Time Series Of Imagery Research Articles

Time Series approach to map areas of Agricultural Plastic Waste generation

Abstract. The escalating presence of plastics in agriculture has raised concerns regarding Agricultural Plastic Waste (APW). Hitherto, a lack of comprehensive plasticulture data impedes effective waste management strategies, potentially resulting in plastic pollution and contributing to microplastic formation. APW locations and quantities are pivotal for territorial planning and the formulation of public policies on waste management and land use. Remote detection of agri-plastics has garnered increased consideration, particularly in mapping plastic-mulched farmlands (PMFs) dispersed in extensive regions. This study investigates whether the use of satellite image time series can map PMFs accurately. We assessed pixel-based classification using a 16-day composite time series of Sentinel-2 imagery (S2-16D) obtained from the Brazil Data Cube project. The Plastic-Mulched Landcover Index (PMLI) was joined into S2-16D bands because it is an index focused on PMF detection. Four classifiers (RF, MLP, L-TAE, and TCNN) were compared through agreed classification metrics. The most promising outcome showed an overall accuracy of 100% employing L-TAE but with visible noise in the map. The time series enhanced accuracy while minimizing background confusion, offering a viable solution for PMF detection. The PMF map presented herein represents an initial stride toward fostering circularity in plasticulture throughout South America.

Divergent dynamics of coastal wetlands in the world’s major river deltas during 1990–2019

Coastal wetlands provide vital dynamic ecosystem services. They have become increasingly important after being linked to several sustainable developmental goals, resulting in a focus on their protection, management, and restoration. Therefore, there is an increasing need to detect and compare coastal wetland spatiotemporal dynamics in deltas at a global scale. In this study, we mapped and characterized coastal wetland spatiotemporal patterns for 1990–2019 in the world’s major river deltas using pixel frequency algorithms and Landsat-4/5 (TM), −7 (ETM + ), −8 (OLI), and Sentinel-2 (MSI) time-series imagery obtained from Google Earth Engine (GEE). Our map had a high overall accuracy (91.84 %) for 2019. Tidal flats were primarily distributed in North America (∼6.87 %) and Asia (∼5.91 %), whereas salt marshes were more commonly found in North America (∼45.39 %) and South America (∼10.61 %). Mangroves are more common in South America (∼11.86 %) and Asia (∼5.83 %), primarily because of the Amazon River Delta and tropical and subtropical regions of Asia, which host several large river deltas. South America had the largest coastal delta wetland area (798,569 km2), followed by Asia (640,251 km2), North America (581,977 km2), Africa (181,977 km2), Europe (140,759 km2), and Oceania (15,915 km2). There was a minor difference in the distribution of wetland vegetation and tidal flats in Asian coastal deltas, and the wetland vegetation area in Asia was greater than that in tidal flats on other continents. We found that the coastal wetland areas increased during 1990–2001, decreased during 2001–2012, and steadily increased during 2012–2019. Our study provides a baseline for monitoring the area, status, and health of the coastal wetlands in these river deltas.

Incidence angle dependency and seasonal evolution of L and C-band SAR backscatter over landfast sea ice

Abstract We estimate sea-ice type specific incidence angle (IA) dependencies for dual polarized (HH/HV) L and C-band synthetic aperture radar (SAR) for the winter, melt onset and advanced melt seasons for level and deformed ice, using time-series of Advanced Land Observing Satellite-2 (ALOS-2) and Sentinel-1 imagery off the north-east coast of Greenland. The IA dependencies are used to radiometrically correct the L and C-band backscatter time-series, which enables analysis of their seasonal evolution. From this, we observe that the L-band backscatter intensity increases for both ice types at the transition from winter to melt onset. We use these results to estimate ice type separability and to train an IA aware Bayesian classifier at both frequencies. These results show that while both frequencies are capable of distinguishing level and deformed ice during the winter, only L-band SAR can reliably make this separation during the melt onset season. During the advanced melt season, the overall classification accuracies are similarly low for L and C-band. This study demonstrates the potential of L-band SAR for sea-ice mapping, which is highly relevant in the light of several upcoming L-band SAR missions.

