Land Classification Research Articles

Analysis of Land Use Based on Carrying Capacity of Agricultural Land in Bantul Regency

Population growth requires expansion of land as a place to grow and develop. If development is not controlled, land conversion can occur for activities that are not in accordance with its function, so that it has an impact on decreasing the carrying capacity of the environment. The environmental carrying capacity of an area is an important factor that must be considered so that the development process carried out can be sustainable. This study aims to determine changes in agricultural land and land carrying capacity in Bantul Regency. The method used is to analyze land change, evaluate land use based on land class and subclass and analyze the carrying capacity of land in Bantul Regency. In formulating these recommendations, several stages were carried out in this research. Geographic analysis is carried out by optimizing Geographic Information System (GIS) applications to map land conditions including an assessment of land use changes over several years.

Optimizing Seagrass Detection in Lampung Selatan: Leveraging Machine Learning Algorithms and Sentinel-2A Imagery

Seagrass beds are one of the coastal ecosystems that play an important role in maintaining the stability of blue carbon. However, high community activities threaten the existence of seagrass beds themselves. South Lampung Regency is one of the areas with considerable seagrass potential but the availability of distribution and density data is still minimal. This research aims to identify and map seagrass density as a first step for seagrass management. The data used in this research is Sentinel 2A multispectral image. Machine learning-based classification methods used are random forest (RF) and support vector machine (SVM) because these algorithms have a good ability to distinguish objects based on their features. This study uses a 2-level classification scheme, where level 1 consists of land, shallow sea, and deep-sea classes. Level 2 is the shallow marine bottom benthic habitat class. The type of seagrass found in this area is Enhalus acoroides. Based on the results of the analysis, low, medium, and high-density classes were obtained with an area of low around 20.12 ha, medium around 34.67 ha and high around 320.12ha with a total area of 374.91ha. RF has a higher overall accuracy of 88.00% while SVM accuracy is 84.00% so it can be concluded that Sentinel 2A images can be used to detect seagrass meadows.

Forest Fire Dynamics in Nepal’s Mid-hills: Insights from Spatial Analysis and Risk Mapping of Doti District

Forest fires have significantly contributed to forest degradation and pollution all over Nepal. Despite their high severity, limited studies have been conducted on the spatial-temporal analysis and risk mapping of these fires especially in the Mid-hill region. This study aims to fill this gap by analyzing Visible Infrared Imaging Radiometer Suite (VIIRS) data from 2012 to 2022, along with satellite imagery, using remote sensing and geographic information systems (GIS) technology. Nine variables, including land class, distance to roads, proximity to settlements, temperature, wind speed, precipitation, slope, aspect, and elevation, were utilized to identify forest fire patterns and create a fire risk map. A multi-parametric hierarchical weighted index model was employed, with multicollinearity checked (VIF<2) to ensure model accuracy and validated by Area Under Curve (AUC=0.787). The study revealed an increasing trend of forest fires, particularly during March to May, in broad-leaved forests near human interference. Out of the total forested area of 1447.58 km2, 42.9%, 46.03%, and 11.07% were classified under high, moderate, and low-risk zones, respectively. The study emphasizes the recurring forest fire problem in often-neglected Mid-hill districts and underscores the need to prioritize them in future strategies.

The availability of forest cover and the critical index of forest area over Merangin-Tembesi Watershed, Jambi Province

The current condition of the Watershed (DAS) has undergone many changes from time to time. The opening of new land areas for agricultural, plantation, mining, residential and industrial activities often does not concern to the sustainability of nature, causing ecosystem changes. The study is aimed to analyze the availability of forest cover over the Merangin-Tembesi watershed and the critical index of the forest area existing. The existing forest cover was analyzed by conducting land cover classification using GIS software and groundtruth. The critical index was performed for the forest area in this watershed. The results showed that critical land in the Merangin Tembesi Watershed was divided into 5 classes, namely uncritical land class with an area of 823,294 ha (60.98%), potentially critical 88,552 ha (6.56%), moderately critical 360,792 ha (26.72%), critical area 52,669 ha (3.90%), and very critical 24,791 ha (1.84%). Protected Forests (HL) in Sarolangun Regency and Merangin Regency are dominated by the moderately critical class 75,260 ha (87.47%) followed by the non-critical class with an area of 5,828 ha (6.77%), critical potential class with an area of 2,189 ha (2.54%), critical class with an area of 1,582 ha (1.84%), very critical class 1,179 ha (1.37%).

