Pasar al contenido principal

MAPPING OF FOREST HEALTH, SPECIES AND FOREST FIRE RISKS USING NOVEL ICT DATA AND APPROACHES

PCI2020-120705-2

•
Nombre agencia financiadora Agencia Estatal de Investigación
Acrónimo agencia financiadora AEI
Programa Programa Estatal de I+D+i Orientado a los Retos de la Sociedad
Convocatoria Proyectos Programación Conjunta Internacional
Año convocatoria 2020
Plan Estatal 2017-2020 Plan Estatal de Investigación Científica y Técnica y de Innovación
Centro beneficiario UNIVERSIDAD DE VIGO

Resultados relacionados

Resultados totales (Incluyendo duplicados): 8
Encontrada(s) 1 página(s)

BULERIA. Repositorio Institucional de la Universidad de León
Publicación Artículo científico (JournalArticle)

Drivers and trends in the size and severity of forest fires endangering WUI areas: a regional case study

  • Rodríguez Jiménez, Fernando
  • Fernandes, Paulo Alexandre Martins
  • Fernández Guisuraga, José Manuel
  • Álvarez Bermúdez, Xana 1986-
  • Lorenzo Cimadevilla, Enrique
[EN] This study explored, for the first time, the drivers shaping large fire size and high severity of forest fires classified as level-2 in Spain, which pose a great danger to the wildland–urban interface. Specifically, we examined how bottom-up (fuel type and topography) and top-down (fire weather) controls shaped level-2 fire behavior through a Random Forest classifier at the regional scale in Galicia (NW Spain). We selected for this purpose 93 level-2 forest fires. The accuracy of the RF fire size and severity classifications was remarkably high (>80%). Fire weather overwhelmed bottom-up controls in controlling the fire size of level-2 forest fires. The likelihood of large level-2 forest fires increased sharply with the fire weather index, but plateaued at values above 40. Fire size strongly responded to minimum relative humidity at values below 30%. The most important variables explaining fire severity in level-2 forest fires were the same as in the fire size, as well as the pre-fire shrubland fraction. The high-fire-severity likelihood of level-2 forest fires increased exponentially for shrubland fractions in the landscape above 50%. Our results suggest that level-2 forest fires will pose an increasing danger to people and their property under predicted scenarios of extreme weather conditions, SI, Project 4Map4Health is selected in the call ERA-Net CHIST-ERA (2019) and funded by the State Research Agency of Spain (reference PCI2020-120705-2/AEI/10.13039/501100011033). PMF and JMFG were supported by National Funds from FCT—Portuguese Foundation for Science and Technology, under the project UIDB/04033/2020. JMFG was supported by a postdoctoral fellowship granted by the Ramón Areces Foundation
Investigo. Repositorio Institucional de la Universidade de Vigo
Publicación Artículo científico (JournalArticle)

Modelling of live fuel moisture content in different vegetation scenarios during dry periods using meteorological data and spectral indices

  • Rodriguez Jimenez, Fernando
  • Lorenzo Cimadevila, Henrique Remixio
  • Novo Gómez, Ana
  • Acuña Alonso, Carolina
  • Álvarez Bermúdez, Xana
Wildfires are of increasing concern to society, the scientific community and managers. Each year the forest fire season is becoming longer or deseasonalised. One of the main factors in fire behaviour is the live fuel moisture content. Climatological variables such as air humidity, temperature and precipitation directly affect the fuel that is likely to burn. Soil moisture also contributes to fire spread and severity. For these reasons, an analysis of fuel moisture is carried out at 3 sampling sites in the interior of Galicia (Spain), through multiple linear regression models. The 3 plots represent different characteristics for shrub growth: one plot is composed of pine trees, another plot contains a high density of undergrowth and the last plot was burned in 2020 and is regenerating undergrowth. A specific model was generated for each plot from field samples of live fuel and soil field samples collected from September to May, meteorological variables obtained from the stations and Sentinel-2 satellite spectral indices for the field sampling days.The best resulting spectral index for plots without trees was the NDI45, but with a low correlation (R2 < 0.2), while in the plot with pine trees it was the SAVI index (R2 = 0.85). On the other hand, specific model equations were estimated combining these spectral indices with the soil moisture samples and the average temperature of the last 7 days. In all cases the result was considerably improved (R2 between 0.72 and 0.91). These models can be helpful in estimating the probability of fire danger, outside the summer season, with long periods of drought, enabling decision making by managers in each region., Agencia Estatal de Investigación | Ref. PCI2020-120705-2, Universidade de Vigo/CISUG
Investigo. Repositorio Institucional de la Universidade de Vigo
Publicación Artículo científico (JournalArticle)

Heuristic generation of multispectral labeled point cloud datasets for deep learning models