Integrating time series Sentinel-2 images and tide height to mapping tidal flats in the Chinese mainland

As a unique and important ecosystem, tidal flats provide a variety of ecosystem functions and services. Mapping tidal flats is essential for the protection and management of coastal ecosystems. However, large-scale tidal flats mapping still faces challenges due to the tidal variation and spectral similarity between tidal flats and inland wetlands. Previous methods rely on the coastlines or maximum seawater extent to exclude inland areas, which is limited by its inability to effectively differentiate tidal flats from spectrally similar inland wetlands. To address these issues, we proposed a new tidal-flat mapping method by integrating Sentinel-2 time series imagery with tide height (TH) data from ground-based tide stations on Google Earth Engine. We first generated images at the lowest and highest tidal stages, established the statistical relationship between the Normalized Difference Water Index (NDWI) of each pixel and TH, and then concatenated them into a Random Forest classifier for further classification. The statistical relationship between NDWI and TH amplified the difference between tidal flats and inland wetlands, thus significantly reducing the influence of spectral similarity. This method could produce a high-precision tidal flats map with an overall accuracy of 97.30% in the coastal zone of the Chinese mainland. By quantitatively comparing with the previous tidal flat maps, we found that the strategies of tidal-level information simulation and inland area exclusion were the two main reasons producing the differences among the maps. The proposed method does not rely on space constraints to exclude inland wetlands and can capture more estuarine tidal flats, so it can be used as a reliable means to monitor the tidal flats in large-scale areas.

Wave runup and total water level observations from time series imagery at several sites with varying nearshore morphologies

Coastal imaging systems have been developed to measure wave runup and total water level (TWL) at the shoreline, which is a key metric for assessing coastal flooding and erosion. However, extracting quantitative measurements from coastal images has typically been done through the laborious task of hand-digitization of wave runup timestacks. Timestacks are images created by sampling a cross-shore array of pixels from an image through time as waves propagate towards and run up a beach. We utilize over 7000 hand-digitized timestacks from six diverse locations to train and validate machine learning models to automate the process of TWL extraction. Using these data, we evaluate two deep learning model architectures for the task of runup detection. One is based on a fully convolutional architecture trained from scratch, and the other is a transformer-based architecture trained using transfer learning. The deep learning models provide a probability of each pixel being either wet or dry. When contoured at the 50% level (equal chance of being wet or dry), the deep learning models more accurately identified TWL maxima than minima at all sites. This resulted in accurate predictions of 2% exceedance runup, but under predictions of significant swash and over predictions of wave setup. Improved agreement with the complete TWL time series was obtained through post-processing by utilizing the wet/dry probability of each pixel to weight the contouring toward lower dryness probabilities for runup minima (maxima agreed well with observations without tuning). Overall, a transformer-based model using transfer learning provided the best agreement with wave runup statistics, including a) the 2% exceedance runup, b) significant swash, and c) wave setup at the shoreline. For a random subset of images, the model was found to be within the uncertainty range of hand-digitization. The relative success of the transfer learning model suggests that fine-tuning a large model has advantages compared to training a smaller model from scratch. Models provide per-pixel probabilistic estimates in less than 10 s per timestack on a single computational unit, versus the more than 5 min required for hand-digitization. The model is therefore well-suited for near real-time applications, allowing for the development of early warning systems for difficult to forecast events. Real-time wave runup and total water level observations can also be incorporated into coastal hazards forecasts for data assimilation and continual model validation and improvement.

Ensemble Machine Learning on the Fusion of Sentinel Time Series Imagery with High-Resolution Orthoimagery for Improved Land Use/Land Cover Mapping

In the United States, several land use and land cover (LULC) data sets are available based on satellite data, but these data sets often fail to accurately represent features on the ground. Alternatively, detailed mapping of heterogeneous landscapes for informed decision-making is possible using high spatial resolution orthoimagery from the National Agricultural Imagery Program (NAIP). However, large-area mapping at this resolution remains challenging due to radiometric differences among scenes, landscape heterogeneity, and computational limitations. Various machine learning (ML) techniques have shown promise in improving LULC maps. The primary purposes of this study were to evaluate bagging (Random Forest, RF), boosting (Gradient Boosting Machines [GBM] and extreme gradient boosting [XGB]), and stacking ensemble ML models. We used these techniques on a time series of Sentinel 2A data and NAIP orthoimagery to create a LULC map of a portion of Irion and Tom Green counties in Texas (USA). We created several spectral indices, structural variables, and geometry-based variables, reducing the dimensionality of features generated on Sentinel and NAIP data. We then compared accuracy based on random cross-validation without accounting for spatial autocorrelation and target-oriented cross-validation accounting for spatial structures of the training data set. Comparison of random and target-oriented cross-validation results showed that autocorrelation in the training data offered overestimation ranging from 2% to 3.5%. The XGB-boosted stacking ensemble on-base learners (RF, XGB, and GBM) improved model performance over individual base learners. We show that meta-learners are just as sensitive to overfitting as base models, as these algorithms are not designed to account for spatial information. Finally, we show that the fusion of Sentinel 2A data with NAIP data improves land use/land cover classification using geographic object-based image analysis.

