Classification of Subdivision Land Use and Land Cover Using Deep Learning Models

Land cover type is a crucial parameter that is required for various land surface models that simulate water and carbon cycles, ecosystem dynamics, and climate change. Many land use/land cover maps used in recent years have been derived from field investigations and remote-sensing observations. However, no land cover map that is derived from a single source (such as satellite observation) properly meets the needs of land surface simulation in China. This article presents a decision-fuse method to produce a higher-accuracy land cover map by combining multi-source local data based on the Dempster–Shafer (D–S) evidence theory. A practical evidence generation scheme was used to integrate multi-source land cover classification information. The basic probability values of the input data were obtained from literature reviews and expert knowledge. A Multi-source Integrated Chinese Land Cover (MICLCover) map was generated by combining multi-source land cover/land use classification maps including a 1:1,000,000 vegetation map, a 1:100,000 land use map for the year 2000, a 1:1,000,000 swamp-wetland map, a glacier map, and a Moderate-Resolution Imaging Spectroradiometer land cover map for China in 2001 (MODIS2001). The merit of this new map is that it uses a common classification system (the International Geosphere-Biosphere Programme (IGBP) land cover classification system), and it has a unified 1 km resolution. The accuracy of the new map was validated by a hybrid procedure. The validation results show great improvement in accuracy for the MICLCover map. The local-scale visual comparison validations for three regions show that the MICLCover map provides more spatial details on land cover at the local scale compared with other popular land cover products. The improvement in accuracy is true for all classes but particularly for cropland, urban, glacier, wetland, and water body classes. Validation by comparison with the China Forestry Scientific Data Center (CFSDC)–Forest Inventory Data (FID) data shows that overall forest accuracies in five provinces increased to between 42.19% and 88.65% for our MICLCover map, while those of the MODIS2001 map increased between 27.77% and 77.89%. The validation all over China shows that the overall accuracy of the MICLCover map is 71%, which is higher than the accuracies of other land cover maps. This map therefore can be used as an important input for land surface models of China. It has the potential to improve the modeling accuracy of land surface processes as well as to support other aspects of scientific land surface investigations in China.

Characterization of climate uncertainties due to different land cover maps in regional climate models is essential for adaptation strategies. The spatiotemporal heterogeneity in surface characteristics is considered to play a key role in terrestrial surface processes. Here, we quantified the sensitivity of model results to changes in land cover input data (GlobCover 2009, GLC 2000, CCI, and ECOCLIMAP) in the regional climate model (RCM) COSMO-CLM (v5.0_clm16). We investigated land cover changes due to the retrieval year, number, fraction and spatial distribution of land cover classes by performing convection-permitting simulations driven by ERA5 reanalysis data over Germany from 2002 to 2011. The role of the surface parameters on the surface turbulent fluxes and temperature is examined, which is related to the land cover classes. The bias of the annual temperature cycle of all the simulations compared with observations is larger than the differences between simulations. The latter is well within the uncertainty of the observations. The land cover class fractional differences are small among the land cover maps. However, some land cover types, such as croplands and urban areas, have greatly changed over the years. These distribution changes can be seen in the temperature differences. Simulations based on the CCI retrieved in 2000 and 2015 revealed no accreditable difference in the climate variables as the land cover changes that occurred between these years are marginal, and thus, the influence is small over Germany. Increasing the land cover types as in ECOCLIMAP leads to higher temperature variability. The largest differences among the simulations occur in maximum temperature and from spring to autumn, which is the main vegetation period. The temperature differences seen among the simulations relate to changes in the leaf area index, plant coverage, roughness length, latent and sensible heat fluxes due to differences in land cover types. The vegetation fraction was the main parameter affecting the seasonal evolution of the latent heat fluxes based on linear regression analysis, followed by roughness length and leaf area index. If the same natural vegetation (e.g. forest) or pasture grid cells changed into urban types in another land cover map, daily maximum temperatures increased accordingly. Similarly, differences in climate extreme indices are strongest for any land cover type change to urban areas. The uncertainties in regional temperature due to different land cover datasets were overall lower than the uncertainties associated with climate projections. Although the impact and their implications are different on different spatial and temporal scales as shown for urban area differences in the land cover maps. For future development, more attention should be given to land cover classification in complex areas, including more land cover types or single vegetation species and regional representative classification sample selection. Including more sophisticated urban and vegetation modules with synchronized input data in RCMs would improve the underestimation of the urban and vegetation effect on local climate.

