Land Cover Types Research Articles

Habitat suitability and representation of the potential distribution of the lowland tapir (Tapirus terrestris) within land cover types and protected areas of eastern Colombia

Habitat suitability and representation of the potential distribution of the lowland tapir (Tapirus terrestris) within land cover types and protected areas of eastern Colombia

The Effect of Land Cover Change on Soil Erosion in Awach Kibuon Sub-basin, Kenya

Land cover change is a significant driver of soil erosion. While soil erosion is a natural process, human activities can significantly alter the landscape, making soil more vulnerable to erosion. This erosion reduces a watershed's capacity to sustain vital natural resources and ecosystem services. This study investigated the impact of these changes on soil erosion within four hydrological units (Awach Kibuon, Awach Owade, Awach Kasipul, and Awach Kabondo) of the Awach Kibuon sub-basin between 2018 and 2023. The specific objective of the study was to quantify the effect of land cover change on soil erosion rate and determine how specific land cover types affect soil erosion in the study area. This study employed a quasi-longitudinal design to assess the influence of land cover changes on soil erosion. Sentinel-2 NDVI satellite imagery provided land cover data. The land cover maps, soil data, rainfall data and the Digital Elevation Model were used in the Revised Universal Soil Loss Equation Model within a GIS environment to estimate soil erosion rates. The study revealed a consistent decline in vegetation cover across all hydrological units, as indicated by a decrease in NDVI. The mean NDVI decreased by 12.88%, 10.92%, 4.78%, and 11.92% in Awach Kibuon, Awach Owade, Awach Kasipul and Awach Kabondo respectively. Conversely, the mean soil erosion rate increased by 23.9% in Awach Kibuon, 17.85% in Awach Owade, 24.43% in Awach Kasipul, and 20.54% in Awach Kabondo. Sediment yield increased by 33% in Awach Kibuon, 18% in Awach Owade, 17% in Awach Kasipul, and 23% in Awach Kabondo. These findings suggest a direct relationship between reduced vegetation and elevated soil erosion. The relationship between land cover and erosion varies depending on the density of vegetation. Areas with dense vegetation cover have an inverse relationship, highlighting the protective role of vegetation cover. However, the study also observed that very dense vegetation areas which were also found in high-sloped areas experienced high soil erosion rates. The erosion rate increases even in areas that have experienced an increase in vegetation cover. This is because these areas are also found in high-sloped areas. The slope factor superseded the ability of vegetation cover to protect against soil loss. In conclusion, the change in land cover has significantly increased soil erosion in the Awach Kibuon Sub-basin, however, the slope factor also accelerated soil loss in the basin. Therefore, a holistic approach that combines promoting vegetation cover with land management techniques like terracing and drainage channels is crucial for mitigating soil degradation and water sedimentation in sub-basin.

Moths respond to key habitat structures in conifer plantations managed as irregular high forest

Conifer plantations need to deliver greater multi-functionality beyond the production of timber. This includes the support of biodiversity and associated ecosystem functioning. Forest managers are increasingly adopting novel, more sustainable forest management techniques such as irregular silviculture, a type of continuous cover forestry, to manage plantations, in response to climate change. This may confer benefits to important elements of woodland biodiversity such as moths, by creating plantations that have diverse vertical structure and a permanent canopy (called irregular high forest). In this study at two estates in south-west Britain we specifically tested whether moth species richness and abundance (at the guild and overall level) were associated with (1) surrounding land cover types and (2) habitat structures within stands undergoing transformation using irregular silviculture. A total of 268 species of 11,582 individuals were recorded. Moths responded to all four landscape cover types, negatively to arable and conifer woodland, positively for broadleaf woodland, and both positively and negatively for improved grassland. Moth richness and abundance had a positive association with basal area (albeit below standard even aged stocking rates with most plots around 40–45 m2ha) and a weak positive association with higher vertical structural complexity (indicated by a more developed understorey and sub-canopy layers). Moth abundance had a moderate positive association with broadleaf canopy where we recommend plantations should have at least 10–15 % broadleaf canopy cover. Moth richness and abundance were negatively associated to mean tree diameter and canopy openness. Many of these features are more prevalent in stands further progressed towards irregular high forest. We show how moth communities can be maintained via more sustainable forestry management techniques that increase overall habitat and structural complexity.