  • Comesaña Cebral, Lino Jose
  • Martínez Sánchez, Joaquín
  • Rúa Fernández, Erik
  • Arias Sánchez, Pedro
Deep Learning (DL) models need big enough datasets for training, especially those that deal with point clouds. Artificial generation of these datasets can complement the real ones by improving the learning rate of DL architectures. Also, Light Detection and Ranging (LiDAR) scanners can be studied by comparing its performing with artificial point clouds. A methodology for simulate LiDAR-based artificial point clouds is presented in this work in order to get train datasets already labelled for DL models. In addition to the geometry design, a spectral simulation will be also performed so that all points in each cloud will have its 3 dimensional coordinates (x, y, z), a label designing which category it belongs to (vegetation, traffic sign, road pavement, …) and an intensity estimator based on physical properties as reflectance., Agencia Estatal de Investigación | Ref. PCI2020-120705-2
Investigo. Repositorio Institucional de la Universidade de Vigo
Publicación Artículo científico (JournalArticle)

Transport infrastructure management based on LiDAR synthetic data: a deep learning approach with a ROADSENSE simulator

  • Comesaña Cebral, Lino Jose
  • Martínez Sánchez, Joaquín
  • Núñez Seoane, Antón
  • Arias Sánchez, Pedro
In the realm of transportation system management, various remote sensing techniques have proven instrumental in enhancing safety, mobility, and overall resilience. Among these techniques, Light Detection and Ranging (LiDAR) has emerged as a prevalent method for object detection, facilitating the comprehensive monitoring of environmental and infrastructure assets in transportation environments. Currently, the application of Artificial Intelligence (AI)-based methods, particularly in the domain of semantic segmentation of 3D LiDAR point clouds by Deep Learning (DL) models, is a powerful method for supporting the management of both infrastructure and vegetation in road environments. In this context, there is a lack of open labeled datasets that are suitable for training Deep Neural Networks (DNNs) in transportation scenarios, so, to fill this gap, we introduce ROADSENSE (Road and Scenic Environment Simulation), an open-access 3D scene simulator that generates synthetic datasets with labeled point clouds. We assess its functionality by adapting and training a state-of-the-art DL-based semantic classifier, PointNet++, with synthetic data generated by both ROADSENSE and the well-known HELIOS++ (HEildelberg LiDAR Operations Simulator). To evaluate the resulting trained models, we apply both DNNs on real point clouds and demonstrate their effectiveness in both roadway and forest environments. While the differences are minor, the best mean intersection over union (MIoU) values for highway and national roads are over 77%, which are obtained with the DNN trained on HELIOS++ point clouds, and the best classification performance in forested areas is over 92%, which is obtained with the model trained on ROADSENSE point clouds. This work contributes information on a valuable tool for advancing DL applications in transportation scenarios, offering insights and solutions for improved road and roadside management., Agencia Estatal de Investigación | Ref. PID2022-140662OB-I00, Agencia Estatal de Investigación | Ref. PCI2020-120705-2-I00
Investigo. Repositorio Institucional de la Universidade de Vigo
Publicación Artículo científico (JournalArticle)

Wildfire response of forest species from multispectral LiDAR data. A deep learning approach with synthetic data

  • Comesaña Cebral, Lino Jose
  • Martínez Sánchez, Joaquín
  • Suárez Fernández, Gabriel Eduardo
  • Arias Sánchez, Pedro
Forests play a crucial role as the lungs and life-support system of our planet, harbouring 80% of the Earth's biodiversity. However, we are witnessing an average loss of 480 ha of forest every hour because of destructive wildfires spreading across the globe. To effectively mitigate the threat of wildfires, it is crucial to devise precise and dependable approaches for forecasting fire dynamics and formulating efficient fire management strategies, such as the utilisation of fuel models The objective of this study was to enhance forest fuel classification that considers only structural information, such as the Prometheus model, by integrating data on the fire responses of various tree species and other vegetation elements, such as ground litter and shrubs. This distinction can be achieved using multispectral (MS) Light Detection and Ranging (LiDAR) data in mixed forests. The methodology involves a novel approach in semantic classifications of forests by generating synthetic data with semantic labels regarding fire responses and reflectance information at different spectral bands, as a real MS scanner device would detect. Forests, which are highly intricate environments, present challenges in accurately classifying point clouds. To address this complexity, a deep learning (DL) model for semantic classification was trained on synthetic point clouds in different formats to achieve the best performance when leveraging MS data Forest plots in the study region were scanned using different Terrestrial Laser Scanning sensors at wavelengths of 905 and 1550 nm. Subsequently, an interpolation process was applied to generate the MS point clouds of each plot, and the trained DL model was applied to classify them. These classifications surpassed the average thresholds of 90% and 75% for accuracy and intersection over union, respectively, resulting in a more precise categorisation of fuel models based on the distinct responses of forest elements to fire. The results of this study reveal the potential of MS LiDAR data and DL classification models for improving fuel model retrieval in forest ecosystems and enhancing wildfire management efforts, Ministerio de Universidades | Ref. FPU16/00855, Agencia Estatal de Investigación | Ref. PCI2020-120705-2, Universidade de Vigo/CISUG
Investigo. Repositorio Institucional de la Universidade de Vigo
Publicación Artículo científico (JournalArticle)

PLS-PM analysis of forest fires using remote sensing tools. The case of Xurés in the Transboundary Biosphere Reserve

