RESUMO
Data from the Operational Land Imager (OLI) and the Thermal Infrared Sensor (TIRS) instruments onboard the Landsat 8 and Landsat 9 satellite platforms are subject to contamination by cloud cover, with cirrus contributions being the most difficult to detect and mask. To help address this issue, a cirrus detection channel (Band 9) centered within the 1.375-µm water vapor absorption region was implemented on OLI, with a spatial resolution of 30 m. However, this band has not yet been fully utilized in the Collection 2 Landsat 8/9 Level 2 surface temperature data products that are publicly released by U.S. Geological Survey (USGS). The temperature products are generated with a single-channel algorithm. During the surface temperature retrievals, the effects of absorption of infrared radiation originating from the warmer earth's surfaces by ice clouds, typically located in the upper portion of the troposphere and re-emitting at much lower temperatures (approximately 220 K), are not taken into consideration. Through an analysis of sample Level 1 TOA and Level 2 surface data products, we have found that thin cirrus cloud features present in the Level 1 1.375-µm band images are directly propagated down to the Level 2 surface data products. The surface temperature errors resulting from thin cirrus contamination can be 10 K or larger. Previously, we reported an empirical and effective technique for removing thin cirrus scattering effects in OLI images, making use of the correlations between the 1.375-µm band image and images of any other OLI bands located in the 0.4-2.5 µm solar spectral region. In this article, we describe a variation of this technique that can be applied to the thermal bands, using the correlations between the Level 1 1.375-µm band image and the 11-µm BT image for the effective removal of thin cirrus absorption effects. Our results from three data sets acquired over spatially uniform water surfaces and over non-uniform land/water boundary areas suggest that if the cirrus-removed TOA 11-µm band BT images are used for the retrieval of the Level 2 surface temperature (ST) data products, the errors resulting from thin cirrus contaminations in the products can be reduced to about 1 K for spatially diffused cirrus scenes.
RESUMO
This study assessed the accuracy of various methods for estimating lake evaporation in arid, high-wind environments, leveraging water temperature data from Landsat 8. The evaluation involved four estimation techniques: the FAO 56 radiation-based equation, the Schendel temperature-based equation, the Brockamp & Wenner mass transfer-based equation, and the VUV regression-based equation. The study focused on the Chah Nimeh Reservoirs (CNRs) in the arid region of Iran due to its distinctive wind patterns and dry climate. Our analysis revealed that the Split-window algorithm was the most precise for satellite-based water surface temperature measurement, with an R2 value of 0.86 and an RMSE of 1.61 °C. Among evaporation estimation methods, the FAO 56 stood out, demonstrating an R2 value of 0.76 and an RMSE of 4.36 mm/day in comparison to pan evaporation measurements. A subsequent sensitivity analysis using an artificial neural network (ANN) identified net radiation as the predominant factor influencing lake evaporation, especially during both wind and no-wind conditions. This research underscores the importance of incorporating net radiation, water surface temperature, and wind speed parameters in evaporation evaluations, providing pivotal insights for effective water management in arid, windy regions.
Assuntos
Lagos , Água , Temperatura , Redes Neurais de Computação , Clima DesérticoRESUMO
To identify areas of high biodiversity and prioritize conservation efforts, it is crucial to understand the drivers of species richness patterns and their scale dependence. While classified land cover products are commonly used to explain bird species richness, recent studies suggest that unclassified remote-sensed images can provide equally good or better results. In our study, we aimed to investigate whether unclassified multispectral data from Landsat 8 can replace image classification for bird diversity modeling. Moreover, we also tested the Spectral Variability Hypothesis. Using the Atlas of Breeding Birds in the Czech Republic 2014-2017, we modeled species richness at two spatial resolutions of approx. 131 km2 (large squares) and 8 km2 (small squares). As predictors of the richness, we assessed 1) classified land cover data (Corine Land Cover 2018 database), 2) spectral heterogeneity (computed in three ways) and landscape composition derived from unclassified remote-sensed reflectance and vegetation indices. Furthermore, we integrated information about the landscape types (expressed by the most prevalent land cover class) into models based on unclassified remote-sensed data to investigate whether the landscape type plays a role in explaining bird species richness. We found that unclassified remote-sensed data, particularly spectral heterogeneity metrics, were better predictors of bird species richness than classified land cover data. The best results were achieved by models that included interactions between the unclassified data and landscape types, indicating that relationships between bird diversity and spectral heterogeneity vary across landscape types. Our findings demonstrate that spectral heterogeneity derived from unclassified multispectral data is effective for assessing bird diversity across the Czech Republic. When explaining bird species richness, it is important to account for the type of landscape and carefully consider the significance of the chosen spatial scale.