Limitations of mapping land surface properties and their conversion into climate model boundary conditions are major sources of uncertainty in climate simulations. In this paper, the range of the largest possible uncertainty in satellite-derived land cover (LC) map is estimated and its impact on climate simulations is quantified with the Earth System Model of the Max-Planck Institute for Meteorology utilizing prescribed sea surface temperature and sea ice. Two types of uncertainty in the LC map are addressed: (i) uncertainty due to classification algorithm of spectral reflectance into LC classes, and (ii) uncertainty due to conversion of LC classes into the climate model vegetation distribution. For forest cover, each of them is about the same order of magnitude as the uncertainty range in recent observations (∼± 700 Mha). Superposing two sources of uncertainty results in LC maps that feature the range of vegetation deviation that is about the same order of magnitude as the recent (since year 1700) forest loss due to agriculture (forest cover uncertainty range ∼± 1700 Mha). These uncertainties in vegetation distribution lead to noticeable variations in near-surface climate variables, local, regional, and global climate forcing. Temperature does not show significant uncertainty in global mean, but rather exhibits regional deviations with an opposite response to LC uncertainty that compensate each other in the global mean (e.g., albedo feedback controls temperature in boreal North America resulting in cooling (warming) with decrease (increase) of vegetation while evaporative cooling controls temperature in South America and sub-Saharan Africa resulting in cooling (warming) with increase (decrease) of vegetation). Large-scale circulation is also affected by the LC uncertainty, and consequently precipitation pattern as well. It is demonstrated that precipitation uncertainty in the monsoonal regions are about the same order of magnitude as in previous studies with idealized perturbations of vegetation. These findings indicate that the range of uncertainty in satellite-derived vegetation maps for climate models is about the same order of magnitude as the uncertainty in recent observations of forest cover or as the forest lost due to agriculture. Consequently, climate simulations have a similar range of uncertainty in variables representing near-surface climate as the observed climate change due to land use. Hence, more accurate methods are needed for mapping and converting LC properties into model vegetation in order to increase reliability of climate model simulations.

Understanding grassland habitat dynamics in space and time is crucial for evaluating the effectiveness of protection measures and developing sustainable management practices, specifically within the Natura 2000 network and in light of the European Biodiversity Strategy. Land cover maps, derived from remote sensing data, are essential for understanding long-term changes in vegetation cover and land use and assessing the impact of land use changes on grassland ecosystems. In this study, we conducted a 20-year land cover analysis of grassland landscapes in Umbria, Italy, using Random Forest classifications of Landsat data in Google Earth Engine. Our analysis was based on the years 2000, 2010, and 2020. We integrated harmonic modeling, Gray-Level Co-occurrence Matrix (GLCM) textural analysis, statistical image and gradient analysis, and other spectral and Digital Terrain Model (DTM)-derived indices to enhance the classification capabilities. The LandTrendr (LT) algorithm was used in GEE to collect ground control points in no-change areas automatically. We used a method based on Multilayer Perceptron-Artificial Neural Networks (MLP-ANNs) to forecast 2040 land cover. Our land cover classifications and the scenario model validation achieved an overall accuracy of over 90%. However, the classification of shrublands proved challenging due to their mixed composition and unique spatial patterns, resulting in lower accuracies. Feature importance analysis demonstrated the value of the enhanced map composition, and applying the LandTrendr algorithm simplified the diachronic land use and land cover (LULC) classification and change analysis by supporting automatic training data collection. Results support the interpretation of grassland dynamics in Umbria over the past two decades and identify areas affected by encroachment from shrubs, woody plants, or those with reduced green biomass. The forecasting method along with the selection of spatial drivers to predict land cover change, demonstrated high efficiency compared to other studies. A specific analysis was developed to identify areas where conservation measures related to the Natura 2000 network have been more or less effective in preserving grasslands. Overall, the research provides a scientific foundation for a methodology helpful in informing policy decisions and defining spatially explicit management strategies to enhance grassland conservation inside and outside Natura 2000 areas.