Land Use Classification Study based on Landsat 8 OLI Time Series Imagery and Support Vector Machine

Land Use Classification Study based on Landsat 8 OLI Time Series Imagery and Support Vector Machine

Estimating forest aboveground biomass in tropical forests with Landsat time-series data and recurrent neural network

ABSTRACT The insensitivity and saturation of sensor signals significantly limit the application of Landsat data on estimating biomass in tropical forests with a high level of biomass. The previous studies with single-date Landsat data have lower accuracies in tropical forests compared to boreal and temperate forests. Landsat time-series data provide a promising opportunity to improve the accuracy by enhancing the relationship between Landsat spectral reflectance and forest aboveground biomass with disturbance and recovery dynamics. Compared to the single-date image, Landsat time-series data can capture abrupt spectral changes (e.g. harvesting and fire) and show the regrowth process in forested pixels. However, very limited studies take advantage of Landsat time-series data to estimate aboveground biomass in tropical forests. Recurrent neural networks (RNNs) are powerful deep learning techniques to capture time dependencies in sequence data. However, the application of RNNs in estimating forest biomass has not been explored yet. Therefore, we integrate the long short-term memory network (LSTM) and the fully connected neuron network (FNN) to establish an RNN-FNN model for estimating forest biomass with Landsat time-series imagery and airborne LiDAR data. We compared the proposed model with the commonly used Random Forest model and linear regression model which are implemented with single-date data. The results show that the RNN-FNN model can deal with Landsat time-series sequence data to enhance the relationship between Landsat spectral reflectance and forest aboveground biomass. The proposed model achieves the R2 of 0.63 and RMSE of 25.5 Mg/ha, which significantly outperformed the Random Forest model and linear regression model with Landsat single-date data. This study demonstrates the value of RNN and Landsat time-series imagery in estimating forest biomass for tropical forests.

Impact assessment of agricultural droughts on water use efficiency in different climatic regions of Punjab Province Pakistan using MODIS time series imagery

AbstractDrought is the most destructive phenomenon that distresses the terrestrial carbon cycle balance and crop production. The variation in evapotranspiration (ET) and gross primary productivity (GPP) is a significant cause of agricultural drought effects on water use efficiency. This study aims to evaluate the impact of agricultural drought on WUE and it's anomalies in different climate regions. The standard vegetation index was used to measure the extent of agricultural drought. WUE was calculated using the ratio of ET, GPP, and climate classification using the De Martonne method. This study was conducted over the last 22 years, from 2001 to 2022. Meanwhile, 2001, 2002, 2014, and 2018 were considered high drought years based on a 22‐year analysis. According to remote sensing analysis of ET and GPP, the WUE increased throughout drought in all climate regions and more strongly in the arid zone than in humid regions. Humid areas were more vital due to drought than arid ones. WUE anomalies badge with increased drought severity across all climates except the very humid zone. Humid climates saw faster recovery times than arid ones, and areas experienced severe droughts. The findings of this research are essential for understanding the carbon and water cycles for agriculture management. The analysis of drought severity and WUE across arid climate regions helped analyse the varying impact of drought on WUE due to climate change. The significance of this study includes informing agricultural, water resource, and drought management planning in the Punjab Province, an agricultural region vulnerable to climate change. This study holds important learnings for agricultural regions worldwide. It has practical and scientific importance regarding agricultural systems' specific stresses and responses to droughts.