  • Rodriguez Jimenez, Fernando
  • Lorenzo Cimadevila, Henrique Remixio
  • Acuña Alonso, Carolina
  • Álvarez Bermúdez, Xana
Forest fires have environmental, social and economic impacts in many areas. Various factors related to territory directly influence both the number and the surface area of each fire. The link between different variables (climate, social and environmental) in the risk of fire and in the characteristics of fires is studied here through Partial Least Squares - Path Models. In addition, images from the Sentinel-2 sensor and geographic information systems are used to create a cartographic base of fires in the Transboundary Biosphere Reserve of Galicia and the Site of Community Importance of Xurés (Galicia) between 2015 and 2020. In all, seven variables are analyzed in this study area using the partial least squares-path modeling method: climate, topography, land use, type of environmental protection, the anthropogenic factor, fire defense, and fire data (severity and area). The parameters for each variable are used to obtain weights and thus determine the importance of each one. The areas where the problem of forest fires is greatest are those with the greatest environmental protection. Up to 31% of the surface area of the Natura 2000 Network was burned in the 6-year study period. Topography and land use are also shown to be relevant factors in the effects of forest fires in this territory. By contrast, higher population density and the development of infrastructures such as roads and water tanks mitigate the impact of fires. The problem of forest fires encompasses many variables that need to be studied. By contextualizing each study area as far as possible, specific measures to prevent and reduce damage can be drawn up., Agencia Estatal de Investigación | Ref. PCI2020–120705-2, Financiado para publicación en acceso aberto: Universidade de Vigo/CISUG
Investigo. Repositorio Institucional de la Universidade de Vigo
Publicación Artículo científico (JournalArticle)

Spatial analysis of tree species before forest fires

  • Novo Gómez, Ana
  • González Jorge, Higinio
  • Martínez Sánchez, Joaquín
  • Balado Frías, Jesús
Spain is included in the top five European countries with the highest number of wildfires. The occurrence and magnitude of forest fires involves aspects of a very diverse nature, from those of a socio-economic, climatic, or physiographic nature, to those concerning fuel or the availability and quantity of resources and means of extinction. The distribution of wildfires in Galicia is not random and that fire occurrence may depend on ownership conflicts also a spatial dependence between productive or non-productive area exists. Satellite data play a major role in providing knowledge about fires by delivering rapid information to map fire-damaged areas precisely and promptly. In addition, the availability of large-scale data and the high temporal resolution offered by the Sentinel-2 satellite enables to classify and determine the land cover changes with high accuracy. This study describes a methodology to detect burned areas and analyse the Land Cover and Land Use (LCLU) classes present in these areas during the period of 5 years (2016–2021) by Sentinel-2 images. The training areas were obtained by photointerpretation and the image classification was performed using the Random Forest algorithm which shows an overall accuracy range between 80–85%. The methodology concluded that Lobios and Muiños were the most affected municipalities by wildfires. Additionally, the spatial analysis determined that the Deciduous Forest mainly composed by Quercus sp. were the most affected in 2017 followed by Coniferous Forest mainly composed by Pinus sp.in 2016. Although, Scrub and Rock are the classes more affected for wildfire during 2016–2020 period., Universidade de Vigo, Agencia Estatal de Investigación | Ref. PCI2020-120705-2, Xunta de Galicia | Ref. ED481B-2019-061, Xunta de Galicia | Ref. ED431C 2020/01
Investigo. Repositorio Institucional de la Universidade de Vigo
Publicación Artículo científico (JournalArticle)

Drivers and trends in the size and severity of forest fires endangering WUI areas: a regional case study

  • Rodriguez Jimenez, Fernando
  • Fernandes, Paulo M.
  • Fernández Guisuraga, José Manuel
  • Álvarez Bermúdez, Xana
  • Lorenzo Cimadevila, Henrique Remixio
This study explored, for the first time, the drivers shaping large fire size and high severity of forest fires classified as level-2 in Spain, which pose a great danger to the wildland–urban interface. Specifically, we examined how bottom-up (fuel type and topography) and top-down (fire weather) controls shaped level-2 fire behavior through a Random Forest classifier at the regional scale in Galicia (NW Spain). We selected for this purpose 93 level-2 forest fires. The accuracy of the RF fire size and severity classifications was remarkably high (>80%). Fire weather overwhelmed bottom-up controls in controlling the fire size of level-2 forest fires. The likelihood of large level-2 forest fires increased sharply with the fire weather index, but plateaued at values above 40. Fire size strongly responded to minimum relative humidity at values below 30%. The most important variables explaining fire severity in level-2 forest fires were the same as in the fire size, as well as the pre-fire shrubland fraction. The high-fire-severity likelihood of level-2 forest fires increased exponentially for shrubland fractions in the landscape above 50%. Our results suggest that level-2 forest fires will pose an increasing danger to people and their property under predicted scenarios of extreme weather conditions., Fundação para a Ciência e Tecnologia | Ref. UIDB/04033/2020, Fundación Ramón Areces | Ref. Postdoctoral fellowship, Agencia Estatal de Investigación | Ref. PCI2020-120705-2