Satellite Images and Spectral Vegetation Indices as Auxiliary Tools

to Monitoring Fuel Availability in Areas Prone to Wildfire:

Study Case in the Northern Region of Portugal

Bárbara Pavani-Biju

1,2,3 a

, José G. Borges

Susete Marques

and Ana C. Teodoro

1,2 d

Department of Geosciences, Environment and Land Planning, University of Porto,

Rua Campo Alegre, 687, 4169-007, Porto, Portugal

Earth Sciences Institute (ICT), Pole of the FCUP, University of Porto, 4169-007 Porto, Portugal

Forest Research Centre, School of Agriculture, University of Lisbon, Tapada da Ajuda, 1349-017 Lisbon, Portugal

Forest Research Centre, Associate Laboratory TERRA, School of Agriculture, University of Lisbon, Tapada da Ajuda,

1349-017 Lisbon, Portugal

Keywords: Remote Sensing, Wildfire, Spectral Vegetation Indices, Sentinel-2, Wildfire Fuels.

Abstract: Remote sensing data has led to the development of spectral indices for monitoring ecosystems, land surface

changes, and water quality. These indices are used in various applications, including agricultural and wildfire

monitoring, to understand vegetation cycles and disturbances. Wildfire research focuses on the effects of

extreme occurrences, and understanding forest ecology after severe events is crucial for evaluating forest

health. Vegetation Indices (VIs) are frequently used in forest and wildfire monitoring studies to account for

plant biophysical, biochemical, and physiological characteristics. Normalized Difference Vegetation Index

(NDVI), Normalized Burn Ratio (NBR), Normalized Difference Infrared Index (NDII), and Plant Senescence

Reflectance Index (PSRI) are indices used to assess vegetation conditions. VIs are valuable resources for

monitoring post-wildfire occurrences, as they measure biophysical changes and provide comprehensive

monitoring of the affected area, playing a crucial role in assessing the health of forests. Pre-wildfire vegetation

conditions monitoring is also important for implementing preventative measures in critical regions to increase

wildfire defense and identifying wildland fuels is crucial for improving fuel management actions. This

research aims to demonstrate the effectiveness of chosen VIs and fuel models as tools to assess pre-fire

conditions, enabling decision-makers to increase wildfire surveillance and landscape resilience in Vale do

Sousa, Portugal's northern area. Despite limitations, this approach is valuable, especially in terms of financial

or logistical constraints. Moreover, combining VIs with fuel hazard models can improve fuel reduction efforts.

1 INTRODUCTION

With the rapid development and accessibility of

remote sensing data, several spectral indices have

been developed to monitor ecosystems, land surface

changes, and water quality (Chughtai et al., 2021; Ma

et al., 2019). Spectral indices have been used in a

variety of applications, including agricultural and

wildfire monitoring, to better understand vegetation

cycles or disturbances (Zeng et al., 2022).

https://orcid.org/0000-0001-7721-008X

https://orcid.org/0000-0002-0608-5784

https://orcid.org/0000-0001-7922-5680

https://orcid.org/0000-0002-8043-6431

A significant percentage of wildfire research

focuses on the effects of extreme occurrences (Dos

Santos et al., 2020). Understanding the forest ecology

after severe events is crucial for evaluating forest

health (Avetisyan et al., 2023). The information

acquired may help stakeholders make key decisions

about post-fire restoration, management, and actions.

Most post wildfire assessments rely on field surveys,

which are expensive, time-consuming, and limited to

specific regions of the affected area (Fernandes et al.,

2006).

Pavani-Biju, B., Borges, J. G., Marques, S. and Teodoro, A. C.

Satellite Images and Spectral Vegetation Indices as Auxiliary Tools to Monitoring Fuel Availability in Areas Prone to Wildﬁre: Study Case in the Northern Region of Portugal.

