Land Cover Dataset Research Articles

Pazarcık (Kahramanmaraş) İlçesi Arazi Kullanımının Zamansal Değişimi (1990-2018)

Bu çalışmada Pazarcık ilçesinin coğrafi potansiyeli ve sahanın arazi kullanımının zamana bağlı değişimi CORİNE arazi örtüsü veri setleri kullanılarak belirlenmiş ve değerlendirilmiştir. Çalışmada literatür taraması, arazi gözlemleri yapılmış, yerel kamu kurum ve kuruluşlardan veriler temin edilmiştir. ArcGIS 10.5 programında haritalar yapılarak tablo ve grafik ile de desteklenmiştir. 1990-2018 yılları arasında ilçe genel olarak tarım alanları bakımından büyük değişimler göstermiştir. 1990 yılında 85.242 hektar olan tarım alanları 2018 yılında 70.787 hektara düşmüştür. İlçe arazisi üzerinde en büyük değişim ise 1990-2000 yılları arasındaki dönemde yaşanmıştır. Bu 10 yıllık süreçte tarım alanlarında 277.9 hektar bir azalma söz konusu iken, yapay alanlarda ise 372 hektarlık bir artış gözlenmiştir. En az değişim 2012-2018 yıllarını içine alan dönemde yaşanmıştır. Verimli tarım arazilerinin büyük ölçüde kentsel kullanıma sunulması olumsuz sonuçlara yol açabilmektedir. İlçe arazisinden istenilen düzeyde yarar sağlayabilmek için sürdürülebilir arazi kullanım yöntemlerinin uygulanması gerekmektedir. Özellikle yerel kurum ve kuruluşların desteğiyle hâlihazırdaki tarım arazilerinin iskâna açılması engellenebilir.

The urgent need to develop a new grassland map in China: based on the consistency and accuracy of ten land cover products.

Grasslands are the most dominant terrestrial ecosystem in China, but few national grassland maps have been generated. The grassland resource map produced in the 1980s is widely used as background data, but it has not been updated for almost 40 years. Therefore, a reliable map depicting the current spatial distribution of grasslands across the country is urgently needed. In this study, we evaluated the grassland consistency and accuracy of ten land cover datasets (GLC2000, GlobCover, CCI-LC, MCD12Q1, CLUD, GlobeLand30, GLC-FCS30, CGLS-LC100, CLCD, and FROM-GLC) for 2000, 2010, and 2020 based on extensive fieldwork. We concluded that the area of these ten grassland products ranges from 107.80×104 to 332.46×104 km2, with CLCD and MCD12Q1 having the highest area consistency. The spatial and sample consistency is highest in the regions of east-central Inner Mongolia, the Qinghai-Tibet Plateau and northern Xinjiang, while the distribution of southern grasslands is scattered and differs considerably among the ten products. MCD12Q1 is significantly more accurate than the other nine products, with an overall accuracy (OA) reaching 77.51% and a kappa coefficient of 0.51; CLCD is slightly less accurate than MCD12Q1 (OA=73.02%, kappa coefficient=0.45) and is more conducive to the fine monitoring and management of grassland because of its 30-meter resolution. The highest accuracy of grassland was found in the Inner Mongolia-Ningxia region and Qinghai-Tibet Plateau, while the accuracy was worst in the southeastern region. In the future grassland mapping, cartographers should improve the accuracy of the grassland distribution in South China and regions where grassland is confused with forest, cropland and bare land. We specify the availability of valuable data in existing land cover datasets for China's grasslands and call for researchers and the government to actively produce a new generation of grassland maps.

Assessment of Spatial-Temporal Changes of Landscape Ecological Risk in Xishuangbanna, China from 1990 to 2019

