EO4WildFires: An Earth Observation multi-sensor, time-series

machine-learning-ready benchmark dataset for wildfire impact

prediction
Dr. Dimitrios Sykas*a, Dimitrios Zografakisa, Konstantinos Demestichasa, Constantina
Costopouloua, Pavlos Kosmidisb
a
Department of Agricultural Economy & Development of the Agricultural University of Athens –
[email protected], bCatalink Ltd

ABSTRACT

This paper presents a benchmark dataset called EO4WildFires; a multi-sensor (multi spectral; Sentinel-2, Synthetic-
Aperture Radar - SAR; Sentinel-1, meteorological parameters; NASA Power) time-series dataset that spans 45 countries,
which can be used for developing machine learning and deep learning methods targeted for the estimation of the area that
a forest wildfire might cover.
This novel EO4WildFires dataset is annotated using EFFIS (European Forest Fire Information System) as forest fire
detection and size estimation data source. A total of 31,730 wildfire events are gathered from 2018 to 2022. For each event,
Sentinel-2 (multispectral), Sentinel-1 (SAR) and meteorological data are assembled into a single data cube. The
meteorological parameters that are included in the data cube are: ratio of actual partial pressure of water vapor to the partial
pressure at saturation, average temperature, bias corrected average total precipitation, average wind speed, fraction of land
covered by snowfall, percent of root zone soil wetness, snow depth, snow precipitation, as well as percent of soil moisture.
The main problem that this dataset is designed to address, is the severity forecasting before wildfires occur. The dataset is
not used to predict wildfire events, but rather to predict the severity (size of area damaged by fire) of a wildfire event, if
that happens in a specific place under the current and historical forest status, as recorded from multispectral and SAR
images, and meteorological data.
Using the data cube for the collected wildfire events, the EO4WildFires dataset is used to realize three (3) different
preliminary experiments, to evaluate the contributing factors for wildfire severity prediction. The first experiment evaluates
wildfire size using only the meteorological parameters, the second one utilizes both the multispectral and SAR parts of the
dataset, while the third exploits all dataset parts. In each experiment, machine learning models are developed, and their
accuracy is evaluated. The results show that the size of wildfire events can be estimated better using Sentinel-2 data.
Second in terms of accuracy is Sentinel-1, while the usage of only meteorological data presented the lowest accuracy
among the three.
The dataset is published with an Open Access license and is hosted at: 10.5281/zenodo.7762564.
Keywords: forest wildfires, earth observation, machine learning dataset, multi-sensor data

1. INTRODUCTION
Forest wildfires constitute a pressing global concern, causing widespread environmental, economic, and social damage.
The increasing frequency and intensity of these events have been attributed to factors such as climate change, deforestation,
and human activities. Accurate and timely prediction of wildfire severity, particularly the size of the area that a fire might
cover, is crucial for effective disaster management, resource allocation, and mitigation efforts. Remote sensing data,
especially from satellite platforms such as Sentinel-1 and Sentinel-2, as well as meteorological data from sources like
NASA Power or European Centre for Medium-Range Weather Forecasts (ECMWF), provide a wealth of information that
can be harnessed to develop machine learning and deep learning models for wildfire severity estimation.
In this paper, we introduce a novel benchmark dataset called EO4WildFires, which combines multispectral (Sentinel-2),
Synthetic-Aperture Radar - SAR (Sentinel-1), and meteorological data from 30 countries to create a comprehensive, multi-
sensor time-series dataset. Annotated using the European Forest Fire Information System (EFFIS) for wildfire detection

Ninth International Conf. on Remote Sensing and Geoinformation of the Environment (RSCy2023), edited by
K. Themistocleous, et al., Proc. of SPIE Vol. 12786, 1278603 · © The Authors. Published under a
Creative Commons Attribution CC-BY License · doi: 10.1117/12.2680777

Proc. of SPIE Vol. 12786 1278603-1

and size estimation, EO4WildFires spans 31,742 wildfire events between 2018 and 2022. By assembling Sentinel-2,
Sentinel-1, and meteorological data into a single data cube for each event, this dataset allows for a more in-depth analysis
of the factors contributing to wildfire severity.
This paper aims to address the challenge of forecasting forest wildfire severity before the events occur, focusing on the
potential size of the area damaged by fire in a specific location given the current and historical forest status. To this end,
we design three experiments to evaluate the contribution of various factors in predicting wildfire size using the
EO4WildFires dataset. Each experiment explores different combinations of meteorological parameters, multispectral, and
SAR data, and assesses the accuracy of machine learning models developed based on these inputs. Through these
experiments, we aim to provide insights into the development of more accurate and reliable wildfire severity prediction
models that can better inform decision-makers and support wildfire management efforts worldwide.