This study investigates how a land cover map is produced using Sentinel 2 satellite images and how the resulting map is integrated with the calculation of hydrological parameters. Due to its capacity to deliver high-resolution spatial data, satellite imagery is increasingly being used for land cover mapping (1). The process involves several stages, including pre-processing of the Sentinel 2 imagery, image classification using machine learning algorithms, and post-processing of the classified image to generate a land cover map. The resulting land cover map is then integrated with hydrological parameters such as precipitation, evapotranspiration, and soil characteristics to calculate various hydrological parameters. Debrecen, north-east Hungary is the study area. This study's method is suitable for usage in other cities. Understanding the relationship between land cover and water availability requires the integration of the land cover map with hydrological parameters (2). The study demonstrates the potential of using remote sensing data for hydrological studies and highlights the importance of integrating various data sources for accurate estimation of hydrological parameters. The results show that the land cover classes have a significant impact on the water balance of urban sites. This study outlines the key steps involved in creating a land cover map using Sentinel 2 satellite imagery and integrating it with hydrological parameters calculation. The application of this strategy supports sustainable water management techniques and offers insightful information about a region's hydrological processes.

Classifying multi-temporal image data to produce thematic maps and quantify land cover changes is one of the most common applications of remote sensing. Mapping land cover changes at the regional level is essential for a wide range of applications including land use planning, decision making, land cover database generation, and as a source of information for sustainable management of natural resources. Land cover changes in Lake Hawassa Watershed, Southern Ethiopia, were investigated using Landsat MSS image data of 1973, and Landsat TM images of 1985, 1995, and 2011, covering a period of nearly four decades. Each image was partitioned in a GIS environment, and classified using an unsupervised algorithm followed by a supervised classification method. A hybrid approach was employed in order to reduce spectral confusion due to high variability of land cover. Classification of satellite image data was performed integrating field data, aerial photographs, topographical maps, medium resolution satellite image (SPOT 20m), and visual image interpretation. The image data were classified into nine land cover types: water, built-up, cropland, woody vegetation, forest, grassland, swamp, bare land, and scrub. The overall accuracy of the LULC maps ranged from 82.5 to 85.0%. The achieved accuracies were reasonable, and the observed classification errors were attributable to coarse spatial resolution and pixels containing a mixture of cover types. Land cover change statistics were extracted and tabulated using the ERDAS Imagine software. The results indicated an increase in built-up area, cropland, and bare land areas, and a reduction in the six other land cover classes. Predominant land cover is cropland changing from 43.6% in 1973 to 56.4% in 2011. A significant portion of land cover was converted into cropland. Woody vegetation and forest cover which occupied 21.0 and 10.3% in 1973, respectively, diminished to 13.6 and 5.6% in 2011. The change in water body was very peculiar in that the area of Lake Hawassa increased from 91.9km(2) in 1973 to 95.2km(2) in 2011, while that of Lake Cheleleka whose area was 11.3km(2) in 1973 totally vanished in 2011 and transformed into mud-flat and grass dominated swamp. The "change and no change" analysis revealed that more than one third (548.0km(2)) of the total area was exposed to change between 1973 and 2011. This study was useful in identifying the major land cover changes, and the analysis pursued provided a valuable insight into the ongoing changes in the area under investigation.

Remote sensing and GIS for mapping and monitoring land cover and land-use changes in the Northwestern coastal zone of Egypt

Nepal, located in a unique transition zone spanning from plains to mountains and then to the plateau, is characterized by diverse and complex land cover. Based on an object-oriented method and decision tree classifier, a land cover product covering the whole of Nepal in 2010 (hereinafter referred to as the NepalCover-2010) was produced using 30 m-resolution Landsat TM images, consisting of 8 classes at Level I and 31 classes at Level II. The accuracy of the NepalCover-2010 product at Level II was validated using samples collected from high-resolution Google Earth images. The result showed that the overall accuracy of the product was 87.17%, with a Kappa coefficient of 0.85, making it the most accurate product among similar land cover products. The product can accurately reflect the spatial patterns of land cover in Nepal. Forests are the main land cover classes, accounting for 41% of the land, followed by croplands covering about 25%. The areal proportion of paddy fields to dry farmlands was approximately two to three. Topographical and meteorological factors presented as the determining effects on the spatial patterns of land cover in Nepal. With elevation uplift from south to north, land cover classes showed a vertical zonality ordered thus: paddy fields, evergreen broadleaf forests, dry farmlands, evergreen broadleaf shrubs, evergreen needleleaf forests, grasslands, sparse vegetation, and permanent ice/snow. Land cover mapping in Nepal contributes significantly to the basic data collection in this country, and can also be of a benefit to China’s international regional economic cooperation strategy entitled “the Belt and Road Initiative”.