LAND MOVEMENT VULNERABILITY ZONING IN THE HILLY AREA OF NORTHERN CILACAP, CENTRAL JAVA

The northern part of Cilacap has an area with hilly morphology with slopes that are relatively gentle to steep. The geological setting in this region is quite complex, the combination of tectonics and the physical properties of the rocks means that this region is strongly deformed. This has an impact on areas prone to landslides as one of land movement types. To what extent can this area be classified based on the vulnerability of its land to landslides is the problem be studied. There are five aspects used in this research, namely rainfall, rock type, slope, land cover and soil type. The multi-aspect overlay method is used to determine the spatial distribution of land movement-prone zones. Geographic Information System-based software support makes the data analysis process easier. Based on the results of the spatial analysis of five aspects, it was concluded that the North Cilacap area could be grouped into 3 land movement vulnerability zones, namely the low vulnerability zone with a score of 8-17, the medium vulnerability zone with a score of 18-27, and the high vulnerability zone with a score of 28-37. The complex tectonic order is the main controller of land movement events in this region.

Runoff Coefficient in the Air Bengkulu Watershed and the Evaluation of the Existing Spatial Planning

Runoff coefficient plays a crucial role in estimating the peak discharge of a river basin. Therefore, this research aimed to investigate runoff coefficient in the Air Bengkulu watershed based on temporal land cover data and evaluate the existing spatial planning. Land cover data from 1998, 2002, 2016, and 2023, with spatial patterns derived from current regional planning were used. The temporal and spatial pattern-based runoff coefficients were determined using land cover data and spatial pattern function, respectively. Meanwhile, descriptive and comparative methods were adopted based on time. The calculated runoff coefficient was 0.073, 0.093, 0.276, and 0.273 for 1998, 2002, 2016, and 2023, while the value obtained based on spatial patterns was 0.306. Additionally, a general trend of increasing the values over time was observed. Land cover change, particularly the decline of forest areas and the expansion of settlement and plantation, contributed to the rising runoff coefficient. The results showed that runoff coefficient (0.306) exceeded the current land cover, similar to other analyses with higher runoff in the predicted scenario. This research suggested a need for a more detailed classification system and scale to accommodate land cover types with relatively low runoff coefficient. In risk assessment, land cover-like spatial patterns with low runoff coefficient should be placed as capacity other than vulnerability components.

Spatiotemporal changes of urban heat island effect relative to land surface temperature: a case study of Jinan City, China.

The urban heat island (UHI) effect has become increasingly prevalent and significant with the accelerated pace of urbanization, posing challenges for urban planners and policymakers. To reveal the spatiotemporal variations of the urban heat island effect in Jinan City, this study utilized Landsat satellite images from 2009, 2014, and 2019, employing the classic Mono-Window algorithm to extract land surface temperature (LST). Additionally, Geodetector was introduced to conduct a detailed analysis of the relationship between LST in Jinan City and land cover types (vegetation, water bodies, and buildings). The results indicate a significant increase in the severity of the urban heat island effect in Jinan from 2009 to 2019, with the central urban area consistently exhibiting a high-intensity core heat island. Suburban areas of Jinan show a clear trend of merging their heat island effects with the central urban area. The combined area of strong cool island effect zones and cool island effect zones within water bodies reaches 89.7%, while the combined proportion of heat island and strong heat island effect zones in building areas is 62.2%. Vegetation cover (FVC) exerts the greatest influence among all factors on the intensity level of the urban heat island effect. These findings provide a reliable basis for decision-making related to urban planning and construction in Jinan City.