Xishuangbanna is a major natural rubber and tea production base in China and a national nature reserve with the best-preserved tropical ecosystem. However, the extensive exploitation and use of land resources impact the land use/land cover (LULC) and the processes of regional landscape ecology, further causing a battery of ecological and environmental problems. It is necessary to evaluate landscape ecological risk objectively and quantitatively for improving the ecological environment and maintaining ecological balance. First, this study selected China Land Cover Dataset (CLCD) to analyze the changes in LULC. Second, we constructed the landscape ecological risk index (ERI) using LULC changes based on Google Earth Engine (GEE) platform. Third, the spatial-temporal pattern and spatial autocorrelation of landscape ecological risk were assessed in our study area. The results showed that the significant change in LULC was that the areas of cropland increased, and the areas of forests decreased during 1990–2019; the forests of a total area of 859.93 km2 were transferred to croplands. The landscape ecological risk kept a low and stable level from 1990 to 2019, more than 75% of the study area remained at the lower or lowest risk level, and in about 70% of the total study area, the ERI level maintained stability. In addition, the landscape ecological risk of the Xishuangbanna increased during 1990–2010 and decreased during 2010–2019. The ecological risk was a significant spatial autocorrelation and has been an aggregation trend in space from 1990 to 2019. Our research can identify key risk areas and provide a reference for the management and sustainable use of land resources, which promotes the understanding of landscape ecological risk and sustainable development of the ecological environment.

Global 10 m Land Use Land Cover Datasets: A Comparison of Dynamic World, World Cover and Esri Land Cover

The European Space Agency’s Sentinel satellites have laid the foundation for global land use land cover (LULC) mapping with unprecedented detail at 10 m resolution. We present a cross-comparison and accuracy assessment of Google’s Dynamic World (DW), ESA’s World Cover (WC) and Esri’s Land Cover (Esri) products for the first time in order to inform the adoption and application of these maps going forward. For the year 2020, the three global LULC maps show strong spatial correspondence (i.e., near-equal area estimates) for water, built area, trees and crop LULC classes. However, relative to one another, WC is biased towards over-estimating grass cover, Esri towards shrub and scrub cover and DW towards snow and ice. Using global ground truth data with a minimum mapping unit of 250 m2, we found that Esri had the highest overall accuracy (75%) compared to DW (72%) and WC (65%). Across all global maps, water was the most accurately mapped class (92%), followed by built area (83%), tree cover (81%) and crops (78%), particularly in biomes characterized by temperate and boreal forests. The classes with the lowest accuracies, particularly in the tundra biome, included shrub and scrub (47%), grass (34%), bare ground (57%) and flooded vegetation (53%). When using European ground truth data from LUCAS (Land Use/Cover Area Frame Survey) with a minimum mapping unit of <100 m2, we found that WC had the highest accuracy (71%) compared to DW (66%) and Esri (63%), highlighting the ability of WC to resolve landscape elements with more detail compared to DW and Esri. Although not analyzed in our study, we discuss the relative advantages of DW due to its frequent and near real-time data delivery of both categorical predictions and class probability scores. We recommend that the use of global LULC products should involve critical evaluation of their suitability with respect to the application purpose, such as aggregate changes in ecosystem accounting versus site-specific change detection in monitoring, considering trade-offs between thematic resolution, global versus. local accuracy, class-specific biases and whether change analysis is necessary. We also emphasize the importance of not estimating areas from pixel-counting alone but adopting best practices in design-based inference and area estimation that quantify uncertainty for a given study area.

Assessing Multi-Temporal Global Urban Land-Cover Products Using Spatio-Temporal Stratified Sampling

In recent years, the availability of multi-temporal global land-cover datasets has meant that they have become a key data source for evaluating land cover in many applications. Due to the high data volume of the multi-temporal land-cover datasets, probability sampling is an efficient method for validating multi-temporal global urban land-cover maps. However, the current accuracy assessment methods often work for a single-epoch dataset, and they are not suitable for multi-temporal data products. Limitations such as repeated sampling and inappropriate sample allocation can lead to inaccurate evaluation results. In this study, we propose the use of spatio-temporal stratified sampling to assess thematic mappings with respect to the temporal changes and spatial clustering. The total number of samples in the two stages, i.e., map and pixel, was obtained by using a probability sampling model. Since the proportion of the area labeled as no change is large while that of the area labeled as change is small, an optimization algorithm for determining the sample sizes of the different strata is proposed by minimizing the sum of variance of the user’s accuracy, producer’s accuracy, and proportion of area for all strata. The experimental results show that the allocation of sample size by the proposed method results in the smallest bias in the estimated accuracy, compared with the conventional sample allocation, i.e., equal allocation and proportional allocation. The proposed method was applied to multi-temporal global urban land-cover maps from 2000 to 2010, with a time interval of 5 years. Due to the spatial aggregation characteristics, the local pivotal method (LPM) is adopted to realize spatially balanced sampling, leading to more representative samples for each stratum in the spatial domain. The main contribution of our research is the proposed spatio-temporal sampling approach and the accuracy assessment conducted for the multi-temporal global urban land-cover product.