This paper proposed a new land cover mapping algorithm for remote sensing images under the presence of clouds. Here, we modelled the observed image as the weighted sum between the reflectances from the cloud-free image and clouds where weights depend on the degree of cloud contamination called the ‘cloud thickness’ map. Next, we represented the cloud thickness map using a level set function. Similarly, we also represented the land cover map as functions of different level set functions whose values indicate the class label. Next, the maximum a posteriori (MAP) criterion is employed where the most likely land cover and cloud thickness maps are chosen. Under the MAP criterion, the corresponding energy function whose values depend on the level set functions can be derived where the minimum energy point corresponds to the optimum land cover and cloud thickness maps. Since the level set functions are continuous, the optimum solution can be obtained by applying the calculus of variation where the gradient descent algorithm is employed. From the observation model, the cloud-free data can only be reconstructed only if thin to medium clouds are present since thick clouds can completely block the reflectance from the land cover materials. Thus, our algorithm is more suitable for thin to medium than thick cloud covers. Our synthesis and real cloud-contamination examples support our statement since our algorithm achieves significantly higher overall accuracies over other classification techniques, especially, in the thin to medium cloud contamination, and fails to recover the true land cover classes over the thick clouds.

Land cover (LC) information plays an important role in different geoscience applications such as land resources and ecological environment monitoring. Enhancing the automation degree of LC classification and updating at a fine scale by remote sensing has become a key problem, as the capability of remote sensing data acquisition is constantly being improved in terms of spatial and temporal resolution. However, the present methods of generating LC information are relatively inefficient, in terms of manually selecting training samples among multitemporal observations, which is becoming the bottleneck of application-oriented LC mapping. Thus, the objectives of this study are to speed up the efficiency of LC information acquisition and update. This study proposes a rapid LC map updating approach at a geo-object scale for high-spatial-resolution (HSR) remote sensing. The challenge is to develop methodologies for quickly sampling. Hence, the core step of our proposed methodology is an automatic method of collecting samples from historical LC maps through combining change detection and label transfer. A data set with Chinese Gaofen-2 (GF-2) HSR satellite images is utilized to evaluate the effectiveness of our method for multitemporal updating of LC maps. Prior labels in a historical LC map are certified to be effective in a LC updating task, which contributes to improve the effectiveness of the LC map update by automatically generating a number of training samples for supervised classification. The experimental outcomes demonstrate that the proposed method enhances the automation degree of LC map updating and allows for geo-object-based up-to-date LC mapping with high accuracy. The results indicate that the proposed method boosts the ability of automatic update of LC map, and greatly reduces the complexity of visual sample acquisition. Furthermore, the accuracy of LC type and the fineness of polygon boundaries in the updated LC maps effectively reflect the characteristics of geo-object changes on the ground surface, which makes the proposed method suitable for many applications requiring refined LC maps.

Approaches to fractional land cover and continuous field mapping: A comparative assessment over the BOREAS study region

Urban land cover and land use mapping plays an important role in urban planning and management. In this paper, novel multi-scale deep learning models, namely ASPP-Unet and ResASPP-Unet are proposed for urban land cover classification based on very high resolution (VHR) satellite imagery. The proposed ASPP-Unet model consists of a contracting path which extracts the high-level features, and an expansive path, which up-samples the features to create a high-resolution output. The atrous spatial pyramid pooling (ASPP) technique is utilized in the bottom layer in order to incorporate multi-scale deep features into a discriminative feature. The ResASPP-Unet model further improves the architecture by replacing each layer with residual unit. The models were trained and tested based on WorldView-2 (WV2) and WorldView-3 (WV3) imageries over the city of Beijing. Model parameters including layer depth and the number of initial feature maps (IFMs) as well as the input image bands were evaluated in terms of their impact on the model performances. It is shown that the ResASPP-Unet model with 11 layers and 64 IFMs based on 8-band WV2 imagery produced the highest classification accuracy (87.1% for WV2 imagery and 84.0% for WV3 imagery). The ASPP-Unet model with the same parameter setting produced slightly lower accuracy, with overall accuracy of 85.2% for WV2 imagery and 83.2% for WV3 imagery. Overall, the proposed models outperformed the state-of-the-art models, e.g., U-Net, convolutional neural network (CNN) and Support Vector Machine (SVM) model over both WV2 and WV3 images, and yielded robust and efficient urban land cover classification results.