1.1 Deep Learning Architectures

Deep learning technologies are evolving at fast pace. This sub-section provides a brief overview of recent notable deep
learning architectures that make use of large-scale visual data.
The authors in [1] introduce the technique of Contrastive Captioner (CoCa), a novel image-text encoder-decoder
foundation model that combines contrastive loss and captioning loss. This approach incorporates elements of both CLIP
and Simple Visual Language Model (SimVLM). CoCa is designed to first focus on unimodal text representations and then
on multimodal image-text representations. The model uses contrastive loss between unimodal embeddings and captioning
loss on multimodal decoder outputs. CoCa is pretrained end-to-end on web-scale alt-text data and annotated images,
unifying natural language supervision for representation learning. It achieves state-of-the-art performance on various tasks,
including visual recognition, cross-modal retrieval, multimodal understanding, and image captioning. Notably, CoCa
reaches 91.0% top-1 accuracy on ImageNet with a finetuned encoder.
The authors in [2] explore large-scale models in computer vision and address three major challenges: training instability,
resolution gaps, and dependence on labeled data. The authors propose three techniques: 1) a residual-post-norm method
with cosine attention for training stability, 2) a log-spaced continuous position bias method to transfer models between
different resolution tasks, and 3) SimMIM, a self-supervised pretraining method that reduces the need for labeled images.
The paper successfully trains a 3 billion-parameter Swin Transformer V2 model, the largest dense vision model to date,
capable of handling images up to 1,536x1,536 resolution. This model sets new performance records on ImageNet-V2,
COCO, ADE20K, and Kinetics-400 tasks. Moreover, the training is 40 times more efficient in terms of labeled data and
training time compared to Google's billion-level visual models.
The study in [3] demonstrates that the Transformer architecture, which has become the standard for natural language
processing tasks, can also be effectively applied to computer vision without relying on convolutional networks. The authors
introduce Vision Transformer (ViT), a pure transformer applied directly to sequences of image patches. When pre-trained
on large datasets and transferred to various image recognition benchmarks (ImageNet, CIFAR-100, VTAB, etc.), ViT
achieves excellent results compared to state-of-the-art convolutional networks while requiring significantly fewer
computational resources to train.
A method for algorithm discovery as program search, focusing on optimization algorithms for deep neural network training
is presented in [4]. The authors introduce a technique called Evolved Sign Momentum and codenamed Lion, a simple,
memory-efficient optimization algorithm that outperforms widely used optimizers like Adam and Adafactor in various
tasks. On image classification, Lion improves ViT’s accuracy by up to 2% on ImageNet and reduces pre-training compute
on JFT datasets by up to 5x. In vision-language contrastive learning, Lion achieves 88.3% zero-shot and 91.1% fine-tuning
accuracy on ImageNet. For diffusion models, it outperforms Adam by achieving a better Fréchet inception distance (FID)
score and reducing training compute by up to 2.3x. Lion performs similarly or better compared to Adam in autoregressive,
masked language modeling, and fine-tuning tasks. Its performance gain grows with larger training batch sizes and requires
a smaller learning rate due to the larger norm of the update produced by the sign function.

1.2 Wildfires and Earth Observation

Satellites have a role to play in detecting, monitoring and characterizing fires [5]. This sub-section provides an overview
of notable recent research studies that produce or utilize Earth Observation data for purposes of wildfire detection.