Quality indexes of land ownership and land cover as indicators of small lake catchment degradation

Changes in land ownership, generally resulting in changes in land cover and progressive fragmentation of catchments, often result in the inability to take pro-environmental measures. The aim of this paper is to assess the impact of land ownership on the environmental status of small, but environmentally valuable lake catchments. The study of environmental quality in terms of ownership and land cover was conducted in lake catchments located in the central part of the West Polesie Transboundary Biosphere Reserve. In total, the ownership type was determined for 5661 plots and the land cover type was determined for 2502 subdivisions. The quality of the analysed elements of the study catchments was subjected to expert evaluation using a modified Stapel Scale and was presented using two Qualitative Indexes proposed: Land Cover [Lc] and Land Ownership [Lo]. The obtained values of Index Lc ranged from −0.18 to +1.97, and Index Lo from −1.87 to +2.00. Application of the Average Land Quality Index [LQ], in turn, made it possible to randomise the environmental quality of the analysed catchments, for which it was from −1.02 to +1.99 (areas of very good and very poor environmental quality, respectively). The analyses indicated, for example, catchments that were significantly degraded because of a very large number of private plots in their area − in the five analysed catchments, this ranged from 505 to 719 and their average area was 0.19–1.10 ha. It is noteworthy that very unfavourable land ownership conditions and land cover structure were found in the Poleski National Park catchments – the area with the highest protection status. The proposed assessment system can be used for preliminary analysis of actual and potential threats to lake catchments so as to indicate possible directions of conservation measures and their sensibility.

Capturing Snowmelt Runoff Onset Date under Different Land Cover Types Using Synthetic Aperture Radar: Case Study of Sierra Nevada Mountains, USA

Snow plays a crucial role in the global water and energy cycles, and its melting process can have a series of impacts on hydrological or climatic systems. Accurately capturing the timing of snowmelt runoff is essential for the utilization of snow resources and the early warning of snow-related disasters. A synthetic aperture radar (SAR) offers an effective means for capturing snowmelt runoff onset dates (RODs) over large areas, but its accuracy under different land cover types remains unclear. This study focuses on the Sierra Nevada Mountains and surrounding areas in the western United States. Using a total of 3117 Sentinel-1 images from 2017 to 2023, we extracted the annual ROD based on the Google Earth Engine (GEE) platform. The satellite extraction results were validated using the ROD derived from the snow water equivalent (SWE) data from 125 stations within the study area. The mean absolute errors (MAEs) for the four land cover types—tree cover, shrubland, grassland, and bare land—are 24, 18, 18, and 16 d, respectively. It indicates that vegetation significantly influences the accuracy of the ROD captured from Sentinel-1 data. Furthermore, we analyze the variation trends in the ROD from 2017 to 2023. The average ROD captured by the stations shows an advancing trend under different land cover types, while that derived from Sentinel-1 data only exhibits an advancing trend in bare land areas. It indicates that vegetation leads to a delayed trend in the ROD captured by using Sentinel-1 data, opposite to the results from the stations. Meanwhile, the variation trends of the average ROD captured by both methods are not significant (p > 0.05) due to the impact of the extreme snowfall in 2023. Finally, we analyze the influence of the SWE on RODs under different land cover types. A significant correlation (p < 0.05) is observed between the SWE and ROD captured from both stations and Sentinel-1 data. An increase in the SWE causes a delay in the ROD, with a greater delay rate in vegetated areas. These findings will provide vital reference for the accurate acquisition of the ROD and water resources management in the study area.