DOI: 10.5220/0013249400003935

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 11th International Conference on Geographical Information Systems Theory, Applications and Management (GISTAM 2025), pages 51-60

ISBN: 978-989-758-741-2; ISSN: 2184-500X

Vegetation indices (VIs) are frequently used in

forest and wildfire monitoring studies to account for

plant biophysical (detecting vegetation structure and

changes), biochemical (e.g., pigments, water), and

physiological (photosynthetic light use efficiency)

characteristics (Zeng et al., 2022). The Normalized

Difference Vegetation Index (NDVI) is one method

for assessing vegetation conditions (Rouse et al.,

1973). The Normalized Burn Ratio (NBR) and

Differenced Normalized Burn Ratio (dNBR) are used

to evaluate damage to vegetation and regeneration

(Roy et al., 2006). The Plant Senescence Reflectance

Index (PSRI) is used to measure chlorophyll

concentration (Merzlyak et al., 1999).

VIs are a valuable resource for monitoring post-

wildfire occurrences since they can measure

biophysical changes, allowing for comprehensive

monitoring of the affected area (Avetisyan et al.,

2023). Most common VIs products are easily

accessible through frameworks like Copernicus

Browser, which produces NDVI, Moisture Index, and

other metrics from Sentinel-2 data (ESA, 2024).

For example, NDVI can also be calculated from

Landsat imagery using Google Earth Engine

(Gorelick et al., 2017) or obtained as a product

directly from the United States Geological Survey

(U.S. Geological Survey, 2024). Other VIs can be

computed using Geographic Information System

(GIS) tools like QGIS, ArcGIS®, and SNAP (ESA,

2024; Esri & Environmental Systems Research

Institute, 2015; QGIS Association, 2024).

VIs are calculated as a ratio of certain bands,

differences, or derivatives of reflectance from sensor

imagery at specific spectral wavelengths (de Almeida

et al., 2020). Sentinel-2 constellation provides high-

resolution images for monitoring vegetation at visible

and infrared wavelengths, such as NDVI, which is

calculated by combining the Near Infrared (NIR) and

red bands.

With a rising rate of severe weather events

associated with climate change, VIs also play an

important role as reliable tools for assessing the

health of forests (Lee et al., 2024). As an example, the

forest in Portugal's northern region is particularly

vulnerable to severe wildfires, which are becoming

increasingly regular in this region (Marques et al.,

2017; San-Miguel-Ayanz et al., 2017; Teodoro &

Duarte, 2013).

As mentioned before, post-wildfire event

monitoring is crucial for forest management and

decision-making processes. However, studying

vegetation before a fire occurs is just as important as

assessing post-fire conditions, physical features, and

climate data for the affected area (Lee et al., 2024).

When stakeholders monitor pre-wildfire vegetation

conditions, they can implement preventative

measures in critical regions to increase wildfire

defense. VIs are helpful to evaluate the condition of

vegetation in terms of greenness (NDVI), humidity

(Normalized Difference Formulation Index - NDII),

and senescence (PSRI) (Bento-Gonçalves et al.,

2019b; Hardisky et al., 1983; Merzlyak et al., 1999).

Wildfire behavior and impacts have a connection

to the type of vegetation and its attributes; in the

words of Fernandes et al., (2006), describing wildland

fuels is crucial in order to improve fuel management

actions. This means that it is essential to classify the

study area's land use and land cover (LULC) in order

to identify vegetation cover type and structure, as well

as classify them according to their fire hazard

potential. When combined NDVI, PSRI and NDII to

identify fuel availability, with other information, such

as fuel hazard potential, they may assist in identifying

areas where fuel management should be prioritized

(Bento-Gonçalves et al., 2017).

The main objective of the present research is to

demonstrate the effectiveness of the chosen VIs as a

tool to assess pre-fire vegetation conditions. To assess

fuel availability, a vegetation fuel hazard model

developed for Portugal by Fernandes et al., (2006)

was considered. Enabling decision-makers to take

action to increase wildfire surveillance and landscape

resilience in Vale do Sousa, Portugal's northern area,

principally in critical areas such as the Wildland

Urban Interface.

2 MATERIALS AND METHODS

2.1 Study Area

The research area (Figure 1), a forested area with

about 29k ha, Vale do Sousa, which is situated in

northern Portugal, encompasses two Forest

Intervention Zones (ZIFs): Entre-Douro-e-Sousa and

Castelo de Paiva. Extreme fires have taking place in

the Vale do Sousa woodland area over the past few

years (Pavani-Biju et al., 2024). Large wildfires

occurred in 2017, burning more than 5000 hectares

and causing major environmental and financial

impacts (ICNF, 2017).

