Landscape metrics are essential for understanding the spatial patterns associated with land-use changes. This study investigates the correlation among landscape metrics calculated for the native vegetation class in the Paraíba Valley (São Paulo, Brazil), between 2000 and 2022. Class-level metrics were derived, and Pearson’s correlation analysis was applied to identify linear associations between different metrics. The results revealed a positive correlation between Total Edge and Number of Patches (r = 0.69), indicating that increased fragmentation leads to greater total edge length. In contrast, Fractal Dimension Index and Aggregation Index showed a negative correlation (r = –0.31), suggesting that more aggregated fragments tend to exhibit simpler shapes. These findings indicate a spatial reorganization of native vegetation in the study area, driven by secondary regeneration and conservation policies, and underscore the relevance of landscape metrics in environmental analysis and planning.

The operational phase of an airport causes several environmental impacts, especially related to air quality and noise pollution. These effects can reach areas several kilometers away from the site, depending on factors such as weather conditions, topography, the number and frequency of flights, and the type of emission sources, both fixed and mobile, within the airport facilities (ICAO, 2023).

This research focuses on analyzing the technical and digital tools available to assess the environmental impact of airport activities, using the Felipe Ángeles International Airport (AIFA) as a case study.

According to the Airport Carbon Accreditation (ACA) program from the Airports Council International (ACI Europe, 2025), the main sources of emissions in airport operations include both direct emissions—such as energy use in buildings and equipment—and indirect emissions from aircraft operations (takeoff, taxiing, climb, approach, and landing), ground vehicle traffic, cargo activities, ground support equipment (GSE), contracted services, and waste and water management. These sources release pollutants like nitrogen oxides (NOx), sulfur oxides (SOx), particulate matter (PM), carbon monoxide (CO), and carbon dioxide (CO₂), and they also significantly contribute to noise pollution in nearby areas (ICAO, 2022).

The methodology involved a technical review of literature, regulations, and case studies from both national and international airports. Different tools were identified for evaluating environmental impacts. Among the qualitative methods were checklists, cause-effect diagrams, decision trees, and impact networks. Semi-quantitative and quantitative tools included matrices such as Leopold (Ponce, n.d.), Key Sensitivity Indicators Matrix (KSIM), Geographic Sensitivity Indicators Matrix (GSIM), and the Integrated System for Environmental Impact Review and Regulation (IIASA, 2008), which help prioritize impacts based on their magnitude and importance. Satellite images, aerial photographs, and field inspections were also considered.

A key tool identified in recent studies was the Aviation Environmental Design Tool (AEDT), developed by the Federal Aviation Administration (FAA, 2025). AEDT allows users to calculate air emissions and simulate the spread of pollutants and noise levels using real operational data, aircraft configurations, and weather conditions. Its results can be integrated into a Geographic Information System (GIS), helping visualize the environmental impact across space and overlay it with layers such as infrastructure, land use, vegetation, populated areas, and transportation routes (Esri, n.d.).

For this study, the GIS analysis included geospatial data from AIFA, environmental conditions, land use, vegetation, and population distribution within a 15-kilometer radius. This integration helped identify cumulative, synergistic, direct, and indirect impacts (Aspectum, n.d.).

Therefore, combining AEDT with GIS offers a powerful and specialized approach for environmental assessments in airports. It supports compliance with standards like Annex 16 of the International Civil Aviation Organization (ICAO, 2023) and contributes to the Sustainable Development Goals (ACI Europe, 2025) by providing a reliable technical basis for mitigation strategies, urban planning, land management, and environmental protection.

Additionally, guidance documents such as the ICAO Environmental Management Systems for Airports support the implementation of long-term sustainable practices in airport planning and operation (ICAO, n.d.-b). Also, tools like NOAA’s Environmental Sensitivity Index (ESI) maps provide complementary spatial information useful for airport-area assessments (NOAA, n.d.).

Referencias

ACI Europe, 2025. Airport Carbon Accreditation Program. Available at: https://www.airportcarbonaccreditation.org

Aspectum, n.d. GIS for Environmental Impact Analysis. Available at: https://aspectum.com/gis-for-environmental-impact-analysis/

Esri, n.d. Aviation Sustainability & Environment Using GIS. Available at: https://www.esri.com/en-us/industries/aviation/strategies/environmental

FAA, 2025. Aviation Environmental Design Tool (AEDT) Version 2d. Available at: https://aedt.faa.gov

ICAO, 2022. ICAO Environmental Report 2022, Annex 16 Vol. II. International Civil Aviation Organization.

ICAO, 2023. Annex 16: Environmental Protection – Volumes I-III. International Civil Aviation Organization.

