
Agee JK, Skinner CN (2005) Basic principles of forest fuel reduction treatments. Forest Ecology and Management 211, 83-96.
| Crossref | Google Scholar |
Argañaraz JP, Radeloff VC, Bar-Massada A, Gavier-Pizarro GI, Scavuzzo CM, Bellis LM (2017) Assessing wildfire exposure in the wildland-urban interface area of the mountains of central Argentina. Journal of Environmental Management 196, 499-510.
| Crossref | Google Scholar |
Bar-Massada A (2021) A comparative analysis of two major approaches for mapping the wildland-urban interface: a case study in California. Land 10, 679.
| Crossref | Google Scholar |
Bar-Massada A, Radeloff VC, Stewart SI (2014) Biotic and abiotic effects of human settlements in the wildland–urban interface. BioScience 64, 429-437.
| Crossref | Google Scholar |
Bittner K, Adam F, Cui S, Körner M, Reinartz P (2018) Building Footprint Extraction From VHR Remote Sensing Images Combined With Normalized DSMs Using Fused Fully Convolutional Networks. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing 11, 2615-2629.
| Crossref | Google Scholar |
Bowman DMJS, Balch JK, Artaxo P, Bond WJ, Carlson JM, Cochrane MA, D’Antonio CM, DeFries RS, Doyle JC, Harrison SP, Johnston FH, Keeley JE, Krawchuk MA, Kull CA, Marston JB, Moritz MA, Prentice IC, Roos CI, Scott AC, Swetnam TW, van der Werf GR, Pyne SJ (2009) Fire in the Earth System. Science 324, 481-484.
| Crossref | Google Scholar |
Buechi H, Weber P, Heard S, Cameron D, Plantinga AJ (2021) Long-Term Trends in Wildfire Damages in California. International Journal of Wildland Fire 30, 757-762.
| Crossref | Google Scholar |
CAL FIRE (2020) 2020 Incident Archive. Available at https://www.fire.ca.gov/incidents/2020/ [verified 7 November 2022]
CAL FIRE (2021) Top 20 Most Destructive California Wildfires. Available at https://www.fire.ca.gov/media/t1rdhizr/top20_destruction.pdf [verified 25 April 2022]
CAL FIRE FRAP (2020) Fire Perimeters. Available at https://frap.fire.ca.gov/frap-projects/fire-perimeters/ [verified 18 April 2022]
Carlson AR, Helmers DP, Hawbaker TJ, Mockrin MH, Radeloff VC (2022) The wildland–urban interface in the United States based on 125 million building locations. Ecological Applications 32, e2597.
| Crossref | Google Scholar |
Chuvieco E, Martínez S, Román MV, Hantson S, Pettinari ML (2014) Integration of ecological and socio-economic factors to assess global vulnerability to wildfire. Global Ecology and Biogeography 23, 245-258.
| Crossref | Google Scholar |
Cohen JD (2000) Preventing disaster: home ignitability in the wildland-urban interface. Journal of Forestry 98, 15-21.
| Crossref | Google Scholar |
Cruz MG, Sullivan AL, Gould JS, Sims NC, Bannister AJ, Hollis JJ, Hurley RJ (2012) Anatomy of a catastrophic wildfire: The Black Saturday Kilmore East fire in Victoria, Australia. Forest Ecology and Management 284, 269-285.
| Crossref | Google Scholar |
Dong Z, Wang G, Amankwah SOY, Wei X, Hu Y, Feng A (2021) Monitoring the summer flooding in the Poyang Lake area of China in 2020 based on Sentinel-1 data and multiple convolutional neural networks. International Journal of Applied Earth Observation and Geoinformation 102, 102400.
| Crossref | Google Scholar |
U.S. Geological Survey - Earth Resources Observation and Science Center (2017) National Agriculture Imagery Program. Available at https://www.usgs.gov/centers/eros/science/usgs-eros-archive-aerial-photography-national-agriculture-imagery-program-naip [verified 9 August 2022]
Esri (2021) Building Footprint Extraction – USA – Overview. Available at https://www.arcgis.com/home/item.html?id=a6857359a1cd44839781a4f113cd5934/ [verified 25 April 2022]
Fibæk CS, Keßler C, Arsanjani JJ (2021) A multi-sensor approach for characterising human-made structures by estimating area, volume and population based on sentinel data and deep learning. International Journal of Applied Earth Observation and Geoinformation 105, 102628.