Vale do Sousa has also a large rural area, with

Eucalyptus (Eucalyptus globulus Labill) as the

dominating tree species, followed by Maritime Pine

(Pinus pinaster A.), small areas of other deciduous

oak such as cork oak (Quercus suber L.), and

pedunculate oak (Quercus robur L.), and riparian

areas.

GISTAM 2025 - 11th International Conference on Geographical Information Systems Theory, Applications and Management

Figure 1: Study area Vale do Sousa.

However, the study area includes small portions of

the NATURA 2000 Network, a network of protected

areas that covers Europe's most valuable and

vulnerable landscapes and spans 27 European Union

Member States and in order to protect this area,

crucial actions need to be taken to mitigate large

wildfires harmful effects (CEE, 1992).

2.2 Data Acquisition and Analysis

For this study, images from Sentinel-2 between 2017

and 2018 were obtained at Copernicus Browser (table

1), based on the closest date the wildfire started, with

less than 5% cloud cover. First, images were pre-

processed from Level-1C (Top of Atmosphere –

TOA) to Level-2A (Bottom of Atmosphere - BOA)

resulting in orthorectified surface reflectance images,

at SNAP using the algorithm SEN2COR (ESA,

2024). After, VIs were produced for September and

October of 2017 in order to evaluate pre-fire

vegetation conditions.

Table 1: Sentinel-2 images collected for 2017 and 2018.

Sentinel Ima

es Date Wildfire started

tember 02

, 2017 Se

tember 02

, 2017

October 12

, 2017 October 15

th,

of 2017

Furthermore, leaf-on multispectral images from

July 2017 and July 2018 were classified using

Machine Learning (ML)

, Support Vector Machine in

ArcGIS Pro® to determine changes between 2017

and 2018. To train the classifier 200 samples were

collected for each class were collected from the

thematic map of Land Use and Occupation (COS) of

Portugal, from 2015 and 2018. The same spatial

information were adapted to establish the accuracy

assessment points.

The COS is the most widely used national

reference cartography for land use issues. It is a

vector product updated every three years that

represents 83 thematic classes with a minimum

cartographic unit of 1 hectare based on the visual

interpretation of orthophoto maps. There are five

temporally consistent editions available (1995, 2007,

2010, 2015, and 2018), with an accuracy higher than

85% (SNIG, 2022).The sampling approach used was

random points. Approximately 500 points were

generated from the image. Further, the confusion

matrix and accuracy assessment were calculated.

Accuracy ranges from 0 to 1, with 1 representing

100% accuracy, and the Kappa statistic indicate the

overall accuracy of the classification. (McCoy &

Johnston, 2001).The landscape classes chosen to

Satellite Images and Spectral Vegetation Indices as Auxiliary Tools to Monitoring Fuel Availability in Areas Prone to Wildﬁre: Study Case

in the Northern Region of Portugal

perform the classification comprised agriculture, bare

land, building area, eucalyptus, maritime pine, mixed

forest (which includes riparian and non-riparian), and

shrubland. Additionally, the classified data from 2017

was utilized to classify fuel types and hazards (Table

2). According to the Portuguese custom fuel

model(Fernandes et al., 2006; Fernandes & Loureiro,

2021), fuel hazard values range from 1 (very low) to

5 (very high). Vegetated areas were classified on a

scale of 2 to 5 because the fuel model runs from 1 to

4. To encompass additional land cover types in the

classification, such as water, barren land, houses, and

agriculture, were assigned 1 because they are not

considered wildfire fuel.

Table 2: Sentinel-2 images collected for 2017 and 2018.

Fuel Types –

LULC classes

Fuel Models Fire Hazard

Building; Water;

Agriculture;

Bare

1 Very Low

Mixed Forest –

Riparian and

Non-Riparian

2 Low

Maritime Pine

Stands

3 Moderate

Shrubland –

Small

Vegetation

4 High

Eucalyptus

Stands

5 Very High

The VIs were estimated using the ArcGIS Pro®

indices tool, using pre-fire satellite images. In this

case, only the largest forest fires that occurred in

2017, designated as Burned Area 1 (BA1), which

occurred in September and Burned Area 2 (BA2),

which occurred in October (Table 1), were considered

in this analysis. Since both events are classified as

extreme wildfires because they covered more than

100 hectares, BA1 has 5,433 hectares, while BA2 has

666,3 hectares.