Remarkable improvement of ecosystem service values promoted by land use/land cover changes on the Yungui Plateau of China during 2001–2020

Quantitative evaluation of the relationship between land use/land cover (LULC) change and ecosystem service values (ESVs) is of great significance for the scientific optimization of land use structure and the formulation of sustainable land use policy, However, the responses of ESVs to LULC changes in fragile ecosystem regions over long time scales and large spatial scopes in the context of ecological restoration measures have seldom been studied. In this study, the Yungui Plateau in China was used as the study region. Based on annual Moderate–Resolution Imaging Spectroradiometer (MODIS) land cover datasets (MCD12Q1) from 2001 to 2020, using the equivalent coefficient value of ESVs modified for the specific situation of China, and adopting the Getis-Ord Gi* statistic method and Spearman’s rank correlation analysis, the study estimated the values of 11 types of ecosystem services, and analyzed the dynamics of LULC and its impact on ESVs from 2001 to 2020. The results showed that the LULC pattern in the Yungui Plateau changed dramatically in the past two decades, with a significant increase in forestland. The total ESVs improved remarkably from 638.10 billion yuan in 2001 to 746.99 billion yuan in 2020, and was mainly affected by forest restoration and the conversion of grassland to forestland and farmland to grassland. The distribution of ESVs shows obvious spatial and temporal heterogeneity; ESV hot spots were mainly distributed in the edge regions, while ESV cold spots were concentrated on the central and eastern of Yungui Plateau, and the decline in total ESVs still existed and displayed an increasing trend caused by farmland expansion and forestland degradation. Thus, we conclude that the LULC changes caused by ecological restoration measures have a significant positive effect on the improvement of the total ESVs in the Yungui Plateau and suggest that declines in ESVs in local areas should be of high concern in future land use planning and management.

Urban Form and Household Energy Consumption: Evidence from China Panel Data

Compact urban form is of great importance to facilitate low carbon development, while little empirical evidence was found about the impact of urban geometric form on household energy consumption with panel data; this paper uses a multi-time China land cover dataset to calculate the urban form of 253 prefecture-level cities for the years 2000, 2010 and 2020 and examines its impact on urban household energy consumption. We use urban land ruggedness as the instrumental variable for urban form. The results show urban form and household energy consumption are negatively related. The result is robust to several alternative specifications, including measuring the household energy consumption without central heating or controlling extreme weather year effect. Mechanism analysis shows compact geometric form reduces commuting time in larger and medium cities.

Linear and Non-Linear Vegetation Trend Analysis throughout Iran Using Two Decades of MODIS NDVI Imagery

Vegetation is the main component of the terrestrial Earth, and it plays an imperative role in carbon cycle regulation and surface water/energy exchange/balance. The coupled effects of climate change and anthropogenic forcing have undoubtfully impacted the vegetation cover in linear/non-linear manners. Considering the essential benefits of vegetation to the environment, it is vital to investigate the vegetation dynamics through spatially and temporally consistent workflows. In this regard, remote sensing, especially Normalized Difference Vegetation Index (NDVI), has offered a reliable data source for vegetation monitoring and trend analysis. In this paper, two decades (2000 to 2020) of Moderate Resolution Imaging Spectroradiometer (MODIS) NDVI datasets (MOD13Q1) were used for vegetation trend analysis throughout Iran. First, the per-pixel annual NDVI dataset was prepared using the Google Earth Engine (GEE) by averaging all available NDVI values within the growing season and was then fed into the PolyTrend algorithm for linear/non-linear trend identification. In total, nearly 14 million pixels (44% of Iran) were subjected to trend analysis, and the results indicated a higher rate of greening than browning across the country. Regarding the trend types, linear was the dominant trend type with 14%, followed by concealed (11%), cubic (8%), and quadratic (2%), while 9% of the vegetation area remained stable (no trend). Both positive and negative directions were observed in all trend types, with the slope magnitudes ranging between −0.048 and 0.047 (NDVI units) per year. Later, precipitation and land cover datasets were employed to further investigate the vegetation dynamics. The correlation coefficient between precipitation and vegetation (NDVI) was 0.54 based on all corresponding observations (n = 1785). The comparison between vegetation and precipitation trends revealed matched trend directions in 60% of cases, suggesting the potential impact of precipitation dynamics on vegetation covers. Further incorporation of land cover data showed that grassland areas experienced significant dynamics with the highest proportion compared to other vegetation land cover types. Moreover, forest and cropland had the highest positive and negative trend direction proportions. Finally, independent (from trend analysis) sources were used to examine the vegetation dynamics (greening/browning) from other perspectives, confirming Iran’s greening process and agreeing with the trend analysis results. It is believed that the results could support achieving Sustainable Development Goals (SDGs) by serving as an initial stage study for establishing conservation and restoration practices.