RESUMO
Land surface temperatures (LST) are vital parameters in land surface-atmosphere interactions. Constrained by technology and atmospheric interferences, LST retrievals from various satellite sensors usually return missing data, thus negatively impacting analyses. Reconstructing missing data is important for acquiring gap-free datasets. However, the current reconstruction methods are limited for maintaining spatial details and high accuracies. We developed a new gap-free algorithm termed the spatial feature-considered random forest regression (SFRFR) model; it builds stable nonlinear relationships to connect the LST with related parameters, including terrain elements, land coverage types, spectral indexes, surface reflectance data, and the spatial feature of the LST, to reconstruct the missing LST data. The SFRFR model reconstructed gap-free LST data retrieved from the Landsat 8 satellite on 27 July 2017 in Wuhan. The results show that the SFRFR model exhibits the best performance according to the various evaluation metrics among the SFRFR, random forest regression and spline interpolation, with a coefficient of determination (R2) reaching 0.96, root-mean-square error (RMSE) of 0.55, and mean absolute error (MAE) of 0.55. Then, we reconstructed gap-free LST data gathered in Wuhan from 2016 to 2021 to analyze urban thermal environment changes and found that 2020 presented the coolest temperatures. The SFRFR model still displayed satisfactory results, with an average R2 of 0.91 and an MAE of 0.63. We further discuss and discover the factors affecting the visual performance of SFRFR and identify the research priority to circumvent these disadvantages. Overall, this study provides a simple, practical method for acquiring gap-free LST data to help us better understand the spatiotemporal LST variation process.
Assuntos
Algoritmos , Monitoramento Ambiental , Temperatura , Monitoramento Ambiental/métodos , Cidades , ChinaRESUMO
Ecological forests are an important part of terrestrial ecosystems, are an important carbon sink and play a pivotal role in the global carbon cycle. At present, the comprehensive utilization of optical and radar data has broad application prospects in forest parameter extraction and biomass estimation. In this study, tree and topographic data of 354 plots in key nature reserves of Liaoning Province were used for biomass analysis. Remote sensing parameters were extracted from Landsat 8 OLI and Sentinel-1A radar data. Based on the strong correlation factors obtained via Pearson correlation analysis, a linear model, BP neural network model and PSO neural network model were used to simulate the biomass of the study area. The advantages of the three models were compared and analyzed, and the optimal model was selected to invert the biomass of Liaoning province. The results showed that 44 factors were correlated with forest biomass (p < 0.05), and 21 factors were significantly correlated with forest biomass (p < 0.01). The comparison between the prediction results of the three models and the real results shows that the PSO-improved neural network simulation results are the best, and the coefficient of determination is 0.7657. Through analysis, it is found that there is a nonlinear relationship between actual biomass and remote sensing data. Particle swarm optimization (PSO) can effectively solve the problem of low accuracy in traditional BP neural network models while maintaining a good training speed. The improved particle swarm model has good accuracy and speed and has broad application prospects in forest biomass inversion.
RESUMO
The dominant wavelength and hue angle can be used to quantify the color of lake water. Understanding the water color is important because the color relates to the water quality and its related public perceptions. In this paper, we compared the accuracy levels of two methods in calculating dominant wavelength and hue angle values using simulated satellite data calculated from in situ reflectance hyperspectra for 325 lakes and rivers in Minnesota and Wisconsin. The methods developed by van der Woerd and Wernand in 2015 and Wang et al. in 2015 were applied to simulated sensor data from the Sentinel-2, Sentinel-3, and Landsat 8 satellites. Both methods performed comparably when a correction algorithm could be applied, but the correction method did not work well for the Wang method at hue angles < 75°, equivalent to levels of colored dissolved organic matter (CDOM, a440) > ~2 m-1 or chlorophyll > ~10 mg m-3. The Sentinel-3 spectral bands produced the most accurate results for the van der Woerd and Wernand method, while the Landsat 8 sensor produced the most accurate values for the Wang method. The distinct differences in the shapes of the reflectance hyperspectra were related to the dominant optical water quality constituents in the water bodies, and relationships were found between the dominant wavelength and four water quality parameters, namely the Secchi depth, CDOM, chlorophyll, and Forel-Ule color index.
