Explainable artificial intelligence (XAI) detects wildfire occurrence in the Mediterranean countries of Southern Europe.

The impacts and threats posed by wildfires are dramatically increasing due to climate change. In recent years, the wildfire community has attempted to estimate wildfire occurrence with machine learning models. However, to fully exploit the potential of these models, it is of paramount importance to make their predictions interpretable and intelligible. This study is a first attempt to provide an eXplainable artificial intelligence (XAI) framework for estimating wildfire occurrence using a Random Forest model with Shapley values for interpretation. Our findings accurately detected regions with a high presence of wildfires (area under the curve 81.3%) and outlined the drivers empowering occurrence, such as the Fire Weather Index and Normalized Difference Vegetation Index. Furthermore, our analysis suggests the presence of anomalous hotspots. In contexts where human and natural spheres constantly intermingle and interact, the XAI framework, suitably integrated into decision support systems, could support forest managers to prevent and mitigate future wildfire disasters and develop strategies for effective fire management, response, recovery, and resilience.

The wildland-urban interface map of Italy: A nationwide dataset for wildfire risk management.

A wildland-urban interface (WUI) raster map was created for the Italian peninsula with a resolution of 30 m per pixel. The map creation process consisted of three fundamental steps: (1) selection of buildings within the wildland-urban interface areas and subsequent classification of these into isolated, scattered, and clustered buildings; (2) creation of the tree canopy cover layer; (3) generation of WUI map by the intersection of two previous products. According to the WUI map, more than half of the total area of Italy is occupied by interface areas. Areas with buildings classified as clustered (24.61%) and scattered (19.15%) predominate on the territory compared to isolated buildings (14.93%). Most of the buildings are located in areas with a tree cover canopy between up to 64%. This map is functional to the implementation of forest fire prevention plans and to the identification of buildings that are close to fire risk areas such as forests, grasslands, and pastures.

Is Experience the Best Teacher? Knowledge, Perceptions, and Awareness of Wildfire Risk.

Wildfires represent a natural phenomenon with detrimental effects on natural resources and human health. A better knowledge, perception, and awareness of wildfire risk may help communities at risk of exposure to prevent future events and safeguard their own lives. The aim of this study is to explore differences between individuals with and without previous wildfire experience, in terms of (1) subjective and advanced wildfire knowledge, (2) self-reported perceptions, (3) level of information, (4) self-protection measures, and (5) importance of community involvement. As a second step, we investigated differences in the same variables, focusing more deeply on a group of individuals with previous wildfire experience, classifying them according to fire-related employment (fire-related workers vs. non-workers) and wildland-urban interface (WUI) proximity (WUI residents vs. non-WUI residents). The Kruskal-Wallis test was applied to establish differences between the pairs of subsamples. Our results partially confirmed our hypothesis, that direct experience leads individuals to have a greater preparedness on the topic of wildfires. Perception of knowledge is reflected only at a shallow level of expertise, and, therefore, no relevant within-group differences related to fire-related employment or to WUI proximity were detected. Moreover, available information was perceived to be insufficient, thus we report a strong need for developing effective communication to high-risk groups, such as homeowners and fire-related workers.

Association between indoor-outdoor green features and psychological health during the COVID-19 lockdown in Italy: A cross-sectional nationwide study.

Exposure to public green spaces was shown to be associated with psychological health. Nonetheless, evidence is lacking on the role of different green features within and/or surrounding the home environment when public green spaces are inaccessible or not usable. The overarching goal of this study is to shed light on the associations between the presence of greenness within the home and in the surrounding environment and the detrimental effects of quarantine on psychological health during the COVID-19 pandemic lockdown in Italy. A cross-sectional nationwide study involving an online survey was conducted of an Italian population-based sample of 3886 respondents on the association of indoor and outdoor green features (i.e., presence of plant pots, sunlight, green view and accessibility of private green space and natural outdoor environment) with self-reported increases in anxiety, anger, fear, confusion, moodiness, boredom, irritability, recurrent thoughts and/or dreams, poor concentration and sleep disturbance during the COVID-19 lockdown. Single-exposure regression models were performed to estimate associations between single green features and each psychological health outcome adjusted for relevant covariates. In the adjusted models, the presence of plant pots at home was associated with a lower self-reported increase in anxiety, anger, fear, irritability, and sleep disturbance. A greater amount of sunlight in the home was associated with a lower increase in anger, fear, confusion, moodiness, boredom, irritability, poor concentration, and sleep disturbance. A greater amount of green view and access to private green spaces were both associated with a lower increase in each of the psychological health outcomes except for green view and recurrent thought and/or dreams. Natural outdoor environment was associated with anxiety, fear, boredom, irritability, and sleep disturbance. Significant associations remained robust when adjusted for number of confirmed COVID-19 cases. Insights on future investigations are provided.

Green spaces, quality of life, and citizen perception in European cities.

Urban green spaces generate a number of perceived benefits toward human health and well-being, including an overall improvement in the quality of life. To date, processes underlying these benefits are mostly investigated at city level, while very little research has been conducted at a larger scale, such as the European level. Evidence is lacking on the association between green spaces and citizens' perceived benefits based on variations among cities with different socio-economic conditions. This study aims to disentangle the relationship between greenness, citizens' perceived quality of life, the environment, social inclusion and urban management in 51 European cities, as well as the role of per capita income in moderating the effects of greenness on overall perceived quality of life. Through structural equation modeling, the effect of greenness was tested on citizens' perceived overall quality of life (i.e., on the environment, social inclusion, and urban management). The role of metropolitan gross domestic product per capita in moderating the relationship between greenness and citizens' perceptions on overall quality of life was investigated. Our findings confirm the influence of greenness on citizens' quality of life. More importantly, the influence of greenness on the overall perception of quality of life was considerably more marked in lower-income cities. This study represents one of the first attempts at exploring the complex mechanisms underlying the association between green space and citizens' well-being at continent level. Practical implications for European urban planners and policy makers are discussed.

Are Community Gardening and Horticultural Interventions Beneficial for Psychosocial Well-Being? A Meta-Analysis.

Recent literature has revealed the positive effect of gardening on human health; however, empirical evidence on the effects of gardening-based programs on psychosocial well-being is scant. This meta-analysis aims to examine the scientific literature on the effect of community gardening or horticultural interventions on a variety of outcomes related to psychosocial well-being, such as social cohesion, networking, social support, and trust. From 383 bibliographic records retrieved (from 1975 to 2019), seven studies with a total of 22 effect sizes were selected on the basis of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. Meta-analytic findings on 11 comparisons indicate a positive and moderate effect of horticultural or gardening interventions on psychosocial well-being. Moderation analysis shows a greater effect size in individualistic than collectivistic cultures. A greater effect size was also observed in studies involving community gardening compared to horticultural intervention. Nevertheless, an effect of publication bias and study heterogeneity has been detected. Despite the presence of a large number of qualitative studies on the effect of horticulture/gardening on psychosocial well-being, quantitative studies are lacking. There is a strong need to advance into further high-quality studies on this research topic given that gardening has promising applied implications for human health, the community, and sustainable city management.