Suburbanization within City Limits in Hungary—A Challenge for Environmental and Social Sustainability

Suburbanization is one of the most prominent processes of post-socialist urban development, leading to the deconcentration of people, capital, as well as productive and non-productive activities within the functional urban area. This phenomenon also has a significant impact on the traditional rural landscape and leads to environmental and social sustainability challenges. Outmigration from the city center to the rural municipalities of the agglomeration ring is already a thoroughly studied topic. However, less attention is given to migration processes not crossing municipal borders. In Hungary, a significant fraction of them is driven by similar motivations as “traditional” suburbanization. Such movements include flows to built-up residential areas that are physically separated from the urban core. Due to their peculiar development path, the inhabited outskirts can also become destinations for within-city migration. This kind of mobility can be considered suburbanization within the city limits. Because of the general lack of data, this phenomenon is seldom researched. However, this study attempts to address this gap. The prevalence of this process in Hungary is explored by analyzing national statistical data sources. Four case studies are selected for detailed examination—Győr, Zalaegerszeg, Kecskemét, and Szeged. GIS methods, field examinations, surveys, and expert interviews are used to get a detailed picture of the demographic and land cover change processes, as well as the distinctions between the destination areas of the case studies. A comparison of the results drawn from the different methods reveals that land use change in the study area is more widespread than what the land cover datasets indicate. The findings indicate that the Corine Land Cover categories describing mixed land use—especially complex cultivation patterns—are not able to capture the drastic function shift caused by intense suburbanization. Different environmental and social sustainability issues are identified depending on the economic status of the outmigrants. This paper also sheds some light on the urban planning considerations of this emerging challenge.

An International Library for Land Cover Legends: The Land Cover Legend Registry

Information on land cover is vital to numerous United Nations (UN) missions, including achieving the Sustainable Development Goals (SDGs). Because land cover data are developed by a variety of organizations for a range of objectives, they are based on different classification schemes and have discrepancies. In addition, the sustainability for land cover is hampered by limited access to information and documentation. Accordingly, international standards for land cover are developed to improve interoperability between different land cover datasets. However, the use and development of land cover datasets are limited by various factors including availability of properly documented land cover legends in support of different applications including change assessment, comparison, and international reporting. The purpose of this article is to highlight the importance of land cover in achieving several goals and to introduce the first international platform for land cover legend, named Land Cover Legend Registry (LCLR). This registry is a contribution to the international land cover community and the UN in effort to promote and support data harmonization processes and interoperability from local to global level, and vice versa. Users can not only use the registry for preparing consistent datasets, but also contribute to it by providing the latest data to ensure the long-term availability of both updated and existing datasets around the world. Moreover, building on the experience developing land cover legends with different nations, a brief explanation on the preparation of legends is also provided. Additionally, it is more important than ever to develop land cover registers to support the use, expansion, integration, and use uptake of land cover data, particularly for innovative remote sensing, machine learning, and information and communication technologies and techniques that build on existing and national contexts.

Comparing CNNs and Random Forests for Landsat Image Segmentation Trained on a Large Proxy Land Cover Dataset

Land cover mapping from satellite images has progressed from visual and statistical approaches to Random Forests (RFs) and, more recently, advanced image recognition techniques such as convolutional neural networks (CNNs). CNNs have a conceptual benefit over RFs in recognising spatial feature context, but potentially at the cost of reduced spatial detail. We tested the use of CNNs for improved land cover mapping based on Landsat data, compared with RFs, for a study area of approximately 500 km × 500 km in southeastern Australia. Landsat 8 geomedian composite surface reflectances were available for 2018. Label data were a simple nine-member land cover classification derived from reference land use mapping (Catchment Scale Land Use of Australia—CLUM), and further enhanced by using custom forest extent mapping (Forests of Australia). Experiments were undertaken testing U-Net CNN for segmentation of Landsat 8 geomedian imagery to determine the optimal combination of input Landsat 8 bands. The results were compared with those from a simple autoencoder as well as an RF model. Segmentation test results for the best performing U-Net CNN models produced an overall accuracy of 79% and weighted-mean F1 score of 77% (9 band input) or 76% (6 band input) for a simple nine-member land cover classification, compared with 73% and 68% (6 band input), respectively, for the best RF model. We conclude that U-Net CNN models can generate annual land cover maps with good accuracy from proxy training data, and can also be used for quality control or improvement of existing land cover products.