Vegetation Health and Forest Canopy Density Monitoring in The Sundarban Region Using Remote Sensing and GIS

The present study explores vegetation health and forest canopy density in the Sundarbans region using Landsat-8 images. This work analyzes changes in vegetation health using two vegetation indices, the Normalized Difference Vegetation Index (NDVI) and Forest Canopy Density (FCD) values of the Sundarbans, from 2014 to 2020. NDVI, comprising two bands, Red and Near-infrared (NIR), shows a declining trend during the period. Two NDVI land cover classification maps for 2014 and 2020 are produced, and the interest area is divided into five classes: Scanty, Low, Medium, and Densely Vegetated Regions and Water Bodies. A single-band linear gradient pseudo-color is used to assess the land cover difference between 2020 and 2014, showing marked changes in densely vegetative areas. The NDVI difference marks the coastal regions with a higher depletion rate of vegetation than the regions away from the seacoasts. FCD has been taken to compare the results of NDVI with it. FCD consists of another four models: AVI (advanced vegetative index), BI (Bare soil index), SSI (scaled shadow index), and TI (thermal index). FCD is also called crown cover or canopy coverage, which refers to the portion of an area in the field covered by the crown of trees. 2014 and 2015 FCD maps are produced with a single band linear gradient pseudocolor with five land cover classifications: bare soil, Bare Soil, Shrubs, Low, Medium, and Highly vegetated regions. Both maps bear a significant resemblance to NDVI land classification maps. Further, the FCD values of the two maps are scaled between 1 and 100, and the area of each class is calculated. To check the veracity of the NDVI and FCD analysis, a Deep Neural Network (DNN) model has been developed to classify each year’s image taken from Google Earth Engine (GEE). It classifies each year’s image with 99% accuracy. The calculation of the area of each class emphasizes the rapid decline of densely wooded vegetation. Almost 80% of the highly forested zone has been diminished and has become part of the medium-forested region. Area inflation in medium-forested regions corroborates the same. The study also analyzes the migration of vegetation density, i.e., where and how many areas are unchanged, growing, or deforested.

Comparative Analysis of Pixel-Level Fusion Algorithms and a New High-Resolution Dataset for SAR and Optical Image Fusion

Synthetic aperture radar (SAR) and optical images often present different geometric structures and texture features for the same ground object. Through the fusion of SAR and optical images, it can effectively integrate their complementary information, thus better meeting the requirements of remote sensing applications, such as target recognition, classification, and change detection, so as to realize the collaborative utilization of multi-modal images. In order to select appropriate methods to achieve high-quality fusion of SAR and optical images, this paper conducts a systematic review of current pixel-level fusion algorithms for SAR and optical image fusion. Subsequently, eleven representative fusion methods, including component substitution methods (CS), multiscale decomposition methods (MSD), and model-based methods, are chosen for a comparative analysis. In the experiment, we produce a high-resolution SAR and optical image fusion dataset (named YYX-OPT-SAR) covering three different types of scenes, including urban, suburban, and mountain. This dataset and a publicly available medium-resolution dataset are used to evaluate these fusion methods based on three different kinds of evaluation criteria: visual evaluation, objective image quality metrics, and classification accuracy. In terms of the evaluation using image quality metrics, the experimental results show that MSD methods can effectively avoid the negative effects of SAR image shadows on the corresponding area of the fusion result compared with CS methods, while model-based methods exhibit relatively poor performance. Among all of the fusion methods involved in the comparison, the non-subsampled contourlet transform method (NSCT) presents the best fusion results. In the evaluation using image classification, most experimental results show that the overall classification accuracy after fusion is better than that before fusion. This indicates that optical-SAR fusion can improve land classification, with the gradient transfer fusion method (GTF) yielding the best classification results among all of these fusion methods.

Comparison of accuracy and reliability of random forest, support vector machine, artificial neural network and maximum likelihood method in land use/cover classification of urban setting

Accurate land use and land cover (LULC) is crucial for sustainable urban planning and for many scientific researches. However, the demand for accurate LULC maps is increasing; it is required to compare the classification algorithms to choose the best one. Though, machine and deep learning algorithms are widely used across the world their application is limited in Bangladesh. Accurate urban LULC mapping is challenging because urban heterogeneity affects image classification models in specific feature extraction. In this research, the accuracy of machine learning algorithms (MLA) of RF (Random Forest), SVM (Support Vector Machine), deep learning algorithm (DLA) of ANN (Artificial Neural Network) and traditional Maximum Likelihood (MaxL) method was compared in LULC classification of Dhaka city. Model accuracy of MLA and DLA was tested by statistical indices of sensitivity, specificity, precision, recall F1 etc. There is a high correlation between SVM and ANN models were found. The overall accuracy of the maps was 0.93, 0.94, 0.91 and 0.95 and kappa was 0.89, 0.91, 0.86 and 0.93 for the MaxL, RF, SVM and ANN models respectively. The user accuracy and producer accuracy largely varied according to LULC classes in the applied models. The lowest accuracy of the models was found for bare land classification followed by built-up and vegetation. The high mixture of LULC classes affects the accuracy of built up and bare land classification which produces the lowest accuracy in the MaxL model. The findings indicate that the most accurate and reliable model for urban LULC classification was the ANN model.