Detection of emerald ash borer damage using an improved change detection method: Integrating host phenology and pest life history

Invasive Emerald Ash Borer (EAB) damage pose significant challenges for sustainable forest management, necessitating accurate mapping of damaged ash trees. Traditional change detection methods, using time-series imagery, are essential for monitoring forest disturbances but complicated by abnormal fluctuations in original time-series features. Tree phenology also complicates this process by masking the reflectance characteristics indicative of EAB infestation. To address these challenges, we propose an improved change detection method integrating patterns from host tree phenology and EAB life history. This improved method includes: (1) select the indices with time stability to enhance detection reliability by partial least squares method (PLS); (2) correction on negative change values before and positive change values after the phenological peak based on known patterns of tree phenology and EAB life history. Result confirms that this method effectively reflects the seasonal growth and decline dynamics of ash trees, revealing the impacts of phenology and EAB infestation. EAB-damaged trees exhibited slower growth in May and premature decline in July compared with healthy tree, with the damage severity influencing the rate of leaf decline. This proposed method achieved an overall accuracy of 53.4%-76.7% across different months for ash trees with health, light and severe damage. This study highlights the capabilities of integrating pest life history and phenology in change detection method and provide a new method to monitor individual tree health across large areas by high-resolution satellite imagery.

Exploring land use dynamics in rural areas. An analysis of eight cases in the Global North

Rural areas are pivotal in the global economy, hence rural development planners and decision makers face significant technical and information management challenges, hindering their ability to assess sustainable development and formulate strategies. These challenges, coupled with shifts in traditional rural activities, aging populations, and climate change impacts, create inequalities and threaten rural communities. Therefore, developing approaches for the assessment of changes in rural areas are required. In recent years, remote sensing has emerged as a useful tool for monitoring the complex dynamics of rural regions, with a notable increase in the accessibility and use of freely available satellite data and platforms for tracking changes in land use and land cover (LULC). In light of this trend, this article analyzes the spatio-temporal trends in eight different development regions in Japan, Germany, Poland and Spain and integrates them with population data. This research relied on the utilization of openly available datasets - satellite imagery from Landsat and WorldPop population data, processed in Google Earth Engine (GEE). The time series of Landsat imagery were employed to analyze changes in LULC over a 20-year period from 2000 to 2020 in each test site. Unique combinations (i.e. each pixel history) were systematically examined and categorized based on established rules, with improbable scenarios marked as "uncertain." The comparison demonstrates that all eight regions have unique spatial, temporal and thematic characteristics. The most significant changes observed include the transformation of tree-covered areas into built-up areas in the case of Tsukuba and into agricultural lands in Shibushi. In the Bodzentyn and Połaniec communes in Poland, agricultural lands have transitioned into forested areas. In Spain, there is an increase in built-up areas, particularly at the outskirts of established towns. In Germany, the results indicate an expansion of built-up and agricultural areas in the Fürth region, while the Regen region experienced an increase in agricultural land. Incorporating population data reveals relationships between different types of LULC conversions and population trends, showing clear positive trends in areas transformed from agricultural and tree-covered land to built-up areas, with the most profound changes observed in Spanish cases and Tsukuba. The findings from automatic LULC classification highlight the usefulness of this tool in identifying socio-economic changes, especially when it is integrated with and compared to statistical data. Policy makers and planners at local levels should consider these specific values as a foundational reference point for shaping their plans related to sustainable development, land use, environmental preservation, and resource management.

A temporal-spatial deep learning network for winter wheat mapping using time-series Sentinel-2 imagery

Accurate mapping of winter wheat provides essential information for food security and ecosystem protection. Deep learning approaches have achieved promising crop discrimination performance based on multitemporal satellite imagery. However, due to the high dimensionality of the data, sequential relations, and complex semantic information in time-series imagery, effective methods that can automatically capture temporal-spatial features with high separability and generalizability have received less attention. In this study, we proposed a U-shaped CNN-Transformer hybrid framework based on an attention mechanism, named the U-Temporal-Spatial-Transformer network (UTS-Former), for winter wheat mapping using Sentinel-2 imagery. This model includes an “encoder-decoder” structure for multiscale information mining of time series images and a temporal-spatial transformer module (TST) for learning comprehensive temporal sequence features and spatial semantic information. The results obtained from two study areas indicated that our UTS-Former achieved the best accuracy, with a mean MCC of 0.928 and an F1-score of 0.950, and the results of different band combinations also showed better performance than other popular time-series methods. We found that the MCC (MCC/All) of the UTS-Former using only RGB bands decreased by 4.53 %, while it decreased by 13.36 % and 35.02 % for UNet2d-LSTM and CNN-BiLSTM, respectively, compared with that of all the band combinations. The comparison demonstrated that the proposed UTS-Former could capture more global temporal-spatial information from winter wheat fields and achieve greater precision in terms of local details than other methods, resulting in high-quality mapping. The analysis of attention scores for the available acquisition dates revealed significant contributions of both beginning and ending growth images in winter wheat mapping, which is valuable for making appropriate selections of image dates. These findings suggest that the proposed approach has great potential for accurate, cost-effective, and high-quality winter wheat mapping.