RESUMO
In urban areas, industrial and human activities are the prime cause that exacerbates the heating effects, also called the urban heat island (UHI) effect. The land surface temperature (LST), normalized difference vegetation index (NDVI), and the proportion of vegetation (Pv) are indicators of measurement of the heating/urbanization effects. In the present work, we investigated the impact of the COVID-19 lockdown, i.e., restricted human activities. We used Landsat-8 OLI/TIRS (level 1) data to investigate spatial and temporal heterogeneity changes in these urbanization indicators during full and partial lockdown periods in 2020 and 2021, with 2019 as the base year. We have selected three cities in India's eastern coal mining belt, Bokaro, Dhanbad, and Ranchi, for the study. Results showed a significant decrease in LST values over all sites, with a maximum reduction over mining sites, i.e., Bokaro and Dhanbad. The LST value decreased by about 13-19% during the lockdown period. Vegetation indices (i.e., NDVI and Pv) showed a substantial increase of about 15% overall sites. With decreased LST values and increased NDVI values, these quantities' correlations became more negative during the lockdown period. More positive changes are noticed over mining sites than non/less mining sites. This indirectly indicates the reduction in the heat-absorbing particles in the environment and surface of these sites, a possible cause for the reduction in LST values substantially. Reduction in coal particles at the land and vegetation surface likely contributed to decreased LST and enhanced vegetation indices. To check the statistical significance of changes in the UHI indicators in the lockdown period, statistical tests (ANOVA and Tukey's test) are performed. Results indicate that most of the case changes have been significant. The study's finding suggests the lockdown's positive impact on the heating/UHI effects. It emphasizes the need for planned lockdowns as city mitigation strategies to overcome pollution and environmental issues.
Assuntos
COVID-19 , Temperatura Alta , Humanos , Temperatura , Cidades , Monitoramento Ambiental/métodos , COVID-19/epidemiologia , Controle de Doenças Transmissíveis , UrbanizaçãoRESUMO
The research proposes a model to estimate the carbon stock in mangrove forests from multispectral images from Landsat 8 and Sentinel 2B satellites. The Gramame River mangrove, located on the southern coast of Paraíba State, Brazil, was adopted as the study area. Carbon stocks in biomass, below and above ground, were measured from a forest inventory, and vegetation indices were processed on the Google Earth Engine (GEE) platform. To define the fit curves, linear and non-linear regressions were used. The choice of the model considered the highest coefficients of determination (R2), the biomass and carbon stock were estimated from the equations. The biomass carbon stock, calculated from field data, corresponded to 22.27 Gg C, equivalent to 81.75 Gg CO2, with 13.85 Gg C (50.84 Gg CO2) and 8.42 Gg C (30.91 Gg CO2) stored in biomass above and below ground, respectively. Among the models fitted to the indices calculated from Landsat 8 images, NDVI was the one that best explained the spatial distribution of biomass and carbon, with 90.26%. For Sentinel 2B, SAVI was able to explain 80.76%. The total estimated plant carbon stocks corresponded to 26.66 Gg (16.20 Gg C above and 10.36 Gg C below ground) for Landsat 8 and 27.76 Gg C (16.93 Gg C above and 10.83 Gg C below ground) for Sentinel 2B. The proposed work methodology and the suggested mathematical models can be replicated to analyze carbon stocks in other locations, especially in the Americas, because they share the same species.
Assuntos
Dióxido de Carbono , Carbono , Brasil , Carbono/análise , Monitoramento Ambiental/métodos , Florestas , BiomassaRESUMO
This study authorizes processes and approaches using optical and microwave data to determine the availability of water in the study area at any given moment. This will aid in identifying the optimal time and location for irrigation to enhance crop growth. For this purpose, a set of spectral vegetation parameters (from Sentinel-2), soil moisture (from Sentinel-1), evapotranspiration, and surface temperature (from Landsat-8) were used, along with field data on water content and irrigation timing. The results showed that both NDVI and NDMI are highly sensitive to moisture, making them the best indices for determining the timing and location of irrigation. This research contributes to sustainable agricultural development. It has implications for farmers, policymakers, and researchers in optimizing irrigation schedules, developing policies for sustainable agriculture, and enhancing crop productivity while conserving water resources. This approach can be particularly useful in regions facing water scarcity, where the efficient use of water resources is crucial for sustainable agricultural development.