ICAO, n.d.-b. Environmental Management Systems for Airports. Available at: https://www.icao.int/environmental-protection/Documents/EMS_at_Airports.pdf

IIASA, 2008. Adaptive Environmental Assessment and Management. International Institute for Applied Systems Analysis. Available at: https://pure.iiasa.ac.at/id/eprint/823/

NOAA, n.d. Environmental Sensitivity Index (ESI) Maps and Data. Available at: https://response.restoration.noaa.gov

Ponce, V.M., n.d. The Leopold Matrix. Available at: https://ponce.sdsu.edu/the_leopold_matrix.html

Wang, Y., de Vries, W.T., and Zhao, X., 2017. Mapping environmental sensitivity: A systematic online approach to support environmental assessments. Environ. Impact Assess. Rev., 67, pp. 1–12. https://doi.org/10.1016/j.eiar.2017.07.001

Airports Council International, 2025. Airports Responding to Climate Change. Available at: https://www.airportcarbonaccreditation.org

This study aimed to assess the potential for agricultural expansion into pasture areas with varying degrees of degradation in Brazil, providing both quantitative and spatial analyses at different scales – national, biome, microregion, and municipality. The proposed method was based on remote sensing data analysis and on data integration on a geographic information system to map degraded pasturelands, lands with natural agricultural suitability, infrastructure (roads and warehouses), croplands, climate risk agricultural zoning, and legal or environmental constraints posed by indigenous territories, Quilombola communities, agrarian settlements, and conservation units. Results indicate that approximately 28 million hectares of moderately to severely degraded pastures present good to high potential for conversion into croplands, considering the base year of 2022. The Cerrado biome demonstrated the highest potential, accounting for about 14.5 million hectares. The ten municipalities selected by the Digital Agriculture Development Science Center (Semear Digital) initiative and their corresponding microregions presented a total of 34.1 thousand hectares and 266.5 thousand hectares, respectively. The present approach enabled the quantification and spatialization of agricultural potential through a multiscale approach, providing support for decision-making in the formulation of public policies aimed at the restoration of degraded areas.

Urban sprawl constitutes a significant challenge to sustainable development, particularly in rapidly expanding urban environments. In the case of Guayaquil, in Ecuador, the construction of Perimetral Avenue in 1987 facilitated uncontrolled urban transformation. The absence of effective planning and oversight has led to the conversion of mangrove ecosystems into urban areas in specific sectors, with Isla Trinitaria notably experiencing the most substantial transformation, without yet having a clear strategy or methodology for responding to the problem. Within this framework, the objective of the study is to propose a methodology utilizing free software for optimizing resources in territorial planning. This investigation employed data from MapBiomas and the municipal urban cadastre; spatial information was integrated through free software platforms such as QGIS and PostgreSQL/PostGIS. One of the stages of the methodology included the vectorization of raster images, conducting spatial intersections with the cadastral data, and establishing a geographic database to enable SQL queries. The results indicated urban expansion from 2.1% in 1985 to full occupation in 2010, with a peak in construction activity in 1997. Furthermore, limitations such as low image resolution and data absence due to cloud cover were identified. This proposal aims to enhance sustainable urban planning by leveraging accessible tools, thereby contributing to effective territorial planning and informed decision-making in urban contexts.

Urban expansion is a global phenomenon that poses significant challenges for territorial planning and environmental sustainability. In Guayaquil, Ecuador's largest and most populous city, accelerated urban growth has been recorded, particularly in peri-urban areas. This study analyzes the transformation of Guayaquil's urban boundaries between 2016 and 2024, aiming to identify peri-urban zones and their basic infrastructure using Sentinel-2 satellite imagery and Geographic Information Systems (GIS). For this purpose, the analysis integrates Sentinel-2 satellite imagery from 2016, 2020, and 2024 with Geographic Information Systems (GIS) and official data on electricity, potable water, and sanitary sewer systems. The spatial assessment focused on areas beyond the official urban perimeter, highlighting Monte Sinaí as a significant growth zone. Results show an increase of 697.03 hectares of urbanized land during the study period, mainly concentrated in Monte Sinaí, Villa Bonita, and Costa Sol. However, 57.63% of properties in Monte Sinaí lack access to basic services such as electricity, drinking water, or sewage systems, indicating shortcomings in planning and infrastructure provision. The research demonstrates that urban expansion in Guayaquil is heterogeneous and unregulated, with a pronounced gap in basic service coverage. This research contributes meaningful insights that may be used to implement sustainable and equitable strategies for urban development.

Accurate Land Use and Land Cover (LULC) maps support the analysis of land use dynamics and provide a scientific basis for land management around water reservoirs. In recent years, technological advancements have led to significant improvements in the methodological design of these mappings. In the context of large-scale Earth Observation (EO) data, satellite image time series (SITS) represent a powerful approach for capturing and measuring surface changes. This study aimed to map LULC in a contributing watershed of the Batalha Hydroelectric Power Plant reservoir, located in eastern Goias, Brazil. We used the Sentinel-2/MSI time series from Brazil Data Cube (BDC) for the agricultural year (July 2022 to June 2023). The classification included

ten classes: Forest Formations, Savanna Formations, Grass Formations, Silviculture, Pasture, Single-Cycle Agricultural Crops, Multi-Cycle Agricultural Crops, Edification, Seasonally Flooded Areas, and Water. We used the Random Forest algorithm and the

best practices for assessing mapping accuracy. The results demonstrate the potential of SITS for mapping LULC conditions around water reservoirs. The LULC map generated provides valuable information for managing land use around the reservoir, with a focus

on areas where seasonal slope exposure occurs along the immediate reservoir edge. Mapping these areas represents a significant gain in information, as it contributes to the monitoring of sites susceptible to mass movements and marginal erosion processes.