Improving high uncertainty of evapotranspiration products under extreme climatic conditions based on deep learning and ERA5 reanalysis data

Accurate quantification of Terrestrial Evapotranspiration (ET) is crucial for understanding global water resource management amid climate change. However, against the backdrop of frequent global extreme climate events, existing ET products don’t accurately reflect ground data under extreme climate conditions (high temperature and high vapor pressure deficit (VPD)) and no studies have addressed this uncertainty. This study, based on data from 150 flux towers (FLUXNET ET), utilizes four daily-scale ET product datasets (REA, GLEAMv3.6b, REA, ERA5) and ERA5 reanalysis meteorological data as inputs for Convolutional Neural Network (CNN), to explore ways to improve the accuracy of ET products under extreme climatic conditions, while three Machine Learning (ML) models were selected as comparisons. An additional 35 nearby sites provide independent validation for model applicability. The results show that the estimation accuracy of ET products under extreme climatic conditions has an average decrease of 37.5 % in the correlation coefficient (R2) and an average increase of 40 % in the root-mean-square-error (RMSE), whereas the CNN significantly improves this high uncertainty and outperforms the other ML models. The accuracy was significantly improved with the addition of ET products (R2 = 0.924, RMSE=0.323 mm/d) compared to having only meteorological data as input (R2 = 0.888, RMSE=0.369 mm/d). Under extreme climatic conditions, the R2 and RMSE of CNN are 0.786 (171.6 % average improvement) and 0.448 (65.4 % average reduction) compared to ET products, with PBias ranging from −9% to 9 % and the most significant improvement in the America (region), EBF (land cover type), and Cwa (climate type). Furthermore, using ET products as inputs significantly enhances the CNN’s estimation accuracy, yielding R2 and RMSE improvements between 10 % and 30 %. At the 35 independent validation sites, the CNN still performs well, with R2 and RMSE around 0.720 and 0.510 mm/d under extreme climatic conditions, and the accuracy within 500 km is even higher. This study enhances model precision, providing a valuable tool for optimizing water management and tackling climate change. These advancements help formulate precise irrigation strategies and water policies, particularly in extreme weather-prone regions, supporting ecological stability and mitigating water scarcity.

Deep learning-based modeling of land use/land cover changes impact on land surface temperature in Greater Amman Municipality, Jordan (1980–2030)

Modeling the impacts of Land Use/Land Cover changes (LULCC) on Land Surface Temperature (LST) is crucial in understanding and managing urban heat islands, climate change, energy consumption, human health, and ecosystem dynamics. This study aimed to model past, present, and future LULCC on Land Surface Temperatures in the Greater Amman Municipality (GAM) in Jordan between 1980 and 2030. A set of maps for land cover, LST, Normalized Difference Vegetation Index (NDVI), Normalized Difference Built-up Index (NDBI), and topography was integrated into the Cellular Automata-Artificial Neural Network (CA-ANN) and the Long-Short-Term Model (LSTM) models to predict the LULC and LST for 2030. The results showed an expansion of urban areas in GAM from 54.13 km2 (6.6%) in 1980 to 374.1 km2 (45.3%) in 2023. However, agricultural areas decreased from 152.13 km2 (18.5%) in 1980 to 140.38 km2 (17%) in 2023, while barren lands decreased from 54.44 km2 (6.6%) in 1980 to 34.71 km2 (4.22%) in 2023. Forested areas declined from 4.58 km2 (0.56%) in 1980 to 4.35 km2 (0.53%) in 2023. Rangelands/ sparsely vegetated areas declined from 557 km2 (67.7%) in 1980 to 270.71 km2 (32.9%) in 2023. The results of modeling LST showed an increase in average LST for all land cover types, with the most significant increases evident within urban areas and Rangelands/Sparsely vegetated areas. The slightest increase in LST was within forested areas as the average LST increased from 28.42 °C in 1980 to 34.16 °C in 2023. The forecasts for the future showed a continuous increase in LST values in all land cover types. These findings highlight the impact of land surface dynamics and their impact on increasing land surface temperature, which urges the adoption of more sustainable planning policies for more livable and thermally comfortable cities.