Assuntos
Monitoramento Ambiental , Solo , Agricultura , Produção Agrícola , ÁguaRESUMO
Changes in land use due to urbanization, industrialization, and agriculture will adversely affect water quality at all scales. This study examined the possible effects of future land use on the water quality of the Dez River located in Iran. The QUAL2Kw dynamic model was used to simulate the water quality of the Dez River. Data and information available in July 2019 and 2013 were used for calibration and validation. According to the comparison of the RMSE, RMSE%, and percent bias error indices for the model during the calibration and validation period, the QUAL2Kw model of Dez River had high accuracy with acceptable values of errors. The land use changes in the Dez river basin were modeled and predicted by the LCM model after simulating water quality. The images from Landsat 8/OLI were used for 2013, 2016, and 2019, respectively. Based on the accurate evaluation of classified images, Kappa coefficients for 2013, 2016, and 2019 were 88.19, 87.46, and 89.91, respectively. Modeling land use and land cover changes was conducted to predict 2030. As a result of the study, agricultural and built-up areas and water bodies will increase in 2030. The possible effects of land use changes in 2030 on river water quality were examined as a final step. Based on the results of the water quality simulation in 2030, biochemical oxygen demand, chemical oxygen demand, and NO3 parameters exceeded the maximum permissible level of drinking standard. This study recommends frequent water quality monitoring and LULC planning and management to reduce pollution in river basins.
Assuntos
Rios , Qualidade da Água , Monitoramento Ambiental/métodos , Urbanização , AgriculturaRESUMO
Water clarity has long been used as a visual indicator of the condition of water quality. The clarity of waters is generally valued for esthetic and recreational purposes. Water clarity is often assessed using a Secchi disk attached to a measured line and lowered to a depth where it can be no longer seen. We have applied an approach which uses atmospherically corrected Landsat 8 data to estimate the water clarity in freshwater bodies by using the quasi-analytical algorithm (QAA) and Contrast Theory to predict Secchi depths for more than 270 lakes and reservoirs across the continental US. We found that incorporating Landsat 8 spectral data into methodologies created to retrieve the inherent optical properties (IOP) of coastal waters was effective at predicting in situ measures of the clarity of inland water bodies. The predicted Secchi depths were used to evaluate the recreational suitability for swimming and recreation using an assessment framework developed from public perception of water clarity. Results showed approximately 54% of the water bodies in our dataset were classified as "marginally suitable to suitable" with approximately 31% classed as "eminently suitable" and approximately 15% classed as "totally unsuitable-unsuitable". The implications are that satellites engineered for terrestrial applications can be successfully used with traditional ocean color algorithms and methods to measure the water quality of freshwater environments. Furthermore, operational land-based satellite sensors have the temporal repeat cycles, spectral resolution, wavebands, and signal-to-noise ratios to be repurposed to monitor water quality for public use and trophic status of complex inland waters.
Assuntos
Monitoramento Ambiental , Lagos , Monitoramento Ambiental/métodos , Qualidade da Água , Algoritmos , RecreaçãoRESUMO
Utilizing low-rank prior data in compressed sensing (CS) schemes for Landsat 8-9 remote sensing images (RSIs) has recently received widespread attention. Nevertheless, most CS algorithms focus on the sparsity of an RSI and ignore its low-rank (LR) nature. Therefore, this paper proposes a new CS reconstruction algorithm for Landsat 8-9 remote sensing images based on a non-local optimization framework (NLOF) that is combined with non-convex Laplace functions (NCLF) used for the low-rank approximation (LAA). Since the developed algorithm is based on an approximate low-rank model of the Laplace function, it can adaptively assign different weights to different singular values. Moreover, exploiting the structural sparsity (SS) and low-rank (LR) between the image patches enables the restored image to obtain better CS reconstruction results of Landsat 8-9 RSI than the existing models. For the proposed scheme, first, a CS reconstruction model is proposed using the non-local low-rank regularization (NLLRR) and variational framework. Then, the image patch grouping and Laplace function are used as regularization/penalty terms to constrain the CS reconstruction model. Finally, to effectively solve the rank minimization problem, the alternating direction multiplier method (ADMM) is used to solve the model. Extensive numerical experimental results demonstrate that the non-local variational framework (NLVF) combined with the low-rank approximate regularization (LRAR) method of non-convex Laplace function (NCLF) can obtain better reconstruction results than the more advanced image CS reconstruction algorithms. At the same time, the model preserves the details of Landsat 8-9 RSIs and the boundaries of the transition areas.