1. Introduction

Environmental law enforcement agencies face the challenge of acting effectively and in a timely manner to curb actions that violate current legislation. Regarding wildfires and agricultural burning, remote sensing data enables enforcement agents to detect fire hotspots, helping prevent fire spread and allowing the delineation of burned areas for post-incident accountability.

Beyond large-scale fires, in the state of Mato Grosso do Sul, Brazil, some landowners employ fire to eliminate residual vegetation left from deforestation or for pasture clearing. In these cases, biomass is arranged into small piles or linear windrows for burning and later incorporated into the soil. However, Article 58 of Federal Decree No. 6.514/2008 classifies as an environmental infraction the “use fire in agro-pastoral areas without prior authorization from the competent authority or in contravention of the authorization granted”, punishable by a fine of BRL 3,000 per hectare or fraction thereof if no permit has been issued.

As the state-level agency responsible for environmental enforcement and management, the Environment Institute of Mato Grosso do Sul (IMASUL, in Portuguese) is tasked with identifying unauthorized burnings, including windrows, piles, native pasture, and wildfires, and take appropriate administrative sanctions against violators. To achieve this, continuous monitoring of the entire 357,125 km² territory of Mato Grosso do Sul is required—an effort made feasible only through remote sensing data. However, the current tools available pose limitations that challenge law enforcement, such as:

• Spatial resolution: Sentinel-2 imagery is commonly used for burn monitoring due to its free access, 5-day revisit time, and shortwave infrared bands ideal for fire detection. Yet, with a 10-meter pixel resolution, it fails to capture the smaller-scale piles and windrows used in these burnings.

• Temporal resolution: The Visible Infrared Imaging Radiometer Suite (VIIRS), with a spatial resolution of 375 meters, is typically used for active fire detection. However, its 12-hour revisit cycle results in significant temporal gaps, given that windrow fires can last from just a few minutes to a few hours, thus often evading detection. Furthermore, remnants of burning may be quickly incorporated into the soil, requiring prompt enforcement action before all physical evidence is lost.

To address these constraints, IMASUL’s Geoprocessing Unit (UNIGEO) has developed an automated enforcement system that integrates remote sensing techniques using Geostationary Operational Environmental Satellites (GOES) for fire detection and PlanetScope’s very high spatial resolution daily imagery for delineating burned areas.

2. Workflow

The GOES-R Series Advanced Baseline Imager (ABI) generates a Fire/Hot Spot Characterization (FDC) product that assigns a fire mask to each pixel, identifying fire categories. Although the final product has a spatial resolution of 2 km, it is produced every 10 minutes for the Americas, yielding 144 scenes per day with less than 30 minutes latency—allowing near-real-time detection.

Using a Python script, daily layers of fire pixels are generated. These pixels undergo spatiotemporal aggregation during fire events to extract information such as the date and time of first detection and the number of scenes the fire mask was assigned to each pixel. The data is then consolidated into a single monthly vector layer.

Each fire event is then analysed using PlanetScope imagery (~3.7 m resolution), provided via a partnership with the Federal Police (Brazil+ Program). Imagery from before and after the fire event is manually interpreted to delineate the burned area. In cases involving dead vegetation remnants, the linear or dotted patterns characteristic of windrows or piles can be visually distinguished (Figure 1). Once vectorised, A Fire Alert Notification (FAN) is generated for each event, containing fire data and property identification based on Brazil’s Rural Environmental Registry (CAR, in Portuguese), which holds data on all rural properties.

Each FAN undergoes a second technical review to check for valid burn authorizations issued by IMASUL. In the absence of such authorization, the burned area is validated and an administrative violation notice is issued, holding the landholder legally accountable, especially when there is evidence of a causal link—i.e., when the burning clearly aligns with common agro-pastoral practices.

3. Limitations and future perspectives

Preliminary results indicate the proposed workflow is effective for detecting and enforcing regulations against windrow burning, despite the coarse spatial resolution of GOES FDC (2 km). However, some limitations remain:

• False positives: During the evaluation of GOES FDC fire events, false positives were observed, such as fire masks assigned to areas not validated via PlanetScope imagery (Figure 2), or anomalies caused by sunglint in solar farms or water bodies.