Imagery Time Series Cloud Removal and Classification Using Long Short Term Memory Neural Networks

Imagery Time Series Cloud Removal and Classification Using Long Short Term Memory Neural Networks

Thermal imaging can reveal variation in stay-green functionality of wheat canopies under temperate conditions.

Canopy temperature (CT) is often interpreted as representing leaf activity traits such as photosynthetic rates, gas exchange rates, or stomatal conductance. This interpretation is based on the observation that leaf activity traits correlate with transpiration which affects leaf temperature. Accordingly, CT measurements may provide a basis for high throughput assessments of the productivity of wheat canopies during early grain filling, which would allow distinguishing functional from dysfunctional stay-green. However, whereas the usefulness of CT as a fast surrogate measure of sustained vigor under soil drying is well established, its potential to quantify leaf activity traits under high-yielding conditions is less clear. To better understand sensitivity limits of CT measurements under high yielding conditions, we generated within-genotype variability in stay-green functionality by means of differential short-term pre-anthesis canopy shading that modified the sink:source balance. We quantified the effects of these modifications on stay-green properties through a combination of gold standard physiological measurements of leaf activity and newly developed methods for organ-level senescence monitoring based on timeseries of high-resolution imagery and deep-learning-based semantic image segmentation. In parallel, we monitored CT by means of a pole-mounted thermal camera that delivered continuous, ultra-high temporal resolution CT data. Our results show that differences in stay-green functionality translate into measurable differences in CT in the absence of major confounding factors. Differences amounted to approximately 0.8°C and 1.5°C for a very high-yielding source-limited genotype, and a medium-yielding sink-limited genotype, respectively. The gradual nature of the effects of shading on CT during the stay-green phase underscore the importance of a high measurement frequency and a time-integrated analysis of CT, whilst modest effect sizes confirm the importance of restricting screenings to a limited range of morphological and phenological diversity.

Improved Wetland Mapping of a Highly Fragmented Agricultural Landscape Using Land Surface Phenological Features

Wetlands are integral components of agricultural landscapes, providing a wide range of ecological, economic, and social benefits essential for sustainable development and rural livelihoods. Globally, they are vulnerable ecological assets facing several significant threats including water extraction and regulation, land clearing and reclamation, and climate change. Classification and mapping of wetlands in agricultural landscapes is crucial for conserving these ecosystems to maintain their ecological integrity amidst ongoing land-use changes and environmental pressures. This study aims to establish a robust framework for wetland classification and mapping in intensive agricultural landscapes using time series of Sentinel-2 imagery, with a focus on the Gwydir Wetland Complex situated in the northern Murray–Darling Basin—Australia’s largest river system. Using the Google Earth Engine (GEE) platform, we extracted two groups of predictors based on six vegetation indices time series calculated from multi-temporal Sentinel-2 surface reflectance (SR) imagery: the first is statistical features summarizing the time series and the second is phenological features based on harmonic analysis of time series data (HANTS). We developed and evaluated random forest (RF) models for each level of classification with combination of different groups of predictors. Our results show that RF models involving both HANTS and statistical features perform strongly with significantly high overall accuracy and class-weighted F1 scores (p < 0.05) when comparing with models with either statistical or HANTS variables. While the models have excellent performance (F-score greater than 0.9) in distinguishing wetlands from other landcovers (croplands, terrestrial uplands, and open waters), the inter-class discriminating power among wetlands is class-specific: wetlands that are frequently inundated (including river red gum forests and wetlands dominated by common reed, water couch, and marsh club-rush) are generally better identified than the ones that are flooded less frequently, such as sedgelands and woodlands dominated by black box and coolabah. This study demonstrates that HANTS features extracted from time series Sentinel data can significantly improve the accuracy of wetland mapping in highly fragmentated agricultural landscapes. Thus, this framework enables wetland classification and mapping to be updated on a regular basis to better understand the dynamic nature of these complex ecosystems and improve long-term wetland monitoring.