First, the NDVI was computed (1). Its values vary

from -1 to 1. Values near 0 represent bare, rocks,

sand, and snow. Low positive scores (0.2-0.4) are

associated with shrubs and grassland. Values greater

than 0.4 imply live green vegetation (Rouse et al.,

1973).

𝑁𝐷𝑉𝐼 =

(  )

(  )

(1)

The NDII (2) is sensitive to variations in the water

content of the plant canopy (Hardisky et al., 1983).

The value rises as water content rises. NDII

measurements can range from -1 to 1, with green

vegetation often falling between 0.02 and 0.6.

𝑁𝐷𝐼𝐼 =

(  )

(  )

(2)

The PSRI (3) assesses plant senescence

(Merzlyak et al., 1999). PSRI values, like NDVI,

range from -1 to 1. Although values ranging from -

0.1 to 0.2 suggest healthy vegetation, values greater

than 0.2 imply senescence and values less than -0.1

are associated with other landscape characteristics

such as water and buildings.

𝑃𝑆𝑅𝐼 =

( )

( )

(3)

While higher NDVI values suggest greater

greenness, PSRI indicates significant senescence. To

measure fuel availability, a composite using ArcGIS

Pro® composite band tool, of the three VIs was used.

To use PSRI in conjunction with NDVI and NDII, it

was multiplied by -1. Converting positive senescence

numbers to negative values. The compositions were

then sliced based on wildfire fuel availability, which

ranged from 1 to 5, based on what was reported by

Bento-Gonçalves et al., (2019a). Where 1 represents

very low, 2 represents low, 3 represents moderate, 4

represents high, and 5 represents extremely high

wildfire fuel source.

Vector data containing polygons from burned

regions BA1 and BA2 were obtained from the

Portuguese National Authority for Nature

Conservation (ICNF) and compared to VIs and

wildfire fuels data. This was conducted to see if the

spatial information generated corresponds to the

perimeter of the burned regions in official data. The

classified image from 2017 was then compared to

2018 data to determine the percentage change

following the occurrence of these extreme events.

3 RESULTS AND DISCUSSION

VIs were estimated for the two greatest burned areas

(BA1 and BA2) in Vale do Sousa before wildfire

occurrence. In terms of vegetation conditions for

BA1, the NDVI values (figure 2) range from -0.06 to

0.60, with a mean of 0.16 and a standard deviation of

±0.11. It is considered healthy vegetation when NDVI

values are higher than 0.40, and the average value

likely implies that the area is probably primarily

covered by shrubland. This class covers roughly

318,84 ha, accounting for 48% of the total area.

GISTAM 2025 - 11th International Conference on Geographical Information Systems Theory, Applications and Management

Figure 2: NDVI of BA1.

However, the area contains Eucalyptus, Maritime

Pine, and Mixed Forest stands, which is why the

NDVI maximum value is 0.60.

The humidity index, NDII (figure 3), ranges from

0.41 to 0.38, with a mean of -0.07 and a standard

deviation of ±0.09. As can be observed, moisture

content in September had a negative mean, and the

lowest values are below what is considered healthy

green vegetation (between 0.02 and 0.6), showing

that the vegetation of area BA1 was suffering from

water stress.

Figure 3: NDII of BA1

Regarding PSRI (figure 4), which was multiplied by

-1 as described in Bento-Gonçalves et al., (2019a),

negative values now signify senescence, and positive

values represent healthy vegetation. This was done to

create a composition that would evaluate wildfire fuel

availability for both areas of interest. PSRI scores

vary from -0.51 to 0.06, with an average of -0.08 and

a standard deviation of ±0.034.

Figure 4: PSRI of BA1.

This indicates that the majority of the vegetation

in region BA1 falls between the values considered as

healthy green vegetation. While a minor portion

exceeds the threshold for healthy vegetation, ranging

from -0.21 to -0.51.

Figure 5: NDVI of BA2.