• Cloud cover and imagery availability: Despite PlanetScope’s near-daily coverage, dense clouds or thick smoke columns can block optical sensors, impairing fire detection and burned area mapping—specially during the rainy season or intense fire episodes. Given the short duration of windrow burning, even a few consecutive cloudy days can result in missed detections if the post-burn residues are rapidly incorporated into the soil.

• Lack of omission error quantification: Since the workflow begins with GOES FDC data, fire events not detected by this product are not analysed by technicians, making it difficult to quantify undetected windrow burning events.

The next phase involves ground-truth validation through on-site inspections by environmental agents. This step is crucial not only for improving enforcement effectiveness, but also to reinforce the credibility of public institutions responsible for protecting natural resources. Establishing a robust environmental monitoring system directly contributes to law enforcement and fosters accountability, promoting sustainable land use and aligning environmental protection with socioeconomic development.

References

Brasil, Decreto nº 6.514, de 22 de julho de 2008. Dispõe sobre as infrações e sanções administrativas ao meio ambiente, Diário Oficial da União, Brasília, DF, 23 jul. 2008. [Online]. Available: https://www.planalto.gov.br/ccivil_03/_ato2007-2010/2008/decreto/d6514.htm.

European Space Agency (ESA), Sentinel-2 User Handbook, ESA Standard Document, 2021. [Online]. Available: https://sentinel.esa.int/web/sentinel/user-guides/sentinel-2-msi.

GOES-R Algorithm Working Group and GOES-R Program Office, (2018): NOAA GOES-R Series Advanced Baseline Imager (ABI) Level 2 Fire/Hot Spot Characterization (FDC). [Data set]. NOAA National Centers for Environmental Information. doi:10.7289/V5X065CR [accessed 05 Jun 2025].

NASA/NOAA, Visible Infrared Imaging Radiometer Suite (VIIRS) Data, NASA Earth Observing System Data and Information System (EOSDIS), 2023. [Online]. Available: https://earthdata.nasa.gov/viirs.

Planet Team, Planet Application Program Interface: In Space for Life on Earth. San Francisco, CA: Planet Labs PBC, 2023. [Online]. Available: https://www.planet.com.

This work demonstrates that the machine learning application can be used in the prediction of the Air Quality Index (AQI) at Dome A, East Antarctica. The daily meteorological parameters for March 2025, including rainfall, temperatures, wind direction, wind speed, humidity, pressure and solar irradiance data were obtained from the NASA POWER database. K-means clustering was applied in segmentation according to similar meteorological patterns. Using an empirical AQI estimation equation based on weather variables, initial AQI values were simulated. A Random Forest regression model was trained on this estimated AQI and the meteorological input. The model was evaluated with five-fold cross-validation, yielding good results with a root mean square error (RMSE) of about 2.4 AQI units averaged across folds, and an RMSE of about 1.3 and the coefficient of determination (R²) was high near 0.81 on separated test data. As shown in Fig. 3, that the predicted AQI values follows the simulated AQI values within the "Good" air quality category. The precipitation and solar irradiance came out as key factors (refer Fig. 5). These results underline that even in remote regions such as Antarctica, machine learning can effectively simulate AQI using meteorological data alone.

This work proposes a methodology for generating a high-quality dataset of samples to train machine learning models for land use and land cover (LULC) classification, representing an initial step toward the production of large-scale LULC maps. The novel approach is designed to generate a robust set of multi-pixel, multi-temporal, and multi-band samples by integrating label generation and remote sensing data processing across two main phases. In the first phase, multidimensional data are collected from remote sensing sources, while label information is derived by comparing harmonized LULC maps, along with additional processing to enhance data quality. In the second phase, samples are extracted to create a diverse, representative, and balanced set of reliably labeled multidimensional data for LULC classification. In this study, reference maps were obtained from the TerraClass and MapBiomas projects, and satellite imagery was sourced from the Brazil Data Cube platform.

Deforestation poses a significant threat to natural ecosystems, particularly in Argentina’s Chaco region—one of the world’s most rapidly changing forest areas. This study focuses on the detection of sudden deforestation events, where forest cover is rapidly removed within a few months. Monitoring such changes across vast areas requires the use of satellite-based vegetation indices, such as EVI and NDVI from MODIS. However, accurately identifying deforestation events is challenging due to seasonal variability, sensor noise, data gaps, and algorithmic inconsistencies. These factors can obscure true deforestation signals or generate false positives. To address these issues, a robust detection approach must explicitly model time-series dynamics—capturing trends, seasonality, and uncertainty—to reliably distinguish genuine deforestation breakpoints from natural variation and noise. In this paper, three models for the detection of breakpoints in EVI time series were proposed: a simple z-score anomaly detector, and two fully Bayesian models; one temporally uncorrelated and one fully correlated. Results indicate that the Bayesian schemes significantly improve over the naive approach (zscore: AUC=0.921, F1-score=0.870, Bayes: AUC=0.959, F1-score=0.925), for a reasonable cost in computing time (x1000).