A Comprehensive Examination of the Medvezhiy Glacier’s Surges in West Pamir (1968–2023)

The Vanj River Basin contains a dynamic glacier, the Medvezhiy glacier, which occasionally poses a danger to local residents due to its surging, flooding, and frequent blockages of the Abdukahor River, leading to intense glacial lake outburst floods (GLOF). This study offers a new perspective on the quantitative assessment of glacier surface velocities and associated lake changes during six surges from 1968 to 2023 by using time-series imagery (Corona, Hexagon, Landsat), SRTM elevation maps, ITS_LIVE, unmanned aerial vehicles, local climate, and glacier surface elevation changes. Six turbulent periods (1968, 1973, 1977, 1989–1990, 2001, and 2011) were investigated, each lasting three years within a 10–11-year cycle. During inactive phases, a reduction in the thickness of the glacier tongue in the ablation zone occurred. During a surge in 2011, the flow accelerated, creating an ice dam and conditions for GLOF. Using these datasets, we reconstructed the process of the Medvezhiy glacier surge with high detail and identified a clear signal of uplift in the surface above the lower glacier tongue as well as a uniform increase in velocities associated with the onset of the surge. The increased activity of the Medvezhiy glacier and seasonal fluctuations in surface runoff are closely linked to climatic factors throughout the surge phase, and recent UAV observations indicate the absence of GLOFs in the glacier’s channel. Comprehending the processes of glacier movements and related changes at a regional level is crucial for implementing more proactive measures and identifying appropriate strategies for mitigation.

Fire Vulnerability, Resilience, and Recovery Rates of Mediterranean Pine Forests Using a 33-Year Time Series of Satellite Imagery

The modification of fire regimes and their impact on vegetation recovery, soil properties, and fuel structure are current key research areas that attempt to identify the thresholds of vegetation’s susceptibility to wildfires. This study aimed to evaluate the vulnerability of Mediterranean pine forests (Pinus halepensis Mill. and Pinus pinaster Aiton) to wildfires, analyzing two major forest fires that occurred in Yeste (Spain) in 1994 and 2017, affecting over 14,000 and 3200 hectares, respectively. Four recovery regions were identified based on fire severity—calculated using the delta Normalized Burn Ratio (dNBR) index—and recurrence: areas with high severity in 2017 but not in 1994 (UB94-HS17), areas with high severity in 1994 but not in 2017 (HS94-UB17), areas with high severity in both fires (HS94-HS17), and areas unaffected by either fire (UB94-UB17). The analysis focused on examining the recovery patterns of three spectral indices—the Normalized Difference Vegetation Index (NDVI), Normalized Moisture Index (NDMI), and Normalized Burn Ratio (NBR)—using the Google Earth Engine platform from 1990 to 2023. Additionally, the Relative Recovery Indicator (RRI), the Ratio of Eighty Percent (R80P), and the Year-on-Year average (YrYr) metrics were computed to assess the spectral recovery rates by region. These three spectral indices showed similar dynamic responses to fire. However, the Mann–Kendall and unit root statistical tests revealed that the NDVI and NDMI exhibited distinct trends, particularly in areas with recurrence (HS94-HS17). The NDVI outperformed the NBR and NDMI in distinguishing variations among regions. These results suggest accelerated vegetation spectral regrowth in the short term. The Vegetation Recovery Capacity After Fire (VRAF) index showed values from low to moderate, while the Vulnerability to Fire (V2FIRE) index exhibited values from medium to high across all recovery regions. These findings enhance our understanding of how vegetation recovers from fire and how vulnerable it is to fire.