Satellite Images and Spectral Vegetation Indices as Auxiliary Tools to Monitoring Fuel Availability in Areas Prone to Wildﬁre: Study Case

in the Northern Region of Portugal

Area BA2 is substantially larger than BA1,

covering around 5,433 hectares. The NDVI (Figure 5)

suggests that the vegetation is photosynthetically

active, as the values range from -0.21 to 0.88, with a

mean of 0.68 and a standard deviation of ±0.12. The

forested impacted area is mostly covered by

Eucalyptus, followed by Maritime Pine stands, and

small regions of mixed forest. The high average value

is most likely owing to the region's substantial

forested land.

Figure 6: NDII of BA2.

The NDII (figure 6) minimum value is -0.51, and

the maximum value is 0.62, with a mean of 0.22 and

a standard deviation of ±0.14. This indicates that most

of the vegetation's humidity was closer to the

threshold of what is considered water stress than

healthy vegetation. Furthermore, a wildfire occurred

in October of 2017, when the weather was dry and

hot, which is unusual for this month (IPMA, 2017).

All of these elements most likely contributed to the

escalation of these events, which had ramifications

for Vale do Sousa's social, economic, and

environmental conditions.

Regarding the VI senescence (figure 7), the

PSRI's minimum value is -0.93, indicating extreme

senescence in some portion of the wooded area, the

highest value, which indicates healthy vegetation of

0.40. Although the negative mean value of -0.12 with

a standard deviation of ±0.06, is within what is

considered healthy vegetation, demonstrates the

effect of the high negative values on the mean,

changing the distribution of the data.

With this information, it was possible to compute

the fuel availability for both areas (Bento-Gonçalves

et al., 2019a). As shown in Figure 7, the fuel ranges

from very low to moderate in BA1 and very low to

very high in BA2. Nonetheless, only classes above 2

were considered wildfire fuel, implying that, while

both areas are primarily covered by class 1, there are

small patches of class 2 and 3 in BA1, and areas

classified from 2 to 5 in BA2, indicating that fuel was

available in both areas.

Figure 7: PSRI of BA2.

The ML method used to perform the classification of

the selected scenes of 2017 and 2018 was Support

Vector Machine, followed by the computation of a

confusion matrix and accuracy assessment. The

classification was evaluated using the Kappa Index,

which is greater than 0,8, indicating high accuracy of

the classification. Image from 2017 was classified to

identify the fuel type (Fernandes et al., 2006) and, as

a result, fuel hazards maps were created for both

areas, is also shown in Figure 7.

Aside from water stress, senescence, and

vegetation health, wildfires are also affected by

climate, slope and fuel hazards. Because fuel hazard

influences fire spread and magnitude. As can be seen,

a large section of BA1 is classified as 4 (High), with

214 ha classified as 5 (very high). Demonstrating how

forest cover types can have an important influence on

these events. It is worth mentioning that for this study,

the vegetation structure was not evaluated. Classes 3–

5 are related to Maritime Pine, shrubland, and

Eucalyptus. Maritime Pine was classified as 3

(moderate) because it has fewer stands than

Eucalyptus. Shrub was classed as 4 (High) because

the area with this classification is quite close to the

GISTAM 2025 - 11th International Conference on Geographical Information Systems Theory, Applications and Management

Figure 8: Wildfire fuel and Fuel hazard maps of 2017.

above-listed tree species' stands. According to

(Fernandes et al., 2006), the subcover can affect the

spread and fire hazard. Eucalyptus covers a larger

area in BA2, accounting for 52% of the entire area.

Maritime pine represents 23.4% of the total area.

When looking at VIs, it is easy to see lower or

extreme values of the forest's biophysical state. The

significant fuel availability and hazards in the study

area most likely contributed to the large fire. From

2017 to 2018, the change in landcover type of BA1

reveals a 35% decline in Eucalyptus and 52.5% in

maritime pine stands. These percentages reflect the

conversion of those categories to bare and shrubland

cover types in 2018.

There were no eucalyptus losses in the case of

BA2, but the same did not occur to maritime pine

stands. According to the changes observed between

2017 and 2018, the eucalyptus area had an increase of

1%. Whereas maritime pine lost nearly 90% of its

area. However, this loss may be due not only to

wildfire effects but also to harvesting, as Eucalyptus

plantations have increased in recent years in the study

Satellite Images and Spectral Vegetation Indices as Auxiliary Tools to Monitoring Fuel Availability in Areas Prone to Wildﬁre: Study Case

in the Northern Region of Portugal

area. It is worth noting that bare land and shrubland

cover types have also increased. They accounted for

barely 3% in 2017, but almost 30% in 2018.

