Assessing Landscape Fire Hazard by Multitemporal Automatic Classification of Landsat Time Series Using the Google Earth Engine in West-Central Spain

Advancements in remote sensing for active fire detection: A review of datasets and methods

This study comprehensively and critically reviews active fire detection advancements in remote sensing from 1975 to the present, focusing on two main perspectives: datasets and corresponding instruments, and detection algorithms. The study highlights the increasing role of machine learning, particularly deep learning techniques, in active fire detection. Looking forward, the review outlines current challenges and future research opportunities in remote sensing for active fire detection. These include exploring data quality management and multi-modal learning, developing spatiotemporally explicit models, investigating self-supervised learning models, improving explainable and interpretable models, integrating physical-process based models with machine learning, and building digital twins to replicate wildfire dynamics and perform what-if scenario analysis. The review aims to serve as a valuable resource for informing natural resource management and enhancing environmental protection efforts through the application of remote sensing technology.

Soil biogeochemistry and microbial community dynamics in Pinus pinaster Ait. forests subjected to increased fire frequency

Fire frequency might increase in many fire-dominated ecosystems of the world due to the combined effects of global warming, land-use change and increased human pressures. Understanding how changes in fire frequency can affect the main soil biogeochemical dynamics, as well as the microbial community, in the long term is utmost important. Here we determined the effect of changes in fire frequency and other fire history characteristics on soil C and N dynamics and the main microbial groups (using soil fatty acid profiles), in Pinus pinaster forests from central Spain. Stands were chosen to differ in the number of fires (1 to 3) occurred between 1976 and 2018, in the time elapsed since the last fire and the interval undergone between the last two consecutive fires. We found that, in general, most of the studied biogeochemical and microbial variables showed clear differences between unburned and burned stands. The time elapsed since the last fire was the most important fire history covariable and governed the main soil nutrient dynamics and microbial groups. Recovery to pre-fire values took 30-40 years. Increased wildfire frequency only modified total C and nitrification rate, but results were not consistent between stands burned twice and thrice. The time interval (years) between the last two fires was not a significant covariable. The fact that some stands burnt up to thrice in a period of 43 years supports the strong capacity of this ecosystem to recover, even under an increased fire frequency.

Google Earth Engine and Artificial Intelligence (AI): A Comprehensive Review

Remote sensing (RS) plays an important role gathering data in many critical domains (e.g., global climate change, risk assessment and vulnerability reduction of natural hazards, resilience of ecosystems, and urban planning). Retrieving, managing, and analyzing large amounts of RS imagery poses substantial challenges. Google Earth Engine (GEE) provides a scalable, cloud-based, geospatial retrieval and processing platform. GEE also provides access to the vast majority of freely available, public, multi-temporal RS data and offers free cloud-based computational power for geospatial data analysis. Artificial intelligence (AI) methods are a critical enabling technology to automating the interpretation of RS imagery, particularly on object-based domains, so the integration of AI methods into GEE represents a promising path towards operationalizing automated RS-based monitoring programs. In this article, we provide a systematic review of relevant literature to identify recent research that incorporates AI methods in GEE. We then discuss some of the major challenges of integrating GEE and AI and identify several priorities for future research. We developed an interactive web application designed to allow readers to intuitively and dynamically review the publications included in this literature review.

A Highly Accurate Forest Fire Prediction Model Based on an Improved Dynamic Convolutional Neural Network

In this work, an improved dynamic convolutional neural network (DCNN) model to accurately identify the risk of a forest fire was established based on the traditional DCNN model. First, the DCNN network model was trained in combination with transfer learning, and multiple pre-trained DCNN models were used to extract features from forest fire images. Second, principal component analysis (PCA) reconstruction technology was used in the appropriate subspace. The constructed 15-layer forest fire risk identification DCNN model named “DCN_Fire” could accurately identify core fire insurance areas. Moreover, the original and enhanced image data sets were used to evaluate the impact of data enhancement on the model’s accuracy. The traditional DCNN model was improved and the recognition speed and accuracy were compared and analyzed with the other three DCNN model algorithms with different architectures. The difficulty of using DCNN to monitor forest fire risk was solved, and the model’s detection accuracy was further improved. The true positive rate was 7.41% and the false positive rate was 4.8%. When verifying the impact of different batch sizes and loss rates on verification accuracy, the loss rate of the DCN_Fire model of 0.5 and the batch size of 50 provided the optimal value for verification accuracy (0.983). The analysis results showed that the improved DCNN model had excellent recognition speed and accuracy and could accurately recognize and classify the risk of a forest fire under natural light conditions, thereby providing a technical reference for preventing and tackling forest fires.