Assessing the Impacts of Landuse-Landcover (LULC) Dynamics on Groundwater Depletion in Kabul, Afghanistan’s Capital (2000–2022): A Geospatial Technology-Driven Investigation

Land use/land cover (LULC) changes significantly impact spatiotemporal groundwater levels, posing a challenge for sustainable water resource management. This study investigates the long-term (2000–2022) influence of LULC dynamics, particularly urbanization, on groundwater depletion in Kabul, Afghanistan, using geospatial techniques. A time series of Landsat imagery (Landsat 5, 7 ETM+, and 8 OLI/TIRS) was employed to generate LULC maps for five key years (2000, 2005, 2010, 2015, and 2022) using a supervised classification algorithm based on Support Vector Machines (SVMs). Our analysis revealed a significant expansion of urban areas (70%) across Kabul City between 2000 and 2022, particularly concentrated in Districts 5, 6, 7, 11, 12, 13, 15, 17, and 22. Urbanization likely contributes to groundwater depletion through increased population growth, reduced infiltration of precipitation, and potential overexploitation of groundwater resources. The CA-Markov model further predicts continued expansion in built-up areas over the next two decades (2030s and 2040s), potentially leading to water scarcity, land subsidence, and environmental degradation in Kabul City. The periodic assessment of urbanization dynamics and prediction of future trends are considered the novelty of this study. The accuracy of the generated LULC maps was assessed for each year (2000, 2005, 2010, 2015, and 2022), achieving overall accuracy values of 95%, 93.8%, 85%, 95.6%, and 93%, respectively. These findings provide a valuable foundation for the development of sustainable management strategies for Kabul’s surface water and groundwater resources, while also guiding future research efforts.

Early detection of bark beetle infestation using UAV-borne multispectral imagery: a case study on the spruce forest in the Czech Republic

Over the last decade, biotic disturbances caused by bark beetles have represented a serious environmental and economic issue in Central Europe. Great efforts are expended on the early detection and management of bark beetle infestation. Our study analyses a time series of UAV-borne multispectral imagery of a 250-ha forest in the Vysočina region in the Czech Republic. The study site represents a typical European spruce forest with routine silvicultural management. UAV-borne data was acquired three times during the vegetation period, specifically (a) before swarming, (b) at the early stage of infestation, and (c) in the post-abandon phase, i.e., after most bark beetle offspring left the trees. The spectral reflectance values and vegetation indices calculated from orthorectified and radiometrically calibrated imageries were statistically analyzed by quadratic discriminant analysis (QDA). The study shows that healthy and infested trees could be distinguished at the early stage of infestation, especially using NIR-related vegetation indices (NDVI and BNDVI in our case). Detecting infested trees is more significant by vegetation indices than spectral bands and increases with the increasing time after infestation. The study verified the usability of UAV-borne multispectral imageries for early detection of bark beetle infestation at the level of individual trees. Thus, these methods can contribute to precise and effective forest management on a local level.

Monitoring the spatial–temporal distribution of invasive plant in urban water using deep learning and remote sensing technology

Despite water ecosystems being capable of sustaining biodiversity and enhancing the overall resilience of the urban environment, they are highly susceptible to biological invasions. Invasive aquatic plants (IAPs) threaten the natural environment by reducing the diversity of native aquatic plants and animal communities. Detecting IAPs and mapping their distribution is crucial for the protection of urban water ecosystems. This study is the first of its kind to use high-resolution unmanned aerial vehicle (UAV) imagery and deep learning approaches to monitor the expansion of Pistia stratiotes (water lettuce). A DJI Matrice 300 RTK was utilized to capture time-series images with a resolution of 0.018 m on the Guanyin Lake, Cangshan District, Fuzhzou, China, during an outbreak and subsequent rapid growth stage of water lettuce. Three deep learning model architectures and three backbones were combined to detect water lettuce. Model performance and the ability to generalize the model were evaluated to determine the optimal model for water lettuce detection from time-series high-resolution UAV imagery. Results show that the DeepLabv3 + model with ResNet-34 achieved superior performance in detecting water lettuce from time-series imagery, yielding an average accuracy of 90.24 % (85.33 %≤F1_score ≤ 96.54 %) for water lettuce detection on five different dates. For the UAV image acquired on September 26th, the U-Net model with ResNet-18 yielded the highest accuracy (F1_score = 92.46 %), but it was not the optimal model for multi-temporal water lettuce detection on subsequent dates. The distribution of water lettuce can have large variations at different times, with an average change rate of 49.50 % every two days and the highest change rate up to 60.33 %. The study demonstrates that the combination of UAV imagery and a deep learning model can achieve excellent accuracy for water lettuce monitoring and provide a method to map IAPs in dynamic urban water systems over time.