As illustrated, all of these factors may influence

extreme wildfire outbreaks. Although fuel

availability with extreme value only accounts for a

small percentage of the region of interest, when

combined with VIs and fuel hazard score, it revealed

that those locations were prone to wildfires before

they occurred, both PSRI and fuel hazard classes had

higher values. When these remote sensing products

are combined, they can be valuable in pre-fire

management actions by identifying where

precautions should be taken to reduce losses caused

by wildfires, as shown in the change analysis.

One weakness of this study is the classification of

fuel hazards. Because forest structure was not

included in this study. In this manner, the Portuguese

fuel model was modified to be used with the classified

image. Classes were clustered, which could affect

hazardous classification, as the structure becomes

more important to fire behavior than tree species

(Fernandes et al., 2006). It is critical to determine the

structural types of the stands (closed, low, or tall;

open, low, or tall) if the objective is to develop a fuel

hazard model alongside VIs and fuel availability.

Despite this limitation, the primary goal of this

study was to identify available remote sensing-

derived products as support tools for fuel monitoring

and management throughout the wildfire season.

Since both wildfires occurred in areas where VIs

indicated great senescence, low water content and

less greenness, fuel availability could be identified in

the study region. With fuel availability classified as

low to medium in BA1, and low to very high in BA2.

The modified fuel hazards model for both

locations identified vast areas classified as high or

extremely high. It was possible to demonstrate that

the remote sensing-derived data information is

reliable in detecting alterations in forest biophysical

properties. Furthermore, this information could

potentially be employed as an auxiliary tool or as

information when carrying out the most effective fire

control measures in order to manage fuel loading

while decreasing fire severity and fulfil forest fire

management objectives.

4 CONCLUSIONS

Monitoring forest fire fuels is critical for

implementing best practices management approaches

to mitigate wildfire threats to forest ecosystems,

population health, and financial losses. With the

availability of a variety of satellite constellations,

such as Sentinel-2, VIs may be used as additional

tools to monitor fuel loading and hazards, hence

reducing the severity of wildfires.

Data is available and accessible across a variety

of platforms, and VIs resulting from remote sensing

imagery can be computed through different GIS tools.

Furthermore, decision-makers could employ these

spatial data to support forest fire prevention activities.

Despite its limitations, this approach remains a

significant resource, particularly in locations where

fieldwork is not feasible owing to financial or

logistical constraints. The attempts to reduce wildfire

fuel could be improved by looking at VIs along with

fuel hazard models and other management actions,

suggesting which regions need to be investigated

when selecting where the efforts to reduce fuels

should be invested.

Further studies should be performed to verify

additional burned areas from previous years utilizing

alternative satellite constellations such as Landsat. In

addition, other VIs should be explored to verify the

forest's biophysical composition. Although LiDAR

data was not used in this study, it is still an important

remote sensing data as it may provide information

about the structure of the forest in less time and assess

understory fuels.

ACKNOWLEDGEMENTS

Bárbara Pavani-Biju was financially supported by

Portuguese national funds through the Foundation for

Science and Technology I.P. (grant reference

number: 2022.13033.BD), Decision Support for the

Supply of Ecosystem Services under Global Change

(DecisionES) (grant agreement number:

101007950—H2020-MSCA-RISE-2020), and

through FCT – Fundação para a Ciência e Tecnologia,

I.P., in the framework of the ICT project with the

references UIDB/04683/2020 and UIDP/04683/2020.

José G. Borges and Susete Marques were

financially supported by Portuguese national funds

through the Foundation for Science and Technology

I.P. (project references UIDB/00239/2020 and

UIDP/00239/2020 of the Forest Research Centre and

DOI identifier 10.54499/UIDB/00239/2020 and

10.54499/UIDP/00239/2020 and the Associate

Laboratory Terra (TERRA LA/P/0092/2020. Susete

Marques was also funded by national funds through

the FCT, in the scope of Norma Transitória –

DL57/2016/CP13827CT15.