Monitoring land use and land cover is critically important for the development of environmental, social, and economic policies. These analyses not only provide information about ongoing changes but also allow us to relate them to anthropogenic factors or climate change. In recent years, access to large volumes of freely available satellite imagery, along with platforms designed for processing this type of data, has enabled a better understanding of land cover changes and the factors that drive them. In this study, we propose the use of a machine learning model to classify land cover in preservation areas during the period from 2019 to 2025. We then analyze the results of accumulated precipitation data and in situ water quality measurements. The results show that it is possible to achieve land cover classification with an accuracy of 97%. In addition, we analyzed the relationship between precipitation levels and the extent of surface water within the reservoir. The results indicate that fluctuations in surface water are consistent with the classifications derived from the machine learning model and correspond to the accumulated rainfall during the classified periods. Furthermore, we examined the relationship between areas identified as urban and non-urban and the in-situ water quality measurements. The integration of satellite imagery, meteorological data, and in situ measurements provides a robust framework for the interdisciplinary analysis of environmental dynamics and interactions.

The application of geotechnology for accurately estimating peanut maturity is essential for enhancing crop management, facilitating monitoring, and understanding spatial variability. This study aimed to identify and select the most relevant variables for maturity estimation and to evaluate the performance of a multiple linear regression (MLR) model using those variables. The research was conducted in a commercial field during the 2022/2023 season with the IAC 503 cultivar, which has a 150-day growth cycle. Forty sampling points were evaluated across five dates, beginning 28 days before harvest and ending one day prior to harvest. Maturity was assessed using the Hull-Scrape method. Satellite imagery was obtained from the PlanetScope platform, offering 3-meter spatial resolution and daily temporal coverage. Five cloud-free images corresponding to the sampling dates were selected. Dimensionality reduction was performed using principal component analysis (PCA), followed by stepwise regression to identify the most influential variables. The MLR model, despite its simplicity, achieved high performance with an R² of 0.93. The most relevant variables identified were growing degree days (GDD), green band reflectance, NDVI, and SAVI. These results demonstrate that a small set of well-selected variables is sufficient for accurately estimating peanut pod maturity.

Wildfire mapping in remote and ecologically sensitive regions like the Brazilian Pantanal faces challenges due to the high cost of collecting pixel-level annotations required for fully supervised models. In this study, we investigate the use of Weakly Supervised Semantic Segmentation (WSSS) methods—specifically SEAM and Puzzle-CAM—for burned area mapping using multispectral (RGB-NIR) satellite imagery. Both models were adapted to handle four-band data to leverage spectral information relevant for fire detection. Our two-stage pipeline first generates pseudo-labels from image-level annotations and then trains a SegFormer segmentation model on these labels. Experimental results show that Puzzle-CAM, particularly when combined with a stronger ResNeSt-101 backbone, produces high-quality pseudo-labels, leading to segmentation results that closely approach those of fully supervised methods. This approach demonstrates the potential of combining weak supervision and advanced network architectures to reduce labeling costs while enabling scalable wildfire monitoring across the Pantanal. Future work will focus on improving model robustness and extending the methodology to other types of ecological disturbances.

Water quality is essential for sustainable development, especially in Peru, where water scarcity and unequal distribution impact ecosystems and communities. High-altitude Andean reservoirs, such as El Pañe in the Arequipa region, are vital water sources but are increasingly threatened by eutrophication due to nutrient overloading, limited watershed management, and climate variability. Algal blooms caused by eutrophication reduce water quality and pose risks to public health, agriculture, and aquatic ecosystems. However, severe climate conditions and difficult accessibility limit the continued monitoring that is required to control eutrophication events in these environments.

This study presents a predictive model to identify and anticipate eutrophication events in Andean reservoirs of Peru by integrating climatic parameters with satellite spectral data. A dataset was constructed using Normalized Difference Chlorophyll Index (NDCI) for the El Pañe reservoir, climatic data (temperature, precipitation, radiation, wind speed), soil moisture, nitrogen oxide concentrations, and estimated sediment loss. A Random Forest model was trained to predict chlorophyll concentration and contrasted to chlorophyll estimations by an empirical non-linear model based on NDCI values. The model was able to estimate chlorophyll-a concentrations but struggled to adequately predict high chlorophyll-a values. Alternatively, a classification model that groups chlorophyll-a concentrations based on the Carlson’s Trophic State Index (TSI) was also explored. The binary classification model (eutrophic vs. non-eutrophic) achieved 85.9% accuracy and an AUC of 0.92, demonstrating strong performance in detecting algal blooms. Multi-class classification of five trophic states also showed satisfactory results, with AUC values ranging from 0.72 to 0.82. Furthermore, Principal Component Analysis (PCA) was used to assess relative importance of variables used in the Random Forest model. It revealed that radiation, wind speed, and soil moisture were the most influential factors affecting chlorophyll concentrations. Overall, these findings offer valuable insights into the drivers of eutrophication and provide a cost-effective, scalable approach for monitoring water quality in remote Andean reservoirs. The proposed models support data-driven decision-making for water resource management and early intervention in vulnerable ecosystems.