| Crossref | Google Scholar |
Gamal A, Wibisono A, Wicaksono SB, Abyan MA, Hamid N, Wisesa HA, Jatmiko W, Ardhianto R (2020) Automatic LIDAR building segmentation based on DGCNN and Euclidean clustering. Journal of Big Data 7, 102.
| Crossref | Google Scholar |
Godoy MM, Martinuzzi S, Kramer HA, Defossé GE, Argañaraz J, Radeloff VC (2019) Rapid WUI growth in a natural amenity-rich region in central-western Patagonia, Argentina. International Journal of Wildland Fire 28, 473-484.
| Crossref | Google Scholar |
Griffiths D, Boehm J (2019) Improving public data for building segmentation from convolutional neural networks (CNNs) for fused airborne lidar and image data using active contours. ISPRS Journal of Photogrammetry and Remote Sensing 154, 70-83.
| Crossref | Google Scholar |
Hammer RB, Stewart SI, Radeloff VC (2009) Demographic trends, the wildland–urban interface, and wildfire management. Society & Natural Resources 22, 777-782.
| Crossref | Google Scholar |
Kaim D, Radeloff VC, Szwagrzyk M, Dobosz M, Ostafin K (2018) Long-term changes of the wildland–urban interface in the Polish Carpathians. ISPRS International Journal of Geo-Information 7, 137.
| Crossref | Google Scholar |
Kasraee NK, Hawbaker TJ, Radeloff VC (2023) Building locations identified before and after the Camp, Tubbs, and Woolsey wildfires. U.S. Geological Data Release. 10.5066/P9VWV2IO
Kattenborn T, Leitloff J, Schiefer F, Hinz S (2021) Review on Convolutional Neural Networks (CNN) in vegetation remote sensing. ISPRS Journal of Photogrammetry and Remote Sensing 173, 24-49.
| Crossref | Google Scholar |
Keeley JE, Fotheringham CJ (2003) Impact of past, present, and future fire regimes on North American Mediterranean shrublands. In ‘Fire and Climatic Change in Temperate Ecosystems of the Western Americas’. (Eds TT Veblen, WL Baker, G Montenegro, TW Swetnam) pp. 218–262. (Springer Science Publishing: USA)
Keeley JE, Syphard AD (2019) Twenty-first century California, USA, wildfires: fuel-dominated vs. wind-dominated fires. Fire Ecology 15, 24.
| Crossref | Google Scholar |
Keeley JE, Fotheringham CJ, Morais M (1999) Reexamining fire suppression impacts on brushland fire regimes. Science 284, 1829-1832.
| Crossref | Google Scholar |
Khoshboresh-Masouleh M, Alidoost F, Arefi H (2020) Multiscale building segmentation based on deep learning for remote sensing RGB images from different sensors. Journal of Applied Remote Sensing 14, 034503.
| Crossref | Google Scholar |
Kramer HA, Mockrin MH, Alexandre PM, Radeloff VC (2019) High wildfire damage in interface communities in California. International Journal of Wildland Fire 28, 641-650.
| Crossref | Google Scholar |
Li J, Zhang B, Huang X (2022) A hierarchical category structure based convolutional recurrent neural network (HCS-ConvRNN) for Land-Cover classification using dense MODIS Time-Series data. International Journal of Applied Earth Observation and Geoinformation 108, 102744.
| Crossref | Google Scholar |
Marlon JR, Bartlein PJ, Walsh MK, Harrison SP, Brown KJ, Edwards ME, Higuera PE, Power MJ, Anderson RS, Briles C, Brunelle A, Carcaillet C, Daniels M, Hu FS, Lavoie M, Long C, Minckley T, Richard PJH, Scott AC, Shafer DS, Tinner W, Umbanhowar CE, Whitlock C (2009) Wildfire responses to abrupt climate change in North America. Proceedings of the National Academy of Sciences 106, 2519-2524.
| Crossref | Google Scholar |
Mhawej M, Faour G, Adjizian-Gerard J (2017) Establishing the wildland-urban interface building risk index (WUIBRI): The case study of Beit-Meri. Urban Forestry & Urban Greening 24, 175-183.