GISTAM 2025 - 11th International Conference on Geographical Information Systems Theory, Applications and Management

REFERENCES

Avetisyan, D., Stankova, N., & Dimitrov, Z. (2023).

Assessment of Spectral Vegetation Indices

Performance for Post-Fire Monitoring of Different

Forest Environments. Fire, 6(8). https://doi.org/10.33

90/fire6080290

Bento-Gonçalves, A., Vieira, A. A. B., Costa, M. R. M., &

Aranha, J. T. M. (2017). Wildfires Perspectives, Issues

And Challenges Of The 21st Century (1st ed., Vol. 1).

Nova Science Publisher.

Bento-Gonçalves, A., Vieira, A., Baptista, G., Rocha, J., &

Moura, S. (2019a). The 2017 Large Wildﬁre of Braga -

Evaluation of the Different Conditions of the Burned

Vegetation. Geo-Eco-Trop, 43, 627–640. chrome-

extension://efaidnbmnnnibpcajpcglclefindmkaj/https://

www.geoecotrop.be/uploads/publications/pub_434_13

.pdf

Bento-Gonçalves, A., Vieira, A., Baptista, G., Rocha, J., &

Moura, S. (2019b). The 2017 Large Wildfire of Braga -

Evaluation of the Different Conditions of the Burned

Vegetation. International Journal of Tropical Geology,

Geography and Ecology, 4(43), 627–640. chrome-

extension://efaidnbmnnnibpcajpcglclefindmkaj/https://

www.geoecotrop.be/uploads/publications/pub_434_13

.pdf

Chughtai, A. H., Abbasi, H., & Karas, I. R. (2021). A

review on change detection method and accuracy

assessment for land use land cover. Remote Sensing

Applications: Society and Environment, 22, 100482.

https://doi.org/10.1016/j.rsase.2021.100482

de Almeida, D. R. A., Stark, S. C., Valbuena, R., Broadbent,

E. N., Silva, T. S. F., de Resende, A. F., Ferreira, M. P.,

Cardil, A., Silva, C. A., Amazonas, N., Zambrano, A.

M. A., & Brancalion, P. H. S. (2020). A new era in

forest restoration monitoring. Restoration Ecology,

28(1), 8–11. https://doi.org/10.1111/rec.13067

DIRECTIVA 92/43/CEE DO CONSELHO Relativa à

Preservação Dos Habitats Naturais e Da Fauna e Da

Flora Selvagens (1992). https://eur-lex.europa.eu/Lex

UriServ/LexUriServ.do?uri=CONSLEG:1992L0043:2

0070101:PT:HTML#tocId6

Dos Santos, S. M. B., Bento-Gonçalves, A., Franca-Rocha,

W., & Baptista, G. (2020). Assessment of burned forest

area severity and postfire regrowth in chapada

diamantina national park (Bahia, brazil) using dnbr and

rdnbr spectral indices. Geosciences (Switzerland),

10(3). https://doi.org/10.3390/geosciences10030106

ESA. (2024). Copernicus Data Space Ecosystem Browser.

https://browser.dataspace.copernicus.eu/

ESA Sentinel Application Platform. (2024). SNAP (11.0.0).

http://step.esa.int/

Esri, & Environmental Systems Research Institute, I.

(2015). ArcGIS Resources. ArcGIS 10.3 Help.

http://resources.arcgis.com/en/help/

Fernandes, P., & Loureiro, C. (2021). Modelos de

combustível florestal para Portugal. https://www.

researchgate.net/publication/357812218

Fernandes, P., Luz, A., Loureiro, C., Ferreira-Godinho, P.,

& Botelho, H. (2006). Fuel modelling and fire hazard

assessment based on data from the Portuguese National

Forest Inventory. Forest Ecology and Management

234, S229. https://doi.org/10.1016/j.foreco.2006.08.2

Gorelick, N., Hancher, M., Dixon, M., Ilyushchenko, S.,

Thau, D., & Moore, R. (2017). Google Earth Engine:

Planetary-scale geospatial analysis for everyone.

Remote Sensing of Environment, 202, 18–27.

https://doi.org/10.1016/j.rse.2017.06.031

Hardisky, M. A., Klemas, V., & Smart, R. M. (1983). The

Influence of Soil Salinity, Growth Form, and Leaf

Moisture on-the Spectral Radiance of partina

alterniflora Canopies. Photogrammetric Engineering &

Remote Sensin, 49, 77–83.