Reforestation in Southern China Enhances the Convective Afternoon Rainfall During the Post-flood Season

Dynamic and continuous land use and cover change (LUCC) is one of the external forcing factors affecting regional climate in China. Based on the annual dynamic global land cover dataset derived from the Global Land Surface Satellite (GLASS-GLC), this paper modelled and investigated the effects of annual transient LUCC on precipitation over China using the regional climate model RegCM4.8.2. Analysis of the GLASS-GLC data revealed that considerable conversions of cropland to forest, grassland to cropland and bare ground to grassland had occurred during the last 3 decades (1984–2013) in southern, northeastern and northwestern China respectively. By comparing the differences between the two sets of experiments under the fixed LUCC and dynamic LUCC scenarios, the results showed that reforestation in southern China during the past 3 decades significantly enhanced local convective afternoon rainfall (CAR; ∼2 mm/day) during the post-flood season (POF; July-August-September). This reforestation effect on CAR also increased with the intensity of LUCC. However, the realistic LUCC effects were weak and negligible for other periods, regions and large-scale precipitation. Furthermore, we have identified two possible reasons that favored the occurrence of POFCAR, namely that the moisture and instability conditions required to trigger convections were both enhanced by reforestation. This was evidenced by increases (decreases) in convective available potential energy (lifting condensation level), increases in atmospheric water vapor content and declining tendencies of equivalent potential temperature with height. Overall, this study highlights the importance of reforestation impacts on the diurnal variations of the precipitation.

Precipitation Trends Analysis Using Gridded Dynamic Sampling Zones: Case Study Yangtze Delta Megalopolis

As a result of the fast growth of remote sensing and data assimilation technology, many global land use land cover (LULC) and climate reanalysis data sets have been used to advance our understanding of climate and environmental change. This paper investigates the precipitation variations of the Yangtze Delta Megalopolis by using precipitation reanalysis data under conditions of dynamic urban sprawl. Compared with current precipitation characteristic analyses, which are often based on a limited number of ground rainfall stations, the approach followed in this study comprises a grid-based statistical method using large sets of samples with a uniform distribution and a same representative grid area. This novel approach of dynamic sampling is applied in this study to overcome the temporal and spatial inconsistency of stationary sampling. This approach allows to examine the impact of urbanization on regional precipitation characteristics. The Yangtze Delta Megalopolis (YDM) region, one of the most developed regions in China, was selected as a case study to evaluate the impact of urbanization on subsequent precipitation features. The results reveal that the annual total precipitation (TP) and the maximum daily precipitation (MDP) in both urban and non-urban areas of the YDM region generally have increased during the past 30 years. Hence, the region has become increasingly humid. Extrema of annual MDP and TP show obvious spatial characteristics, in which most maxima are located in the southern part of YDM while minima are more concentrated in the northern part. This newly developed approach has potentials for application in studies where underlying surface features exhibit rapid alterations. The findings of this case study provide relevant information for planning and design of regional water resources management, flood risk management, and planning of the urban drainage system of the YDM region.

Machine Learning Approaches for Developing Land Cover Mapping.

In remote sensing data processing, cover classification on decimeter-level data is a well-studied but tough subject that has been well-documented. The majority of currently existent works make use of orthographic photographs or orthophotos and digital surface models that go with them (DSMs). Urban land cover classification plays a significant role in the field of remote sensing to enhance the quality of different applications including environment protection, sustainable development, and resource management and planning. Novelty of the research done in this area is focused on extracting features from high-resolution satellite images to be used in the classification process. However, it is well known in machine learning literature that some of the extracted features are irrelevant to the classification process with a negative or no effect on its accuracy. In this work, a genetic algorithm-based feature selection approach is used to enhance the performance of urban land cover classification. Neural networks (NNs) and random forest (RF) classifiers were used to evaluate the proposed approach on a recent urban land cover dataset of nine different classes. Experimental results show that the proposed approach achieved better performance with RF classifier using only 27% of the features. The random forest tree has achieved highest accuracy 84.27%; it is concluded that the RF algorithm is an appropriate algorithm for classifying cover land.