RESUMO
The imminent threat of Mongolian montane forests facing extinction due to climate change emphasizes the pressing need to study these ecosystems for sustainable development. Leveraging multispectral remote sensing data from Landsat 8 OLI TIRS (2013-2021), we apply Tsallis non-extensive thermodynamics to assess spatiotemporal fluctuations in the absorbed solar energy budget (exergy, bound energy, internal energy increment) and organizational parameters (entropy, information increment, q-index) within the mountain taiga-meadow landscape. Using the principal component method, we discern three functional subsystems: evapotranspiration, heat dissipation, and a structural-informational component linked to bioproductivity. The interplay among these subsystems delineates distinct landscape cover states. By categorizing ecosystems (pixels) based on these processes, discrete states and transitional areas (boundaries and potential disturbances) emerge. Examining the temporal dynamics of ecosystems (pixels) within this three-dimensional coordinate space facilitates predictions of future landscape states. Our findings indicate that northern Mongolian montane forests utilize a smaller proportion of received energy for productivity compared to alpine meadows, which results in their heightened vulnerability to climate change. This approach deepens our understanding of ecosystem functioning and landscape dynamics, serving as a basis for evaluating their resilience amid ongoing climate challenges.
RESUMO
We aimed to map and analyze LAI by using Landsat 8 and Sentinel-2 time series and the corresponding ground measurements collected in pure Anatolian black pine [Pinus nigra J.F. Arnold ssp. pallasiana (Lamb.) Holmboe] stands within seven-month (from June to December) period. A total of 30 sample plots were selected and seven-month changes of LAI values were determined through hemispherical photography for each sample plot. Remote sensing (reflectance values and vegetation indices obtained from Landsat-8 and Sentinel-2) and topographic (elevation, aspect, and slope) data were used to model the LAI for each month using multiple linear regression (MLR) method. Additionally, the data for all months were combined and modeled. In this case, autoregressive modeling techniques were used to solve the temporal autocorrelation problem. Our study indicated that the models developed from Sentinel-2 give more successful results than Landsat 8 on monthly LAI models. The most successful models were obtained in June by using the reflectance values (Radj2 = 0.39, RMSE = 0.3138 m2 m-2), reflectance values-topographic data (Radj2 = 0.59, RMSE = 0.3174 m2 m-2), vegetation indices-topographic data (Radj2 = 0.82, RMSE = 0.2126 m2 m-2), and reflectance values-vegetation indices-topographic data (Radj2 = 0.93, RMSE = 0.1060 m2 m-2). Among the autoregressive modeling techniques, the highest success was obtained with the Landsat 8 OLI using the moving average (2) procedure (R2 = 0.56). This study is significant that it is the first to analyze the monthly effect on LAI modeling and mapping in pure Anatolian black pine stands using both reflectance values, vegetation indices, and topographic data.