High-resolution urban tree-canopy maps are vital for smart-city planning and heat-island mitigation yet remain scarce in Brazil. This work presents a deep-learning workflow that produces such maps from 25 cm RGB orthophotos. Images covering ten São Paulo cities were compiled; seven were used for training/validation and three for independent testing. The DeepLabV3 architecture with a ResNet-152 backbone was assessed under three loss configurations: (i) Balanced Cross-Entropy (BCE) baseline, (ii) BCE plus PointRend boundary refinement, and (iii) BCE combined with a 0.5-weighted Dice term. The BCE baseline delivered the top mean IoU (0.83) and F-score (0.91). PointRend increased recall but introduced systematic false positives in heterogeneous roofs and shaded riparian zones. The BCE+Dice variant recovered recall without raising commission error, achieving the highest balanced accuracy (0.96). The workflow delineates canopy with fine spatial detail and processes 2.8 × 10⁶ m² in under 30 minutes on a single RTX 4000 Ada workstation, demonstrating a practical, scalable solution for statewide tree-inventory production.

This study investigates the impact of light-absorbing impurities (LAIs) from wildfire ashes on seasonal snow albedo in the Patagonian Andes near San Carlos de Bariloche, Argentina. A multi-faceted approach was applied, integrating satellite remote sensing, and radiative transfer simulations to assess albedo reduction and spectral modifications. The methods employed include calculating percentage variation, visible range slope, normalized indices, and polynomial coefficients to quantify albedo change and impurities’ effects on reflectance. The results show a marked drop in albedo after the November 2, 2022, wildfire, particularly in visible and near infrared bands, reflecting the deposition of ashes and black carbon on the snow surface. The combination of multi-source data and methods underscores the utility of satellite observation to track and quantify the effects of such events on snow-covered regions.

This study investigates the use of contact spectroradiometry for the discrimination of grapevine varieties cultivated in the Rio Grande do Sul region, Brazil, focusing on Vitis vinifera and Vitis labrusca. Spectral data were collected from 574 leaf samples using a FieldSpec® 3 ASD spectroradiometer. Analytical methods included spectral ratio computation, first derivative analysis, standard deviation-based variability assessment, and Linear Discriminant Analysis (LDA) with Forward Stepwise variable selection. The results revealed that the most relevant spectral regions for varietal discrimination are concentrated in the visible and near-infrared (green and red-edge) and shortwave infrared (SWIR) ranges, reflecting leaf biochemical and structural properties. A set of ten key wavelengths achieved high classification performance, with the final LDA model reaching an average accuracy of approximately 93%. These findings highlight the potential of contact spectroradiometry as a tool for varietal mapping, traceability, and supporting certification of origin in viticulture.

Forest biomass assessment is a critical element influencing the decisions of stakeholders involved in forest management. In tropical and subtropical biodiversity hotspots, accurate measurement of aboveground biomass (AGB) is crucial for ecosystem sustainability. However, estimating AGB in these forests is challenging due to complex vegetation, necessitating data integration from various sources. Methods for mapping AGB in forest ecosystems have progressed from a destructive sampling approach to non-destructive sampling techniques that utilize allometric models and remotely sensed vegetation variables (Muhe and Argaw, 2022). Remote sensing platforms, including active and optical sensors, have proven to be effective alternatives for assessing and monitoring AGB across different landscapes and scales (Su et al., 2016). Among the optical sensors, Sentinel-2 with its enhanced spectral band and improved spatial resolution of 10-20 meters, is considered to provide superior opportunities for estimating AGB in subtropical and tropical areas compared to other low-resolution satellite images (Pandit et al., 2018). However, optical sensors face challenges in providing structural information due to their inability to penetrate deeply into the upper canopy layers and their susceptibility to being obstructed by frequent cloud cover. (Hyde et al., 2006).

Synthetic Aperture Radar (SAR) provides an innovative approach to minimize the challenges in optical remote sensing. The sensitivity of SAR sensors to forest biomass is dependent on its wavelength. Generally, longer wavelengths penetrate deeper into the forest and are better suited for AGB estimation as they can reach the canopy and interact directly with the trunk (Ouchi, 2013). However, using longer wavelength data can be challenging due to the limited number of operational satellites and the high costs associated with such data. While the Sentinel-1 mission from the European Space Agency (ESA) provides free high-resolution C-band SAR data globally, its limited penetration capability makes it insufficient for forest biomass estimation on its own (Ghosh and Behera, 2018). As a result, researchers commonly combine SAR with optical remote sensing data to achieve more accurate and comprehensive forest biomass measurements (Moghimi et al., 2023).