ICNF. (2017). Relatório Extraordinário Provisório de

incêndios Florestais–2017.

IPMA. (2017). Resumo Climático. www.ipma.pt

Lee, K., van Leeuwen, W. J. D., Gillan, J. K., & Falk, D. A.

(2024). Examining the Impacts of Pre-Fire Forest

Conditions on Burn Severity Using Multiple Remote

Sensing Platforms. Remote Sensing, 16(10).

https://doi.org/10.3390/rs16101803

Ma, S., Zhou, Y., Gowda, P. H., Dong, J., Zhang, G.,

Kakani, V. G., Wagle, P., Chen, L., Flynn, K. C., &

Jiang, W. (2019). Application of the water-related

spectral reflectance indices: A review. In Ecological

Indicators (Vol. 98, pp. 68–79). Elsevier B.V.

https://doi.org/10.1016/j.ecolind.2018.10.049

Marques, S., Marto, M., Bushenkov, V., Mcdill, M., &

Borges, J. G. (2017). Addressing Wildfire Risk in Forest

Management Planning with Multiple Criteria Decision

Making Methods. https://doi.org/10.3390/su9020298

McCoy, Jill., & Johnston, K. (2001). Using ArcGIS Spatial

Analyst. In GIS by ESRI.

Merzlyak, M. N., Gitelson, A. A., Chivkunova, O. B., &

Rakitin, V. Y. (1999). Non-destructive optical detection

of pigment changes during leaf senescence and fruit

ripening.

Pavani-Biju, B., Borges, J. G., Marques, S., & Teodoro, A.

C. (2024). Enhancing Forest Site Classification in

Northwest Portugal: A Geostatistical Approach

Employing Cokriging. Sustainability (Switzerland),

16(15). https://doi.org/10.3390/su16156423

QGIS Association. (2024). QGIS.org. https://www.qgis.

org/

Rouse, J. W., Haas, R. H., & Deering, D. W. (1973).

Monitoring Vegetation Systems In The Great Plains

With Erts.

Roy, M., David, N. K., Danao, J. V., Baribault, H., Tian,

H., & Giorgetti, M. (2006). Remote Sensing of Fire

Severity: Assessing the Performance of the Normalized

Burn Ratio. IEEE Geoscience and Remote Sensing

Letters ·, 31(1), 112–120.

San-Miguel-Ayanz, J., Oom, D., Artes, T., Viegas, D. X.,

Fernandes, P., Faivre, N., Freire, S., Moore, P., Rego,

F., & Castellnou, M. (2017). Forest Fires in Portugal in

2017. In A. Casajus Valles, M. Marin Ferrer, K.

Poljanšek, & I. Clark (Eds.), Science for Disaster Risk

Management 2020: acting today, protecting tomorrow,

Satellite Images and Spectral Vegetation Indices as Auxiliary Tools to Monitoring Fuel Availability in Areas Prone to Wildﬁre: Study Case

in the Northern Region of Portugal

(1st ed., Vol. 1). Publications Office of the European

Union. https://doi.org/10.2760/571085

SNIG. (2022). Carta de Uso e Ocupação do Solo . Carta de

Uso e Ocupação Do Solo .

Teodoro, A. C., & Duarte, L. (2013). Forest fire risk maps:

A GIS open source application - a case study in Norwest

of Portugal. International Journal of Geographical

Information Science, 27(4), 699–720. https://doi.org/

10.1080/13658816.2012.721554

U.S. Geological Survey. (2024). Landsat Data Access.

https://www.usgs.gov/landsat-missions/landsat-data-

access

Zeng, Y., Hao, D., Huete, A., Dechant, B., Berry, J., Chen,

J. M., Joiner, J., Frankenberg, C., Bond-Lamberty, B.,

Ryu, Y., Xiao, J., Asrar, G. R., & Chen, M. (2022).

Optical vegetation indices for monitoring terrestrial

ecosystems globally. Nature Reviews Earth and

Environment, 3(7), 477–493. https://doi.org/10.1038/s4

3017-022-00298-5

GISTAM 2025 - 11th International Conference on Geographical Information Systems Theory, Applications and Management