Spatiotemporal analysis of tropical vegetation ecosystems and their responses to multifaceted droughts in Mainland Southeast Asia using satellite-based time series

ABSTRACT Drought ranks among the costliest of all climate-related phenomena and manifests in various forms, posing significant challenges in understanding its influence on agriculture and natural ecosystems. Mainland Southeast Asia (MSEA), a significant region of tropical agriculture and vegetation ecosystems, has become increasingly susceptible to drought hazards. In this study, we characterized and assessed vegetation dynamics and their drought impacts using correlation analysis and explainable machine learning methods under different vegetation types and elevation zones during the dry growing seasons from 2000 to 2022. Specifically, we characterized the vegetation dynamics and their trend in space and time. Next, we assessed vegetation-drought responses in consideration of meteorological, hydrological, and agricultural droughts under different land cover types and elevation characteristics. Lastly, we used an explainable machine learning method to quantify the drivers and impacts of multifaceted droughts on natural and undisturbed vegetation ecosystems. Our findings revealed that nearly 70% of the MSEA region experienced a greening trend despite large areas of vegetative damage during the drought years. Vietnam witnessed increasing vegetation condition in most observed years while the declining trend was mainly found in Cambodia and southern Laos. Vegetation-drought responses showed that tropical vegetation had a high sensitivity to drought conditions, and stronger responses were observed in rainfed crop, mixed forest, and deciduous forest at lower altitude areas. In natural and undisturbed ecosystems, short-term meteorological and agricultural drought disturbances accounted for nearly 93% of variations in tropical vegetation. Among the different examined drought indices, the 3-month Standardized Precipitation Evapotranspiration (SPEI-3) and Temperature Condition Index (TCI) were identified as the factors having the largest influence, together explaining about 55% of the observed natural undisturbed vegetation variations. These findings deepen our understanding of tropical vegetation-drought responses and the underlying drivers of natural and undisturbed ecosystems. Such insights could provide valuable information to assist national and local governments in the MSEA region in developing effective drought management and adaptation programs to safeguard tropical agricultural production and natural ecosystems amidst growing climate challenges.

Land Cover Mapping in a Mangrove Ecosystem Using Hybrid Selective Kernel-Based Convolutional Neural Networks and Multi-Temporal Sentinel-2 Imagery

Mangrove ecosystems provide numerous ecological services and serve as vital habitats for a wide range of flora and fauna. Thus, accurate mapping and monitoring of relevant land covers in mangrove ecosystems are crucial for effective conservation and management efforts. In this study, we proposed a novel approach for mangrove ecosystem mapping using a Hybrid Selective Kernel-based Convolutional Neural Network (HSK-CNN) framework and multi-temporal Sentinel-2 imagery. A time series of the Normalized Difference Vegetation Index (NDVI) products derived from Sentinel-2 imagery was produced to capture the temporal behavior of land cover types in the dynamic ecosystem of the study area. The proposed algorithm integrated Selective Kernel-based feature extraction techniques to facilitate the effective learning and classification of multiple land cover types within the dynamic mangrove ecosystems. The model demonstrated a high Overall Accuracy (OA) of 94% in classifying eight land cover classes, including mangrove, tidal zone, water, mudflat, urban, and vegetation. The HSK-CNN demonstrated superior performance compared to other algorithms, including random forest (OA = 85%), XGBoost (OA = 87%), Three-Dimensional (3D)-DenseNet (OA = 90%), Two-Dimensional (2D)-CNN (OA = 91%), Multi-Layer Perceptron (MLP)-Mixer (OA = 92%), and Swin Transformer (OA = 93%). Additionally, it was observed that the structure of the network, such as the types of convolutional layers and patch sizes, affected the classification accuracy using the proposed model and, thus, the optimum scenarios and values of these parameters should be determined to obtain the highest possible classification accuracy. Overall, it was observed that the produced map could offer valuable insights into the distribution of different land cover types in the mangrove ecosystem, facilitating informed decision-making for conservation and sustainable management efforts.