A Comparison of Raster-Based Forestland Data in Cropland Data Layer and the National Land Cover Database

The National Agricultural Statistics Service, the statistical arm of the US Department of Agriculture, and the Multi-Resolution Land Characteristics Consortium, a group of the US federal agencies, collect and publish several land-use and land-cover data sets. The aim of this study is to analyze the consistency of forestland estimates based on two widely used, publicly available products: the National Land-Cover Database (NLCD) and Cropland Data Layer (CDL). Both remote-sensing-based products provide raster-formatted land-cover categorization at a spatial resolution of 30 m. Although the processing of the yearly published CDL non-agricultural land-cover data is based on less frequently updated NLCD, the consistency of large-area forestland mapping between these two datasets has not been assessed. To assess the similarities and the differences between CDL- and NLCD-based forestland mappings for the state of North Carolina, we overlay the two data products for the years 2011 and 2016 in ArcMap 10.5.1 and analyze the location and attributes of the matched and mismatched forestland. We find that the mismatch is relatively smaller for the areas of the state where forests occupy larger shares of the total land, and that the relative mismatch is smaller in 2011 when compared to 2016. We also find that a large portion of the forestland mismatch is attributable to the dynamics of re-growth of periodically harvested and otherwise disturbed forests. Our results underscore the need for a holistic approach to data preparation, data attribution, and data accuracy when performing high-scale map-based analyses using each of these products.

Medium Spatial Resolution Mapping of Global Land Cover and Land Cover Change Across Multiple Decades From Landsat

Land cover maps are essential for characterizing the biophysical properties of the Earth’s land areas. Because land cover information synthesizes a rich array of information related to both the ecological condition of land areas and their exploitation by humans, they are widely used for basic and applied research that requires information related to land surface properties (e.g., terrestrial carbon models, water balance models, weather, and climate models) and are core inputs to models and analyses used by natural resource scientists and land managers. As the Earth’s global population has grown over the last several decades rates of land cover change have increased dramatically, with enormous impacts on ecosystem services (e.g., biodiversity, water supply, carbon sequestration, etc.). Hence, accurate information related to land cover is essential for both managing natural resources and for understanding society’s ecological, biophysical, and resource management footprint. To address the need for high-quality land cover information we are using the global record of Landsat observations to compile annual maps of global land cover from 2001 to 2020 at 30 m spatial resolution. To create these maps we use features derived from time series of Landsat imagery in combination with ancillary geospatial data and a large database of training sites to classify land cover at annual time step. The algorithm that we apply uses temporal segmentation to identify periods with stable land cover that are separated by breakpoints in the time series. Here we provide an overview of the methods and data sets we are using to create global maps of land cover. We describe the algorithms used to create these maps and the core land cover data sets that we are creating through this effort, and we summarize our approach to accuracy assessment. We also present a synthesis of early results and discuss the strengths and weaknesses of our early map products and the challenges that we have encountered in creating global land cover data sets from Landsat. Initial accuracy assessment for North America shows good overall accuracy (77.0 ± 2.0% correctly classified) and 79.8% agreement with the European Space Agency (ESA) WorldCover product. The land cover mapping results we report provide the foundation for robust, repeatable, and accurate mapping of global land cover and land cover change across multiple decades at 30 m spatial resolution from Landsat.

Evaluating the Accuracy and Spatial Agreement of Five Global Land Cover Datasets in the Ecologically Vulnerable South China Karst