Forest fire probability under ENSO conditions in a semi-arid region: a case study in Guanajuato

Fires can pose a threat to forest ecosystems when those ecosystems are not fire-adapted or when forest community conditions make them vulnerable to wildfires. Thus, investigating fire-prone environmental conditions is urgently needed to create action plans that preserve these ecosystems. In this sense, climate variables can determine the environmental conditions favorable for forest fires. Our study confirms that vapor pressure deficit (VPD) is an essential climate indicator for forest fires, as it is related to maximum temperatures and low humidity, representing the stress conditions for vegetation prone to fires. This study explores the extent to which ENSO phases can modulate climatic conditions that lead to high VPD over Guanajuato, a semi-arid region in central Mexico, during the dry season (March-April-May). Using fire occurrence data from MODIS (2000-2019) and Landsat 5 (1998-1999), we developed a climatic probability model for the occurrence of forest fires using VPD estimated from ERA5 reanalysis for each ENSO phase. We found that VPD and the occurrence of forest fires were higher during El Niño than under Neutral and La Niña years, with a higher risk of forest fire occurrence in Guanajuato's southern region. This study concludes that it is necessary to implement regional and local fire management plans, especially where the largest number of natural protected areas is located.

Fire-induced loss of the world's most biodiverse forests in Latin America

Fire plays a dominant role in deforestation, particularly in the tropics, but the relative extent of transformations and influence of fire frequency on eventual forest loss remain unclear. Here, we analyze the frequency of fire and its influence on postfire forest trajectories between 2001 and 2018. We account for ~1.1% of Latin American forests burnt in 2002-2003 (8,465,850 ha). Although 40.1% of forests (3,393,250 ha) burned only once, by 2018, ~48% of the evergreen forests converted to other, primarily grass-dominated uses. While greater fire frequency yielded more transformation, our results reveal the staggering impact of even a single fire. Increasing fire frequency imposes greater risks of irreversible forest loss, transforming forests into ecosystems increasingly vulnerable to degradation. Reversing this trend is indispensable to both mitigate and adapt to climate change globally. As climate change transforms fire regimes across the region, key actions are needed to conserve Latin American forests.

Contextualizing the 2019–2020 Kangaroo Island Bushfires: Quantifying Landscape-Level Influences on Past Severity and Recovery with Landsat and Google Earth Engine

The 2019–2020 Kangaroo Island bushfires in South Australia burned almost half of the island. To understand how to avoid future severe ‘mega-fires’ and how vegetation may recover from 2019–2020, we can utilize information from the bulk of historical fires in an area. Landsat time-series of vegetation change provide this opportunity, but there has been little analysis of large numbers of fires to build a landscape-level understanding and quantify drivers in an Australian context. In this study, we built a yearly cloud-free surface reflectance normalized burn ratio (NBR) time-series (1988–2020) using all available summer Landsat images over Kangaroo Island. Data were collected in Google Earth Engine and fitted with LandTrendr. Burn severity and post-fire recovery were quantified for 47 fires, with a new recovery metric facilitating comparison where fire frequency is high. Variables representing the current burn, fire history, vegetation structure, and topography were related to severity and yearly recovery with random forest and bivariate analysis. Results show that the 2019–2020 bushfires were the most widespread and severe, followed by 2007–2008. Vegetation recovers quickly, with NBR stabilizing ten years post-fire on average. Severity is most influenced by fire frequency, vegetation capacity and land use with more severe burns in nature conservation areas with dense vegetation and a history of frequent fires. Influence on recovery varied with time since fire, with initial (year 1–3) faster recovery observed in areas with less surviving vegetation. Later (year 6–10) recovery was most influenced by a variable representing burn year and further investigation indicates that precipitation increases in later post-fire years likely facilitated faster recovery. The relative abundance of eucalypt woodlands also has a positive influence on recovery in middle and later years. These results provide valuable information to land managers on Kangaroo Island and in similar environments, who should consider adjusting practices to limit future mega-fire risk and potential ecosystem shifts if severe fires become more frequent with climate change.