Proc. of SPIE Vol. 12786 1278603-2

The authors in [6] compare MODIS fire products [5] with ground wildfire investigation records from December 2002 to
November 2015 in Yunnan Province, Southwest China. The goal is to understand the differences in spatiotemporal patterns
of regional wildfires detected by the two approaches, estimate the omission error of MODIS fire products, and explore
how local environmental factors influence MODIS wildfire detection probability. The results show that MODIS records
at least twice as many wildfire events compared to ground records, but the distribution patterns are inconsistent. Only
11.10% of the 5,145 confirmed ground records were detected using multiple MODIS fire products. The study found that
fire size is a primary limiting factor for MODIS fire detection capacity, with a 50% probability of detecting wildfires at
least 18 hectares in size. Other factors influencing MODIS wildfire detection probability include weather factors (daily
relative humidity and wind speed) and altitude of wildfire occurrence. The study highlights the importance of considering
local conditions and ground inspection in wildfire monitoring, management, and global wildfire simulations.
The authors in [7] present a method that combines Big Data, Remote Sensing, and Data Mining algorithms (Artificial
Neural Network and Support Vector Machines) to process data from satellite images and predict wildfire occurrences. The
authors create a dataset based on Remote Sensing data related to crop conditions (Normalized Difference Vegetation Index
- NDVI), meteorological conditions (Land Surface Temperature – LST), and the fire indicator “Thermal Anomalies” from
the MODIS instrument on Terra and Aqua satellites. The dataset is publicly available on GitHub. Experiments were
conducted using the big data platform “Databricks,” achieving high prediction accuracy (98.32%). The results were
assessed using various validation strategies and compared with existing wildfire early warning systems.
The authors in [8] present "Next Day Wildfire Spread", a large-scale, multivariate dataset of historical wildfires in the
United States, based on nearly a decade of remote-sensing data. Unlike existing fire datasets, this dataset combines 2-D
fire data with multiple explanatory variables (topography, vegetation, weather, drought index, and population density) over
2-D regions, creating a feature-rich dataset for machine learning. The authors implement a neural network that leverages
the spatial information in the data to predict wildfire spread, comparing its performance with logistic regression and random
forest models. This dataset serves as a benchmark for developing remote-sensing-based wildfire propagation models with
a lead time of one day.
The authors in [9] propose a cost-effective, machine-learning-based approach to predict forest fires in Indonesia using
remote sensing data. This addresses the challenges faced by developing countries that cannot afford expensive ground
instruments used in traditional prediction systems. The proposed model achieves over 0.81 area under the receiver operator
characteristic (ROC) curve, significantly outperforming the baseline approach, which never exceeds 0.70 area under the
ROC curve. The model maintains its performance even with reduced data, demonstrating the potential for machine-
learning-based approaches to create reliable and cost-effective forest fire prediction systems.

2. METHODOLOGY
2.1 Scientific Problem
The scientific problem addressed by the EO4WildFires dataset is to enable the development of a model or set of models
for forecasting the severity of a future wildfire event in a specific location, based on current and past (30 days)
meteorological data, multispectral and SAR images. These data should model the forest status before a wildfire event
occurs. The goal is not to predict wildfire events, but rather to predict the severity, specifically the size of the area damaged
by the fire, before it occurs. This problem requires the development of predictive models that can integrate the various
types of data available in the dataset to generate accurate severity forecasts. The solution to this problem has the potential
to help forest protection services and other stakeholders to better prepare for and respond to wildfires, ultimately reducing
their impact on the environment and society.
As a scientific question, it can be articulated as follows:
“What is the potential of the provided dataset, which includes meteorological data, multispectral and SAR images, to
forecast the severity (size of area damaged by fire) of a future wildfire event in a specific location, based on current and
near past forest status?”
To address this, the European Forest Fire Information System (EFFIS), Copernicus Sentinel-1 & 2, and NASA Power data
sources are compiled in a data cube format to enable deep learning modeling.