Land-cover gradients determine alternate drivers of mammalian species richness in fragmented landscapes

ContextUnderstanding habitat fragmentation is a critical concern for nature conservation and the focus of intense debate in landscape ecology. Resolving the uncertainty around the effects of habitat fragmentation on biodiversity remains an ongoing challenge that requires the successful delineation of multiple patch-landscape interactions.ObjectivesWe carried out a regional analysis on species richness of woodland mammals to determine the relative influence of structural, compositional and functional characteristics related to woodland habitat across different land-cover gradients.MethodsWe calculated the Edge-weighted Habitat Index, an area-weighted measure of functional connectivity that incorporates a mechanistic estimate of edge-effects, for interior woodland habitat. We compared its influence on mammalian species richness to that of increasing edge and patch density, landscape diversity, and a habitat-only model, in different contexts of matrix hostility across Northern England in the UK.ResultsOur results demonstrate the relevance of alternative drivers of species richness resulting from patch-landscape interactions across gradients of matrix hostility. Evidence is provided for positive and negative effects of increasing structural (edge density), functional (connected interior habitat) and compositional (landscape diversity) attributes, varying according to matrix type and intensity. Results were sensitive to dominant land-cover types in the matrix and the scale of observation.ConclusionThis study provides new insights into fragmentation effects on biodiversity and clarifies assumptions around the relative influence of structural, compositional and functional habitat characteristics on landscape-level species richness. We highlight the presence of thresholds, related to matrix hostility, that determine alternative drivers of species richness in woodland mammals. These drivers, and related thresholds, were sensitive to the scale of observation and landscape context. Landscape decisions aimed at promoting biodiversity should consider sources of matrix hostility and homogeneity at scales relevant to ecological processes of interest.

Evaluation and interpretation of landscapes from satellite imagery

Satellite imagery allows us to view landscapes from a bird’s eye view, providing a new dimension in appreciating the environments we inhabit. This alternative perspective has the potential to shape individual perceptions of landscapes and play a pivotal role in land management decision-making and communication. However, the interpretation and appreciation of landscapes seen in satellite imagery may vary among observers. This study investigates the relationship between individuals’ ability to interpret images from eye-level and satellite perspectives, their familiarity with the landscape, and their appreciation of land cover from this viewpoint. To achieve this, a survey was conducted presenting respondents with images of land cover classes captured at eye level and from satellite imagery of the Yungay municipality in Chile. Participants were asked to interpret the primary land use land cover (LULC) depicted in the imagery and indicate their appreciation of that landscape. Variation in the interpretation of LULC was observed between the image source and land cover type. For instance, forest classes seen in eye-level imagery were more accurately interpreted compared to satellite imagery, while the reverse was true for agriculture. These differences in interpretation also impacted the appreciation scores assigned to the landscapes in the images. Specifically, if respondents perceived an image to be dominated by a traditionally appreciated land cover (e.g., Native Vegetation), they provided a higher score, even if the image depicted another class (e.g., Plantation Forestry). These findings highlight that considering the influence of satellite imagery in shaping perception is crucial in supporting land management activities.

Building Shadow Detection Based on Improved Quick Shift Algorithm in GF‐2 Images

Shadows in remote sensing images contain crucial information about various features on the ground. In this study, a method for detecting building shadows in GF‐2 images based on improved quick shift was proposed. First, six feature variables were constructed: first principal component (PC1), brightness component (I), normalized difference shadow index (NDSI), morphological shadow index (MSI), normalized difference water index (NDWI), and normalized difference vegetation index (NDVI). Then, the image was segmented to obtain homogeneous objects, which were then classified using a random forest model. Two improvements were added to the quick shift algorithm: using PC1, I, and MSI as input data instead of RGB images; and adding Canny edge constraints. Validation in six research areas yields Kappa coefficients of 0.928, 0.896, 0.89, 0.913, 0.879, and 0.909, confirming method feasibility. In addition, comparative experiments demonstrate its effectiveness and robustness across different land cover types while mitigating the segmentation scale effect.