In this research, we investigated the feasibility of integrating ground-based measurements with multi-sensor SAR and optical remote sensing data to analyze the spatio-temporal changes in biomass within the subtropical forest of Hong Kong, and evaluated our AGB maps through uncertainty analysis, offering insights into the effectiveness of biomass estimation. Random forest (RF) machine learning model was utilized for the AGB mapping. Correlation analysis was conducted to examine the relationship between the measured AGB and radar backscatter polarimetric data, optical reflectance bands, and spectral indices. Recursive Feature Elimination with 5-fold cross-validation (RFECV) was employed to eliminate redundant information in satellite images. Additionally, the gray level co-occurrence matrix (GLCM) method of texture analysis, was employed to extract various texture variables to enhance the discrimination of spatial information regardless of tone whiles minimizing forest structural variations that are not related to biomass.

From the results, Random Forest (RF) model demonstrated enhanced accuracy across all evaluated datasets. It achieved the highest performance for the 2023 dataset (R² = 0.907, RMSE = 20.479 tons/ha, RRMSE = 0.111 %, MAE = 14.027 tons/ha), followed by the 2022 dataset (R² = 0.899, RMSE = 21.319 tons/ha, RRMSE = 0.115 %, MAE = 16.384), and the 2024 dataset (R² = 0.860, RMSE = 25.054 tons/ha, RRMSE = 0.136 %, MAE = 18.216 tons/ha). Additionally, RF modeling approach demonstrated fewer deviations with the residuals exhibiting less variability in the AGB predictions. Further analysis identified specific vegetation indices and ALOS-2 PALSAR-2 backscatter combinations such as (HH+HV)_Cor, GNDVI, and NDI45, as influential predictors across all AGB ranges. In contrast, Sentinel-1 radar backscatter predictors demonstrated a weaker impact on biomass prediction. Finally, the regional distribution map of forest biomass generated via RF model revealed that district with significant biomass changes include Tai Po, Islands, Sai Kung, Sha Tin and Yuen Long. In these districts, significant spatial heterogeneity was observed. This research underscores the potential of machine learning approaches in conjunction with satellite remote sensing for mapping the spatio-temporal changes in biomass, offering valuable insights for forest management and conservation efforts nationwide. The findings contribute to the growing field of remote sensing applications in ecological studies and highlight the importance of selecting appropriate predictors for improving model accuracy.

In this paper it is presented the results of a study aimed at analyzing the evolution of land cover—particularly vegetation—using remote sensing time series. The focus is on monitoring vegetation changes in the Passaúna basin, an important water supply source for Curitiba, Brazil. Vegetation in this basin plays a key role in ensuring both the quantity and quality of water available to the population. The study employed an unsupervised classification scheme based on the Normalized Difference Vegetation Index (NDVI) and a binary encoding approach. It is used a hybrid method, combining classification results from multiple dates. The core of the methodology involves deriving indicators of seasonal or annual pixel variation by analyzing several images from the same year, and then comparing these indicators across different years. This approach enhances the ability to detect seasonal land cover variations, which improves the identification of land cover classes. Using multiple observations per year proved especially effective in distinguishing vegetation types. The analysis aimed at detecting significant land cover changes, with an emphasis on vegetation loss and recovery. The binary encoding technique facilitated the mapping of land cover evolution, particularly changes associated with the filling of the Passaúna reservoir, and helped pinpoint their locations. A key advantage of this method is that it does not require training sample selection to classify the data. Because NDVI is a normalized index, it was used variation ranges to separate certain land cover classes at each time point. The potential for accurate discrimination increased, combining multiple dates within a year, thanks to the integration of seasonal dynamics. From a hydrological perspective, the land cover changes observed between 1988 and 2018 were substantial, though the system has since shown signs of stabilization. The creation of the reservoir led to changes such as the emergence of new agricultural areas around the water body. At the same time, denser vegetation increased in the upper basin. These changes significantly affect infiltration rates and potential surface runoff, highlighting the hydrological impact of land cover dynamics in the region.

Methane is a potent greenhouse gas with significant implications for climate change. This study investigates high-emission point sources, also known as methane super-emitters, in landfills across the Metropolitan Area of São Paulo, using data from two remote sensing platforms: EMIT and Carbon Mapper. EMIT, aboard the International Space Station, and Carbon Mapper, operating through satellite and airborne sensors, provided complementary datasets that enabled detection and analysis of methane plumes in Caieiras, Guarulhos, and Osasco. Emission rates ranged from 964 to 4,562 kilograms of methane per hour, confirming these landfills as significant sources of anthropogenic methane. The study demonstrates the effectiveness of combining spectral imaging platforms for monitoring urban greenhouse gas emissions and supports the development of mitigation strategies aimed at reducing the environmental impact of the waste sector.