Proc. of SPIE Vol. 12786 1278603-3

2.2 European Forest Fire Information System
EFFIS is a platform that provides up-to-date information on wildland fires in Europe to support forest protection services.
The website is maintained by the European Commission Joint Research Centre (JRC) and provides data on historical and
current wildfires, as well as forecasts of wildfire danger levels. The information on EFFIS includes maps, data on the
location, size, and intensity of wildfires, as well as the affected vegetation type and land use. EFFIS also provides a daily
fire danger forecast based on meteorological data and a fire danger rating system. The website is an essential resource for
forest protection services, researchers, and policymakers, who rely on accurate and timely information to monitor and
manage wildfires in Europe.
2.3 Copernicus Sentinel 1 & 2
Sentinel-1 is an Earth observation mission from the Copernicus Programme, developed by the European Space Agency
(ESA). It consists of a constellation of two polar-orbiting satellites that systematically acquire Synthetic Aperture Radar
(SAR) imagery at high spatial resolution over land and coastal waters. The SAR sensor operates in the C-band frequency
and provides all-weather and day-and-night imaging capabilities, making it useful for a wide range of applications,
including monitoring of land and ocean surfaces, disaster management, and maritime surveillance. The data from Sentinel-
1 is freely available and can be accessed through the Sentinel Data Hub or through third-party data providers. The mission
has been designed to provide long-term, global data acquisition, making it an essential tool for monitoring and managing
natural resources and the environment.
Sentinel-2 is an Earth observation mission from the Copernicus Programme, developed by the European Space Agency
(ESA). It consists of a constellation of two polar-orbiting satellites that systematically acquire optical imagery at high
spatial resolution over land and coastal waters. The sensor on board the satellites captures images in 13 spectral bands,
ranging from the visible to the shortwave infrared regions of the electromagnetic spectrum. This enables the detection and
monitoring of a wide range of phenomena, including vegetation cover, land use, natural disasters, and urban development.
The data from Sentinel-2 is freely available and can be accessed through the Sentinel Data Hub or through third-party data
providers. Sentinel-2 has been designed to provide global coverage and high revisit time, making it a valuable tool for
environmental monitoring, land use mapping, and disaster management.
2.4 NASA Power
NASA Power is a scientific resource that provides solar and meteorological data sets from NASA research to support
various applications, such as renewable energy, building energy efficiency, and agricultural needs. The scientific problem
addressed by NASA Power is to make solar and meteorological data more accessible and usable for researchers,
policymakers, and practitioners working in different fields. This problem requires the development of reliable and accurate
methods for measuring and predicting solar and meteorological variables, such as solar radiation, temperature,
precipitation, and wind speed. NASA Power provides users with access to a range of data sets, tools, and services that can
be used to support research, planning, and decision-making related to energy, agriculture, and other sectors. The solutions
provided by NASA Power have the potential to contribute to a more sustainable and resilient future by enabling informed
decisions based on accurate and up-to-date solar and meteorological data.
2.5 Dataset Structure
The dataset is a combination of meteorological data, multispectral and SAR satellite images, and wildfire event labels
provided by the EFFIS system. The EFFIS system is responsible for providing updated and reliable information on
wildland fires in Europe to support forest protection services.
The dataset was created using the Sentinel-hub API and the NASA Power API. The Sentinel-hub API allows users to make
Web Map Service (WMS) and Web Coverage Service (WCS) web requests to download and process satellite images from
various data sources. The NASA Power API provides solar and meteorological data sets from NASA research to support
renewable energy, building energy efficiency, and agricultural needs.
Table 1. Parameters included in each wildfire event for each data source
Meteorological Data Sentinel-1 Sentinel-2

Proc. of SPIE Vol. 12786 1278603-4

 ratio of actual partial pressure of
water vapor to the partial pressure at VV Band 02
saturation (RH2M)
average temperature (T2M) VH Band 03
bias corrected average total
(VV-VH)/(VV+VH) Band 04
precipitation (PRECTOTCORR)
average wind speed (WS2M) Band 05
fraction of land covered by snowfall
Band 08
(PRECSNOLAND)
percent of root zone soil wetness
Band 11
(GWETROOT)
snow depth (SNODP)
snow precipitation (FRSNO)
soil moisture (GWETTOP)