A gap filling method for daily evapotranspiration of global flux data sets based on deep learning

In response to irregular data gaps in evapotranspiration (ET) data obtained using eddy covariance (EC) methods, this study seeks to explore a high-precision interpolation method that is not constrained by spatiotemporal limitations. Leveraging daily observations from 241 global flux towers (FLUXNET ET) as a reference, data gaps are artificially introduced at five different proportions: 5 %, 10 %, 20 %, 35 %, and 50 %, this process is repeated 30 times (Take the average of the results). Four ET products, along with environmental factors (ERA5 meteorological data and NOAA Normalized Difference Vegetation Index), are employed as input to assess the performance of the target model, Convolutional Neural Network (CNN)-Long Short-Term Memory (LSTM), as well as five comparative models: CNN, LSTM, Random Forest (RF), Support Vector Machine (SVM), and Extreme Learning Machine (ELM), in filling the data gaps. These results are then compared with the filling outcomes obtained without ET products as input, aiming to investigate whether ET products contributes to improving the accuracy and stability of ET gap filling. The results indicate that, firstly, the filling performance of the target model CNN-LSTM surpasses that of other models, with minimal impact from gap proportions. Moreover, incorporating ET products as input alongside environmental factors significantly enhances the filling accuracy and stability of the model. Compared to scenarios without ET products as input, the CNN-LSTM model shows an average increase in R2 of 5 % to 15 % and a decrease in root mean square error (RMSE) of 15 % to 35 %. Specifically, as the gap proportion increases from 5 % to 50 %, the average Pearson correlation coefficient (R2) decreases from 0.952 to 0.846 (a decline of 11 %), the average RMSE increases from 0.491 mm/d to 0.620 mm/d (a rise of 26 %), and reached the filling limit at the gap ratio of 93 % (R2 below 0.4). Conversely, without ET products as input, as the gap proportion increases, the average R2 decreases from 0.916 to 0.784 (a decline of 14 %), the RMSE increases from 0.552 mm/d to 0.952 mm/d (a rise of 72 %), and reached the filling limit at the gap ratio of 80 %. Secondly, the contribution of ET products to the model estimation far exceeds that of ERA5 meteorological factors and NDVI, with the contribution of ET products increasing with higher gap proportions from 58.0 % to 73.6 %. Finally, different latitudes, land cover types, and climate types influence the model filling performance, with better performance observed in high-latitude regions, Grassland (GRA), and Dry-Cold climates. In summary, incorporating ET products as input for CNN-LSTM yields high-precision gap-filling results, surpassing existing typical methods for gap filling and its application is not constrained by spatiotemporal limitations.

Not population density, but city size and per capita income influence the urban forest carbon sequestration: A case of growing cities in Arkansas, USA

While urban trees provides significant ecosystem services, the impact of increasing urbanization and population on their carbon sequestration potential remains relatively underexplored in interior sprawling cities, highlighting a critical research gap amid growing urban environmental challenges. Although existing studies suggest a strongly correlation between urban forest cover and carbon sequestration, this study aims to investigate how the urban demography factors influence land cover and carbon sequestration across cities with low, medium, and high population densities. Using the i-Tree Canopy tool, we sampled 18,814 random points to estimate city land cover and carbon sequestration in Arkansas, interior state of the US. Results from the principal component analysis revealed that 74% of the variation in a city's land cover among population density classes is attributed to tree and soil cover percentages. The mean annual (1.40 ± 0.08 T ha-1) and total carbon sequestration rates (35.30 ± 5.57 T ha-1) were significantly higher in the areas with low population density due to a higher average tree cover percentage (46%). Carbon sequestration rate was negatively associated with land cover types such as impervious surfaces like buildings (r = -0.44), roads (r = -0.28), grass cover (r = -0.46), and impervious other (r = -0.53). The log-linear model with scaled variables suggested that factors such as city's size and tree cover has positive influence urban trees' annual and total carbon sequestration, while spatial distribution and per capita income negatively affect these ecosystem befits in growing cities, regardless of population density. Travel time within city may have negative impact, while decadal population change may positively impact on carbon sequestration alongside management effort. Valuing this potential urban forest carbon sequestration for emerging carbon market could provide monetary benefits to the urban communities, city managers, and policymakers, enabling the development of effective urban forest management strategies in the context of urban climate change.