Accurate and reliable land cover information is vital for ecosystem management and regional sustainable development, especially for ecologically vulnerable areas. The South China Karst, one of the largest and most concentrated karst distribution areas globally, has been undergoing large-scale afforestation projects to combat accelerating land degradation since the turn of the new millennium. Here, we assess five recent and widely used global land cover datasets (i.e., CCI-LC, MCD12Q1, GlobeLand30, GlobCover, and CGLS-LC) for their comparative performances in land dynamics monitoring in the South China Karst during 2000–2020 based on the reference China Land Use/Cover Database. The assessment proceeded from three aspects: areal comparison, spatial agreement, and accuracy metrics. Moreover, divergent responses of overall accuracy with regard to varying terrain and geomorphic conditions have also been quantified. The results reveal that obvious discrepancies exist amongst land cover maps in both area and spatial patterns. The spatial agreement remains low in the Yunnan–Guizhou Plateau and heterogeneous mountainous karst areas. Furthermore, the overall accuracy of the five datasets ranges from 40.3% to 52.0%. The CGLS-LC dataset, with the highest accuracy, is the most accurate dataset for mountainous southern China, followed by GlobeLand30 (51.4%), CCI-LC (50.0%), MCD12Q1 (41.4%), and GlobCover (40.3%). Despite the low overall accuracy, MCD12Q1 has the best accuracy in areas with an elevation above 1200 m or a slope greater than 25°. With regard to geomorphic types, accuracy in non-karst areas is evidently higher than in karst areas. Additionally, dataset accuracy declines significantly (p < 0.05) with an increase in landscape heterogeneity in the region. These findings provide useful guidelines for future land cover mapping and dataset fusion.

Short-time-series grassland mapping using Sentinel-2 imagery and deep learning-based architecture

In the present work, a deep learning-based network called LeNet is applied for accurate grassland map production from Sentinel-2 data for the Greater Sydney region, Australia. First, we apply the technique to the base date Sentinel-2 data (non-seasonal) to make the vegetation maps. Then, we combine short time-series (seasonal) data and enhanced vegetation index (EVI) information to the base date imagery to improve the classification results and generate high-resolution grassland maps. The proposed model obtained an overall accuracy (OA) of 88.36% for the mono-temporal data, and 92.74% for the multi-temporal data. The experimental products proved that, by combining the short time-series images and EVI information to the base date, the classification maps' accuracy is increased by 4.38%. Moreover, the Sentinel-2 produced grassland maps are compared with the pre-existing maps such as Australian Land Use and Management (ALUM) 50 m resolution and Dynamic Land Cover Dataset (DLCD) with 250 m resolution as well as some traditional machine learning methods such as Support Vector Machine (SVM) and Random Forest (RF). The results show the effect of the LeNet network's performance and efficiency for grassland map production from short time-series data. As a result, decision-makers and urban planners can benefit from this work in terms of grassland change identification, monitoring, and planning assessment.

Mapping global urban greenspace: An analysis based on open land-cover data

Urban greenspace plays a positive role in improving the urban living environment, and thus extensive studies have paid attention to the mapping of urban greenspace for various applications. Recently, open and high-resolution land-cover (LC) data have been viewed as essential sources for urban greenspace mapping. However, few studies have performed a global-scale analysis; moreover, as there are commonly different LC types in an LC dataset, few studies have investigated how the definition or selection of different LC types as greenspace impacts urban greenspace mapping. Therefore, this study aims to investigate the spatial pattern of urban greenspace at a global scale and the impact on selecting different LC types for urban greenspace mapping. We employed two global open datasets (FROM-GLC10 and GHS-UCDB) and identified 13,079 global urban areas. Various LC types that potentially relate to urban greenspace and three different selections of LC types were analyzed. Results show that: 1) The spatial pattern of global urban greenspace may vary with the use of different selections; 2) The impact of whether or not to select each LC type as urban greenspace relates to not only the LC type itself but also the geographic locations. The findings may benefit urban planners and designers to select proper LC type(s) for urban greenspace studies.

Forest Fragmentation and Fires in the Eastern Brazilian Amazon–Maranhão State, Brazil

Tropical forests provide essential environmental services to human well-being. In the world, Brazil has the largest continuous area of these forests. However, in the state of Maranhão, in the eastern Amazon, only 24% of the original forest cover remains. We integrated and analyzed active fires, burned area, land use and land cover, rainfall, and surface temperature datasets to understand forest fragmentation and forest fire dynamics from a remote sensing approach. We found that forest cover in the Maranhão Amazon region had a net reduction of 31,302 km2 between 1985 and 2017, with 63% of losses occurring in forest core areas. Forest edges extent was reduced by 38%, while the size of isolated forest patches increased by 239%. Forest fires impacted, on average, around 1031 ± 695 km2 year−1 of forest edges between 2003 and 2017, the equivalent of 60% of the total burned forest in this period. Our results demonstrated that forest fragmentation is an important factor controlling temporal and spatial variability of forest fires in the eastern Amazon region. Thus, both directly and indirectly, forest fragmentation can compromise biodiversity and carbon stocks in this Amazon region.