Precise global/regional land cover mapping is of fundamental importance in studies of land surface processes and modelling. Quantitative assessments of the map quality and classification accuracy for existing land cover maps will help to improve accuracy in future land cover mapping. We compare and evaluate four land cover datasets over China. The datasets include the Version 2 global land cover dataset of IGBP, MODIS land cover map 2001, a global land cover map produced by the University of Maryland, and the land cover map produced by the global land cover for the year 2000 (GLC 2000) project coordinated by the Global Vegetation Monitoring Unit of the European Commission Joint Research Centre. The four maps used different classification systems, which made the comparison difficult. So we first aggregated these maps by reclassifying them using a unified legend system. A large-scale, i.e. 1:100 000 land cover map of China was used as the reference data to validate the four maps. The results show that the GLC2000 land cover map represents the highest accuracy. However, it has obvious local labelling errors and a zero labelling accuracy for the wetland type. The MODIS land cover map ranks second for type area consistency and third for sub-fraction overall accuracy compared with reference data, which may be affected by the local labelling error. The IGBP land cover map has good labelling accuracy, although it has a local labelling error and third consistency for type area. The labelling accuracy and type area consistency for the reference data of UMd land cover map is low. We conclude that the accuracies of all the datasets cannot meet the requirements of land surface modelling. For the reference data, i.e. the 1:100 000 land cover map, the classification system needs to be transferred to a well recognized one that has been used commonly in land surface modelling. In addition, we propose an information fusion strategy to produce a more accurate land cover map of China whose classification system should be compatible with the well-accepted classification system used in land surface modelling.

Mapping ecosystem services (ESs) is an important tool for providing the quantitative information necessary for the optimal use and protection of ecosystems and biodiversity. A common mapping approach is to apply established empirical relationships to ecosystem property maps. Often, ecosystem properties that provide services to humanity are strongly related to the land use and land cover, where the spatial allocation of the land cover in the landscape is especially important. Land use and land cover maps are, therefore, essential for ES mapping. However, insight into the uncertainties in land cover maps and how these propagate into ES maps is lacking. To analyze the effects of these uncertainties, we mapped pollination efficiency as an example of an ecosystem function, using two continental-scale land cover maps and two global-scale land cover maps. We compared the outputs with maps based on a detailed national-scale map. The ecosystem properties and functions could be mapped using the GLOBCOVER map with a reasonable to good accuracy. In homogeneous landscapes, an even coarser resolution map would suffice. For mapping ESs that depend on the spatial allocation of land cover in the landscape, a classification of satellite images using fractional land cover or mosaic classes is an asset.

Landscape pattern is a mixture of natural and human-managed patches that vary in size, shape, arrangement and is the result of complex interactions of physical, biological, and social forces. At present, landscape pattern indices (LPIs) have become increasingly popular for characterizing landscape pattern and they are commonly calculated from land cover maps extracted from remotely sensed imagery. Pixel-based hard classification methods assign each pixel to one class and always suffer from the mixed pixel problem, and make the resultant land cover maps and extracted LPIs inaccuracy. Moreover, as the pixel size of hard classification maps is identical with the resolution of the remotely sensed imagery, spatial patterns of patches which are smaller than the resolution of remotely sensed imagery are usually ignored, and hard classification maps thus can hardly represent the spatial pattern of finer scale land cover classes due to the grain size effect. In general, soft classification techniques can estimate the class composition of land cover classes, however, their output fraction images provide no indication of how such classes are spatially distributed. Sub-pixel mapping (SPM) is a technology to transform the resulting fraction images of soft classification technologies into a finer scale hard classification map, which can solve the mixed pixel problem and eliminate the influence of grain size effect to a certain extent while providing the spatial representation of land cover targets at the sub-pixel scale. In this research, 33 landscape-level LPIs (for all land cover classes altogether) and 28 class-level LPIs (for individual land cover class) were applied to assess the performance of SPM method in characterizing landscape pattern. Ikonos imagery of Dujiangyan (Sichuan Province, China) representing urban landscape pattern and Landsat TM imagery of suburb of Wuhan (Hubei Province, China) representing rural landscape pattern were used in this study. We derived LPIs from both SPM maps and hard classification maps and found that SPM can better depict landscape pattern than hard classification methods at both the landscape level and class level in most cases.