Urban tree vegetation plays a key role in sustainable urban planning and ecosystem service provision. This study evaluates the performance of the Segment Anything Model (SAM), developed by Meta AI, in the segmentation of urban tree vegetation from orbital PlanetScope imagery. These images were selected due to their high spatial and temporal resolution, which makes them particularly suitable for urban applications. SAM was applied in zero-shot mode, guided by geometric prompts over representative tree-covered areas. The analysis was conducted across three Brazilian cities—Corumbá (MS), Rio Verde (GO), and Valparaíso de Goiás (GO)—using different spectral band compositions. SAM’s performance was evaluated through a combined quantitative and qualitative approach, using reference masks derived from manually annotated tree canopy polygons. Although SAM had not been previously trained on satellite imagery, it achieved an F1-scores close to 70% and recall values around 75%, independently of the spectral band composition provided as input. These results demonstrate the model’s generalization ability—even under spectrally constrained scenarios involving only three bands. Qualitative analysis confirmed spatial consistency in tree crown delineation, particularly in homogeneous areas, while over-segmentation was observed in spectrally heterogeneous environments. While the results are promising for exploratory and semi-automated vegetation mapping, they also underscore need for fine-tuning SAM on satellite data to enhance spatial precision and thematic discrimination. Overall, SAM's modular and prompt-based architecture offers a robust foundation for scalable, supervised remote sensing workflows focused on urban vegetation monitoring.

The Pantanal wetlands are distinguished by their ecological dynamics, driven by seasonal variations and flood-drought regimes. Located within the floodplain of the Northern Pantanal and influenced by the Cerrado biome, the Pirizal region features vegetation adapted to both fire and flooding, and is frequently impacted by extreme events such as wildfires. This study aimed to analyze the variation in phytophysiognomies in the region using multispectral imagery acquired from the PlanetScope (SuperDove) satellite for the years 2020 (wet and dry seasons) and 2024 (post-fire). The results indicated that, in March 2020, the predominant mapped classes were Campos de Murunduns, Cambarazal, and Campo de Mimoso. By October of the same year, wildfire effects were evident, with a significant reduction in Campos de Murunduns and the emergence of burned areas. Four years after the fire event (November 2024), significant vegetation recovery was observed, notably an increase in the proportion of Cambarazal and Campo de Mimoso, while Campos de Murunduns experienced a decline. It is concluded that, despite the impacts of the 2020 wildfires, the Pirizal vegetation exhibits resili-ence and regenerative capacity. Remote Sensing proved to be an effective tool for monitoring these dynamics and supporting conserva-tion efforts in the region.

This research examines the relationship between urban physical characteristics and land surface temperature within discrete thermal

pixels captured by Landsat 8 imagery in Guadalajara, Mexico. The city’s extensive collection of high-quality geospatial data enables

urban thermal analysis with high granularity. We integrate multiple datasets: a LiDAR-derived urban tree inventory with individual

tree metrics; building footprints with roof material classifications; green space polygons; transportation infrastructure including roads

and sidewalks; and water body delineations. Our methodology focuses on the pixel level—for each 30-meter thermal pixel (900

square meters), we precisely quantify all urban features within its boundaries, creating a comprehensive dataset where each pixel

contains measurements of all elements present. This spatial integration enables multivariate regression modeling using machine

learning, where predictor variables represent the quantity of each urban element and the target variable is pixel temperature. Using

interpretable machine learning techniques, we quantify each element’s influence on thermal patterns, achieving R² values ranging

from 0.69 to 0.83 across different urban contexts. This pixel-level approach provides granular understanding of urban thermal

dynamics, explaining which factors influence land surface temperature and specifying what characteristics new developments should

incorporate to achieve desired LST conditions, contributing to evidence-based urban planning.

Average rates of shoreline change are key indicators for assessing coastal evolution. The study area is located in

Salinópolis, on the northeastern coast of Pará, Brazil, covering urban, estuarine, and natural zones, including Corvina, Maçarico, Farol

Velho, and Atalaia beaches. Between 1984 and 2024, despite an overall trend of shoreline accretion, areas with growing human

occupation experienced significant coastal erosion, causing building retreat, partial home loss, and damage to beach access roads. Using the

Digital Shoreline Analysis System and dense satellite image time series processed in Google Earth Engine, the coastline was analyzed

in four sectors (A, B, C and D). The sector B with the greatest erosion showed an average linear retreat of –1076.31 m, while another sector

reached a maximum advance of +1079.78 m, indicating a spatially alternating pattern between erosion and accretion. The average linear

rate of change showed a slight overall shoreline retreat, accompanied by high morphodynamic variability and low statistical consistency in

linear trends. Urbanized sectors exposed to ocean forces were most vulnerable to erosion, whereas estuarine and mangrove areas were

more stable. The high sediment supply from estuaries contributed positively to shoreline accretion in several regions. Projections for

2034 and 2044 suggest the shoreline will mostly retain its current shape within uncertainty margins. However, sector C faces a potential

pronounced retreat, threatening frontal dunes and coastal infrastructure. These findings emphasize the importance of strategic

coastal management considering natural and human influences on shoreline dynamics.