RESUMO
Forest biodiversity conservation and species distribution modeling greatly benefit from broad-scale forest maps depicting tree species or forest types rather than just presence and absence of forest, or coarse classifications. Ideally, such maps would stem from satellite image classification based on abundant field data for both model training and accuracy assessments, but such field data do not exist in many parts of the globe. However, different forest types and tree species differ in their vegetation phenology, offering an opportunity to map and characterize forests based on the seasonal dynamic of vegetation indices and auxiliary data. Our goal was to map and characterize forests based on both land surface phenology and climate patterns, defined here as forest phenoclusters. We applied our methodology in Argentina (2.8 million km2 ), which has a wide variety of forests, from rainforests to cold-temperate forests. We calculated phenology measures after fitting a harmonic curve of the enhanced vegetation index (EVI) time series derived from 30-m Sentinel 2 and Landsat 8 data from 2018-2019. For climate, we calculated land surface temperature (LST) from Band 10 of the thermal infrared sensor (TIRS) of Landsat 8, and precipitation from Worldclim (BIO12). We performed stratified X-means cluster classifications followed by hierarchical clustering. The resulting clusters separated well into 54 forest phenoclusters with unique combinations of vegetation phenology and climate characteristics. The EVI 90th percentile was more important than our climate and other phenology measures in providing separability among different forest phenoclusters. Our results highlight the potential of combining remotely sensed phenology measures and climate data to improve broad-scale forest mapping for different management and conservation goals, capturing functional rather than structural or compositional characteristics between and within tree species. Our approach results in classifications that go beyond simple forest-nonforest in areas where the lack of detailed ecological field data precludes tree species-level classifications, yet conservation needs are high. Our map of forest phenoclusters is a valuable tool for the assessment of natural resources, and the management of the environment at scales relevant for conservation actions.
Assuntos
Florestas , Árvores , Argentina , Biodiversidade , ClimaRESUMO
Soil salinity has been a major factor affecting agricultural production in the Keriya Oasis. It has a destructive effect on soil fertility and could destroy the soil structure of local land. Therefore, the timely monitoring of salt-affected areas is crucial to prevent land degradation and sustainable soil management. In this study, a typical salinized area in the Keriya Oasis was selected as a study area. Using Landsat 8 OLI optical data and ALOS PALSAR-2 SAR data, the optical remote sensing indexes NDVI, SAVI, NDSI, SI, were combined with the optimal radar polarized target decomposition feature component (VanZyl_vol_g) on the basis of feature space theory in order to construct an optical-radar two-dimensional feature space. The optical-radar salinity detection index (ORSDI) model was constructed to inverse the distribution of soil salinity in Keriya Oasis. The prediction ability of the ORSDI model was validated by a test on 40 measured salinity values. The test results show that the ORSDI model is highly correlated with soil surface salinity. The index ORSDI3 (R2 = 0.656) shows the highest correlation, and it is followed by indexes ORSDI1 (R2 = 0.642), ORSDI4 (R2 = 0.628), and ORSDI2 (R2 = 0.631). The results demonstrated the potential of the ORSDI model in the inversion of soil salinization in arid and semi-arid areas.
Assuntos
Salinidade , Solo , China , Radar , Solo/química , Simulação de Ambiente EspacialRESUMO
Rapid and large-scale estimation of soil salt content (SSC) and organic matter (SOM) using multi-source remote sensing is of great significance for the real-time monitoring of arable land quality. In this study, we simultaneously predicted SSC and SOM on arable land in the Yellow River Delta (YRD), based on ground measurement data, unmanned aerial vehicle (UAV) hyperspectral imagery, and Landsat-8 multispectral imagery. The reflectance averaging method was used to resample UAV hyperspectra to simulate the Landsat-8 OLI data (referred to as fitted multispectra). Correlation analyses and the multiple regression method were used to construct SSC and SOM hyperspectral/fitted multispectral estimation models. Then, the best SSC and SOM fitted multispectral estimation models based on UAV images were applied to a reflectance-corrected Landsat-8 image, and SSC and SOM distributions were obtained for the YRD. The estimation results revealed that moderately salinized arable land accounted for the largest proportion of area in the YRD (48.44%), with the SOM of most arable land (60.31%) at medium or lower levels. A significant negative spatial correlation was detected between SSC and SOM in most regions. This study integrates the advantages of UAV hyperspectral and satellite multispectral data, thereby realizing rapid and accurate estimation of SSC and SOM for a large-scale area, which is of great significance for the targeted improvement of arable land in the YRD.