In our produced dataset, EO4WildFires, each wildfire event is packaged as a datacube in NetCDF format, resulting in a
total of 31,740 files, equal to the number of wildfire events in the dataset. The data for each wildfire event are collected as
follows:
• Bounding box coordinates and the date of the event are used as inputs.
• Meteorological parameters are extracted using the center of the area as the coordinate. Parameters are collected
from one day before the event until 30 days earlier than that date.
• Sentinel-2 images are subset using the bounding box coordinates. A mosaicking process [https://custom-
scripts.sentinel-hub.com/sentinel-2/monthly_composite/#] is applied to overcome cloud cover issues by selecting
the best available pixels of the last 30 days before the wildfire event.
• Sentinel-1 images are also cropped using the bounding box coordinates. Since cloud cover is not an issue with
SAR images, there is no need to generate a mosaic. Thus, the most recent image before the event is selected.
Sentinel-1 images are also cropped using the bounding box coordinates. Both ascending and descending images
are provided.
• Burned area mask, a boolean mask of the burned area is also provided based on the rasterized EFFIS vector data
on the Sentinel-2 grid.
The EO4WildFires dataset contains a rich set of features (Table 1) that can be used for wildfire severity prediction. The
combination of meteorological data and satellite images can provide insights into the environmental conditions that can
lead to the outbreak of wildfires. The wildfire event labels provided by the EFFIS system can be used to train predictive
models for early detection and response to wildfire events.
The mosaicking process is used for creating a monthly composite image using Sentinel-2 satellite imagery. The process
selects the best pixel for each date in the last 31 days based on a ratio of the available bands. The selection process is
designed to avoid cloud cover, and the criteria used depends on the level of blue in the image. If blue is less than 0.12, the
date with the maximum ratio of band B08 against band B02 is chosen. If no pixel is available above this threshold, the
date with the maximum ratio of band B03 against band B02 is selected when blue is less than 0.45. If water is detected in
the image, the date with the maximum ratio of band B02 against band B08 is chosen. If snow is detected, the median of
the scene with snow is used. The resulting composite image provides a cloud-free representation of the last 31 days in the
selected region.
A total of 31,730 wildfire events ranging from 2018 to 2022 across 45 countries, covering 8,707 level-4 administrative
areas. We have used the GDAM (https://gadm.org/) database to correlate the polygons of the detected events from EFFIS

Proc. of SPIE Vol. 12786 1278603-5

with the level-4 administrative boundaries. After analysing the distribution of the data, it was found that the median size
of a wildfire is 31.0 hectares while the mean is 128.77 hectares. An impressive 54,769.0 hectares was the largest recorded
wildfire in 2021 in Antaya, Turkey, while the second largest wildfire was 51,245.0 hectares in 2021 in Evoia, Greece. The
dataset is openly available at: 10.5281/zenodo.7762564
Figure 1 depicts on the map the level-4 administration boundaries of the recorded wildfire events (2018-2022) within the
EO4WildFires dataset. For instance, we notice that specific areas in the Mediterranean have high concetrations in wildfire
events. Figure 2 illustrates the total number of wildfires events for each year (2018-2022) per country. For instance, we
may notice that the total number of detected events was higher in Ukraine in 2022 compared to previous years. Figure 3
depicts the median size of wildfires in hectares per country, where we can see that the Netherlands, Belgium and Austria
demonstrate the largest values.

Figure 1. Wildfires events (2018-2022) with the corresponding affected level-4 administration boundaries – yellow
polygons: administrative boundaries (level 4) of affected areas, red points: locations of wildfires events.

Proc. of SPIE Vol. 12786 1278603-6

Figure 2. Total number of wildfires events for each year (2018-2022) per country. The size and color of the circles represent
the total number of detected events.

Figure 3. Median size of wildfires in hectares per country.