| Crossref | Google Scholar |
Microsoft (2018) US Building Footprints. Available at https://github.com/microsoft/USBuildingFootprints/ [verified 21 April 2022]
Microsoft (2022) Global ML Building Footprints. Available at https://github.com/microsoft/GlobalMLBuildingFootprints/ [verified 10 December 2022]
Modugno S, Balzter H, Cole B, Borrelli P (2016) Mapping regional patterns of large forest fires in wildland–urban interface areas in Europe. Journal of Environmental Management 172, 112-126.
| Crossref | Google Scholar |
Postadjian T, Le Bris A, Sahbi H, Mallet C (2017) Investigating the potential of deep neural networks for large-scale classification of very high resolution satellite images. ISPRS Annals of Photogrammetry, Remote Sensing and Spatial Information Sciences IV-1/W1, 183-190.
| Crossref | Google Scholar |
Radeloff VC, Hammer RB, Stewart SI (2005) Rural and suburban sprawl in the U.S. Midwest from 1940 to 2000 and its relation to forest fragmentation. Conservation Biology 19, 793-805.
| Crossref | Google Scholar |
Radeloff VC, Helmers DP, Kramer HA, Mockrin MH, Alexandre PM, Bar-Massada A, Butsic V, Hawbaker TJ, Martinuzzi S, Syphard AD, Stewart SI (2018) Rapid growth of the US wildland–urban interface raises wildfire risk. Proceedings of the National Academy of Sciences 115, 3314-3319.
| Crossref | Google Scholar |
Rousset F, Ferdy J-B (2014) Testing environmental and genetic effects in the presence of spatial autocorrelation. Ecography 37, 781-790.
| Crossref | Google Scholar |
RStudio Team (2020) ‘RStudio: Integrated Development for R.’ (RStudio, PBC: Boston, MA) Available at http://www.rstudio.com/
Smith H, de Beurs KM, Neeson TM (2022) Evaluation of low-resolution remotely sensed datasets for burned area assessment within the wildland–urban interface. Remote Sensing Applications: Society and Environment 26, 100752.
| Crossref | Google Scholar |
Song J, Gao S, Zhu Y, Ma C (2019) A survey of remote sensing image classification based on CNNs. Big Earth Data 3, 232-254.
| Crossref | Google Scholar |
Stewart SI, Radeloff VC, Hammer RB, Hawbaker TJ (2007) Defining the wildland–urban interface. Journal of Forestry 105, 201-207.
| Crossref | Google Scholar |
Syphard AD, Radeloff VC, Keeley JE, Hawbaker TJ, Clayton MK, Stewart SI, Hammer RB (2007) Human Influence on California Fire Regimes. Ecological Applications 17, 1388-1402.
| Crossref | Google Scholar |
Syphard AD, Brennan TJ, Keeley JE (2014) The role of defensible space for residential structure protection during wildfires. International Journal of Wildland Fire 23, 1165-1175.
| Crossref | Google Scholar |
USDA, USDI (2001) Urban wildland interface communities within the vicinity of federal lands that are at high risk from wildfire. Federal Register 66, 751-777 https://www.federalregister.gov/d/01-52.
| Google Scholar |
Wambugu N, Chen Y, Xiao Z, Wei M, Aminu Bello S, Marcato Junior J, Li J (2021) A hybrid deep convolutional neural network for accurate land cover classification. International Journal of Applied Earth Observation and Geoinformation 103, 102515.
| Crossref | Google Scholar |
Xu X, Lu Y, Zou B (2020) Building Extraction From PolSAR Image Based on Deep CNN with Polarimetric Features. In ‘2020 21st International Radar Symposium (IRS)’, pp. 117–120, October 2020, Warsaw, Poland. Available at https://ieeexplore.ieee.org/document/9253912
Zheng Z, Zhong Y, Wang J, Ma A, Zhang L (2021) Building damage assessment for rapid disaster response with a deep object-based semantic change detection framework: From natural disasters to man-made disasters. Remote Sensing of Environment 265, 112636.
| Crossref | Google Scholar |
Zhu XX, Tuia D, Mou L, Xia GS, Zhang L, Xu F, Fraundorfer F (2017) Deep learning in remote sensing: A comprehensive review and list of resources. IEEE Geoscience and Remote Sensing Magazine 5, 8-36.
| Crossref | Google Scholar |