RESUMO
Soil organic carbon (SOC), as the largest carbon pool on the land surface, plays an important role in soil quality, ecological security and the global carbon cycle. Multisource remote sensing data-driven modeling strategies are not well understood for accurately mapping soil organic carbon. Here, we hypothesized that the Sentinel-2 Multispectral Sensor Instrument (MSI) data-driven modeling strategy produced superior outcomes compared to modeling based on Landsat 8 Operational Land Imager (OLI) data due to the finer spatial and spectral resolutions of the Sentinel-2A MSI data. To test this hypothesis, the Ebinur Lake wetland in Xinjiang was selected as the study area. In this study, SOC estimation was carried out using Sentinel-2A and Landsat 8 data, combining climatic variables, topographic factors, index variables and Sentinel-1A data to construct a common variable model for Sentinel-2A data and Landsat 8 data, and a full variable model for Sentinel-2A data, respectively. We utilized ensemble learning algorithms to assess the prediction performance of modeling strategies, including random forest (RF), gradient boosted decision tree (GBDT) and extreme gradient boosting (XGBoost) algorithms. The results show that: (1) The Sentinel-2A model outperformed the Landsat 8 model in the prediction of SOC contents, and the Sentinel-2A full variable model under the XGBoost algorithm achieved the best results R2 = 0.804, RMSE = 1.771, RPIQ = 2.687). (2) The full variable model of Sentinel-2A with the addition of the red-edge band and red-edge index improved R2 by 6% and 3.2% over the common variable Landsat 8 and Sentinel-2A models, respectively. (3) In the SOC mapping of the Ebinur Lake wetland, the areas with higher SOC content were mainly concentrated in the oasis, while the mountainous and lakeside areas had lower SOC contents. Our results provide a program to monitor the sustainability of terrestrial ecosystems through a satellite perspective.
Assuntos
Carbono , Solo , Algoritmos , Ecossistema , Lagos , Aprendizado de Máquina , Tecnologia de Sensoriamento Remoto , Áreas AlagadasRESUMO
This work evaluates the performance of three machine learning (ML) techniques, namely logistic regression (LGR), linear regression (LR), and support vector machines (SVM), and two multi-criteria decision-making (MCDM) techniques, namely analytical hierarchy process (AHP) and the technique for order of preference by similarity to ideal solution (TOPSIS), for mapping landslide susceptibility in the Chitral district, northern Pakistan. Moreover, we create landslide inventory maps from LANDSAT-8 satellite images through the change vector analysis (CVA) change detection method. The change detection yields more than 500 landslide spots. After some manual post-processing correction, the landslide inventory spots are randomly split into two sets with a 70/30 ratio for training and validating the performance of the ML techniques. Sixteen topographical, hydrological, and geological landslide-related factors of the study area are prepared as GIS layers. They are used to produce landslide susceptibility maps (LSMs) with weighted overlay techniques using different weights of landslide-related factors. The accuracy assessment shows that the ML techniques outperform the MCDM methods, while SVM yields the highest accuracy of 88% for the resulting LSM.
Assuntos
Deslizamentos de Terra , Sistemas de Informação Geográfica , Modelos Logísticos , Paquistão , Máquina de Vetores de SuporteRESUMO
The increasing wildfire occurrence due to global climate changes urged the improvement of present wildfire growth prediction and evaluation methods. This study aimed to propose novel solutions to their two primary limitations, including the lack of robust fuel classification method and the low spatial resolution of wildfire growth accuracy assessment while ensuring wide applicability using open data satellite missions and software. The first objective was to create a robust two-step fuel model classification method consisted of the supervised machine learning classification of generalized land cover classes in the 1st level and their individual unsupervised classification to vegetation subtypes in the 2nd level. The second objective was creating a wildfire prediction accuracy assessment method using MODIS 250 m images, which overcome the limitations of low spatial resolution while preserving sub-daily temporal resolution. The wildfire on the Korcula island in Croatia was analyzed in the study, being specific for its long duration from 18 to 24 July 2015. The wildfire ignition occurred in the isolated area, which prolonged the response time from emergency agencies. Random Forest (RF) with input Landsat 8 spectral bands and indices resulted in the highest classification accuracy in the 1st classification level with an overall agreement of 83.6%. The vegetation subclasses from the 2nd classification level were matched to the 13 standard fuel models for the input in FARSITE software. The predicted wildfire evaluation showed the highest mean accuracy of 0.906 for the first two days, which decreased to 0.722 in the latter stages of the active wildfire caused by overprediction. The proposed two-step fuel model classification presented a cost-efficient solution to the fuel map creation in any part of the world, with a disadvantage of no in-situ ground truth identification and accuracy assessment for 2nd classification level. The evaluation of wildfire growth prediction with 250 m images enabled high spatial and temporal resolution of the assessment, while its limitations of wildfire overprediction and the negative effects of wildfire smoke in MODIS images should be addressed in future research.