Proc. of SPIE Vol. 12786 1278603-7

 3. EXPERIMENTAL RESULTS FOR WILDFIRE SIZE PREDICTION

3.1 Experiments Description

This section describes three experiments aimed at predicting the severity (size) of a wildfire event utilizing the composed
EO4WildFires dataset. In the first experiment, called Meteo, the input features used for prediction are only meteorological
parameters listed in Table 1. In the second experiment, called S1, the input features used are the backscatter bands from
Sentinel-1, and in the third experiment, called S2, the input features used are the mosaicked spectral bands from Sentinel-
2. Specifically, for the S1 experiment, two different sub-experiments were conducted, one for the ascending and one for
the descending orbits. This subdivision was put in place since the appearance of land features varies significantly depending
on the illumination angle of the objects by the SAR beams. To maximize the potential of SAR data, ascending and
descending data were isolated to minimize response variations due to illumination angles. This allows to focus the detection
capabilities of the deep learning networks on the data variations that contribute into the prediction of the wildfire size.
The original EO4WildFires dataset is split into training, validation, and testing subsets, with 20,307 events in the training
set, 5,077 events in the validation set, and 6,346 events in the test set. The division of the dataset is the same for all three
experiments to ensure comparability of results. The goal of the experiments is to determine which input features provide
the best prediction performance for the size of a wildfire event.
Heavily inspired by popular datasets like coco [10], three index files (train/val/test) are created that operate as literal file
catalogues. Each row in the index files refers to a specific file in the disk.
3.2 Experiments Results
Sentinel-1 and Sentinel-2 can be illustrated as images that incorporate 3-dimension and 6-dimension channels,
respectively. Therefore, it is logical to implement a popular feature extraction architecture to accumulate knowledge to a
final classification layer. For this task, ResNet-32 [11] model was chosen, as it offers robust feature extraction given its
total layer and parameter count. A modification was added to the optimizer, as it was trained using Lion [4] optimizer,
which, given the latest advances, has shown great potential, as explained in Section 1.1.
Our experiments did not rely on pretrained weights for model initialization, but training was rather done from zero. The
only difference between the experiments S1 and S2 was the input channel dimension in ResNet, as stated before. ResNet
by default accepts 3 channels for 224x224 size images. S1 ascending and descending data match that description with
minimum processing, but Sentinel 2A data do not. Finally, vanilla ResNet input dimensions, images and masks were
padded to match integral multiples of 224 in each dimension, and later they were split into 224x224 chunks. The burned
area, i.e. the classification label, was calculated as the sum of 1s in the padded image divided by (224*224).
Meteorological data are tabular and, thus, require minimum processing. Samples have been taken for thirty days prior to
each event, therefore a 30-day sequence of 9 features (RH2M, T2M, PRECTOTCORR, WS2M, FRSNO, GWETROOT,
SNODP, PRECSNOLAND, GWETTOP) can be directly fed into deep learning models primarily used for time series
forecasting. A bidirectional Long Short-Term Memory (LSTM) network was chosen, consisting of 512 hidden dimension
features and 3 layers. Outputs of the last hidden state of the model are propagated into a linear fully connected layer and
then passed into loss and error functions, respectively.
For all training schedules, “Cross Entropy” loss was utilized, outputs of the model were passed into a Sigmoid function
and then “Mean Absolute Error” was calculated to measure the distance between the predictions and the actually burned
area. Training routine implements early stopping with minimum delta of 0.001 in validation loss and patience of 7
consecutive epochs before signaling trainer to stop. Finally, all models’ optimizers have a learning rate scheduler attached
to them, namely “Reduce LROn Plateau” is utilized, which updates learning rate by 0.1 every 3 epochs.
The results of the three experiments, in terms of test loss and test error, are summarized in Table 2 and further discussed
in Section 4.
Table 2. Experiment results.

Test Loss (Cross Entropy) Test Error (MAE)

Meteo 0.422 0.129

Proc. of SPIE Vol. 12786 1278603-8

 S1 Ascending 0.524 0.066

S1 Descending 0.517 0.062

S2 0.371 0.056

4. CONCLUSIONS
This paper presents EO4WildFires, a comprehensive benchmark dataset that integrates multispectral, SAR, and
meteorological data to provide insights into the factors affecting wildfire severity. Through a series of experiments, we
demonstrate the potential of using machine learning models to predict wildfire size based on various combinations of input
data.
The EO4WildFires dataset addresses the scientific problem of forecasting the severity of future wildfire events, based on
recent past meteorological (30-days before the event) data, and most recent past multispectral, and SAR images (earliest
available images before the event). The dataset combines data from the EFFIS, Copernicus Sentinel-1 & 2, and NASA
Power to create a comprehensive source of information for developing predictive models. The dataset, which includes
31,740 files covering 31,730 wildfire events across 45 countries, provides a rich set of features that can be used to train
machine learning models to predict wildfire severity.
Three experiments were conducted to predict the size of a wildfire event using different input features: meteorological
parameters (Meteo), backscatter bands from Sentinel-1 (S1), and mosaicked spectral bands from Sentinel-2 (S2). The
ascending and descending orbits of Sentinel-1 were treated separately. The dataset was divided into training, validation,
and testing subsets, with the same division across all experiments to ensure comparability of results. ResNet-32 was chosen
as the feature extraction architecture for Sentinel-1 and Sentinel-2 data, while a bidirectional LSTM network was used for
meteorological data. The goal of these experiments was to determine which input features provide the best prediction
performance for wildfire size.
As shown in Table 2, the experiment with the lowest error was S2. Features from S2 represent the actual status (vegetation
and soil moisture, fuel and flammability) of the forest, which suggests that such data, when used as input, contribute into
the prediction of wildfire severity. On the other hand, the Meteo experiment has the highest error. Features from Meteo
include only the atmospheric conditions, which have don’t capture the full extent of the status of different vegetation
conditions. The SAR experiments present an reasonably low error, which can be attributed to the fact that although SAR
data do not directly measure the vegetation status, they do provide information about the vegetation canopy and forest
internal structure. In future work, the authors will continue to experiment with further combinations of data inputs and
different network architectures.
By leveraging this dataset, researchers, first responders and policymakers can be better informed by accurate and up-to-
date information on wildfire events and their potential severity.