Utilizing multitemporal indices and spectral bands of Sentinel-2 to enhance land use and land cover classification with random forest and support vector machine

Multitemporal imagery offers a critical advantage by capturing seasonal variations, which are essential for differentiating between land use and land cover (LULC) types. While these types may appear similar when examined at one specific time, they exhibit distinct phenological patterns across different seasons. This temporal depth is crucial for enhancing model accuracy, particularly in heterogeneous landscapes where LULC transitions are frequent and complex. This paper made use of spectral bands and indices of Sentinel-2 from April to September 2020 for LULC classification using two advanced machine learning models: Random forest (RF) and support vector machine (SVM). The spectral indices include the normalized difference vegetation index (NDVI), normalized difference built-up index (NDBI), and modified normalized water index (MNDWI). The dataset was divided into four temporal feature sets: April-May, June-July, August-September, and the entire period from April-September. For each two-month period, the median values of the spectral bands and indices were used. Both models were evaluated based on overall accuracy, F1-score, Kappa coefficient, precision, and recall. Results indicate that incorporating multitemporal features enhanced the performance of the chosen models, with overall accuracy increasing from 82.4% to 94.03% for RF and from 75.4% to 93.54% for SVM. Additionally, the RF algorithm demonstrated higher accuracy than the SVM model across various time periods, with notable improvements in other performance metrics. These improvements also underscore the ability of the models to leverage the rich multitemporal data provided by Sentinel-2 for accurate LULC classification. This study highlights the importance of incorporating the dynamics of features in remote sensing applications to enhance the precision and reliability of LULC classification.

A Data‐Driven Approach to Assess the Impact of Climate Change on a Tropical Mangrove in India

AbstractAs a potential carbon sink, mangroves play an important role in climate mitigation. India houses several major global mangrove patches, which remain vulnerable to climate change. The ecosystem‐atmosphere CO2 exchange is most accurately measured by the eddy covariance method, whereas satellites provide the biophysical parameters for a wider area. In the present study, the Sentinel‐2 satellite data is used to map the land cover types in the Pichavaram mangrove forest and identify two major dominant species (Rhizophora spp. and Avicennia marina), which indicated more than 95% classification accuracy. We used 2 years (2017 and 2018) of in situ gross primary productivity (GPP) and leaf area index (LAI) measurements and rectified the Moderate Resolution Imaging Spectroradiometer (MODIS) GPP and LAI products from 2010 to 2018. The modified MODIS GPP and LAI products were used to develop machine learning models, that is, Random Forest (RF) and Extreme Gradient Boosting (XGBoost) to study the climate influence on mangrove productivity. The RF model (R2 = 0.85 and root mean square error (RMSE) = 0.2) outperformed the XGBoost model (R2 = 0.75 and RMSE = 0.26) and was used to project the impact of climate change on the mangrove GPP for two extreme climate change scenarios, namely SSP1‐1.26 and SSP5‐8.5. The GPP increases and decreases in future during wet and dry periods, respectively. Overall, the projected GPP indicated a reduction of 3.73%–20.3% from 2050 to 2060 and of 4.82%–28.15% from 2090 to 2100, compared to its current average (from 2010 to 2018).

Impact of Drought Duration and Severity on Drought Recovery Period for Different Land Cover Types in Balochistan, Pakistan

Impact of Drought Duration and Severity on Drought Recovery Period for Different Land Cover Types in Balochistan, Pakistan