ACKNOWLEDGEMENT
This study has been conducted in the framework of SILVANUS project. This project has received funding from the
European Union’s Horizon 2020 research and innovation program under grant agreement No 101037247. The contents of
this publication are the sole responsibility of the authors and can in no way be taken to reflect the views of the European
Commission.

REFERENCES

[1] Jiahui Yu, Zirui Wang, Vijay Vasudevan, Legg Yeung, Mojtaba Seyedhosseini, Yonghui Wu, “CoCa: Contrastive
Captioners are Image-Text Foundation Models”, Computer Vision and Pattern Recognition, (2022).
[2] Ze Liu, Han Hu, Yutong Lin, Zhuliang Yao, Zhenda Xie, Yixuan Wei, Jia Ning, Yue Cao, Zheng Zhang, Li Dong,
Furu Wei, Baining Guo, " Swin Transformer V2: Scaling Up Capacity and Resolution," Computer Vision and
Pattern Recognition (2022).

Proc. of SPIE Vol. 12786 1278603-9

[3] Alexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk Weissenborn, Xiaohua Zhai, Thomas Unterthiner,
Mostafa Dehghani, Matthias Minderer, Georg Heigold, Sylvain Gelly, Jakob Uszkoreit, Neil Houlsby, " An Image
is Worth 16x16 Words: Transformers for Image Recognition at Scale," Computer Vision and Pattern Recognition,
(2021).
[4] Xiangning Chen, Chen Liang, Da Huang, Esteban Real, Kaiyuan Wang, Yao Liu, Hieu Pham, Xuanyi Dong,
Thang Luong, Cho-Jui Hsieh, Yifeng Lu, Quoc V. Le, “Symbolic Discovery of Optimization Algorithms,”
Machine Learning (2023).
[5] Team, M. F. (2022, November 18). MODIS Active Fire and Burned Area Products - Home. Retrieved from
https://modis-fire.umd.edu.
[6] Ying, L., Shen, Z., Yang, M., and Piao, S., “Wildfire Detection Probability of MODIS Fire Products under the
Constraint of Environmental Factors: A Study Based on Confirmed Ground Wildfire Records,” Remote Sensing
11(24), 3031 (2019).
[7] Younes Oulad Sayad, Hajar Mousannif, Hassan Al Moatassime, “Predictive modeling of wildfires: A new dataset
and machine learning approach,” Fire Safety Journal (2019)
[8] F. Huot, R. L. Hu, N. Goyal, T. Sankar, M. Ihme and Y. -F. Chen, "Next Day Wildfire Spread: A Machine
Learning Dataset to Predict Wildfire Spreading From Remote-Sensing Data," in IEEE Transactions on
Geoscience and Remote Sensing, vol. 60, pp. 1-13, (2022).
[9] Suwei Yang, Massimo Lupascu, Kuldeep S. Meel, “Predicting Forest Fire Using Remote Sensing Data And
Machine Learning,” The Thirty-Fifth AAAI Conference on Artificial Intelligence (2021).
[10] Tsung-Yi Lin et al., 2014. Microsoft COCO: Common Objects in Context. CoRR, abs/1405.0312. Available at:
http://arxiv.org/abs/1405.0312.
[11] K. He, X. Zhang, S. Ren and J. Sun, "Deep Residual Learning for Image Recognition," 2016 IEEE Conference
on Computer Vision and Pattern Recognition (CVPR), Las Vegas, NV, USA, 2016, pp. 770-778, doi:
10.1109/CVPR.2016.90.

Proc. of SPIE Vol. 12786 1278603-10