A Spatio-Temporal Encoding Neural Network for Semantic Segmentation of Satellite Image Time Series

Remote sensing image semantic segmentation plays a crucial role in various fields, such as environmental monitoring, urban planning, and agricultural land classification. However, most current research primarily focuses on utilizing the spatial and spectral information of single-temporal remote sensing images, neglecting the valuable temporal information present in historical image sequences. In fact, historical images often contain valuable phenological variations in land features, which exhibit diverse patterns and can significantly benefit from semantic segmentation tasks. This paper introduces a semantic segmentation framework for satellite image time series (SITS) based on dilated convolution and a Transformer encoder. The framework includes spatial encoding and temporal encoding. Spatial encoding, utilizing dilated convolutions exclusively, mitigates the loss of spatial accuracy and the need for up-sampling, while allowing for the extraction of rich multi-scale features through a combination of different dilation rates and dense connections. Temporal encoding leverages a Transformer encoder to extract temporal features for each pixel in the image. To better capture the annual periodic patterns of phenological phenomena in land features, position encoding is calculated based on the image’s acquisition date within the year. To assess the performance of this framework, comparative and ablation experiments were conducted using the PASTIS dataset. The experiments indicate that this framework achieves highly competitive performance with relatively low optimization parameters, resulting in an improvement of 8 percentage points in the mean Intersection over Union (mIoU).

The evolution and driving mechanism of land use classification systems in China

It is of great value to explore the evolution and dynamic mechanism of land use classification systems, such as improving the current classification system and providing guidance for scientific land use methodologies. In this study, the evolution process of land use classification systems in China is analyzed, the characteristics of which in terms of content-form, the urban–rural land use relationship and land functionality are compared, and the evolution mechanism of which is evaluated. The findings reveal that: (1) The land use classification systems in China have undergone a three-stage evolution process, comprising an initial "exploration stage", followed by an "improvement stage", culminating in a "maturity stage"; (2) The content and form of these systems exhibit distinct characteristics, marked by the refinement of construction land, stability in hierarchy, and an increase in the number of classifications. The urban–rural land use relationships have transitioned from a state of "urban–rural separation" to "urban–rural coordination", and ultimately to "urban–rural integration". Moreover, land functions have evolved from single to comprehensive; (3) The evolution of land use systems is primarily driven by national policies, socioeconomic development, and resource endowments, and in essence, it is constrained by the man-land relationship. To meet the needs of global village development, future land classification systems should strive to establish universal international standards.

Vulnerability of Water Resources to Drought Risk in Southeastern Morocco: Case Study of Ziz Basin

Water resources in Morocco have been severely influenced by climate change and prolonged drought, particularly in the pre-Saharan zone. The Ziz watershed faces increasing pressure due to the high demographic growth, increased demand for water, excessive groundwater consumption, and investment in agriculture. But how long will water resources withstand these problems? This study, therefore, enters into the context of the assessment of water resources and estimates their vulnerability using the Standardized Precipitation Index (SPI), Standardized Precipitation Evapotranspiration Index (SPEI), and Standardized Groundwater Index (SGI), on data from the Ziz watershed from 1986 to 2016. Additionally, climate projections were utilized to simulate the future SGI from 2017 to 2100. The Water Evaluation and Planning System (WEAP) was employed to evaluate changes in Land Use and Land Cover (LULC) during the period of 1992–2020, and to generate future scenarios for land class inflows and outflows from 2017 to 2100, in comparison to the reference period of 1986–2016, thereby incorporating the SSP climate scenarios. The results indicate that the Ziz Basin experienced significant drought events in 1986–1989 and 2000–2003. The SPI and SPEI significantly correlated with SGI in some monitoring wells and with specific accumulation periods. The LULC analysis showed an increase in agricultural land and urban land and a decrease in barren or sparse land. Climate data analysis and scenarios predict that under SSP5-8.5, minimum and maximum temperatures will increase by 2.61 °C and 2.93 °C, respectively, and precipitation will decrease by 30% over this century. This substantial shift in climate conditions is reflected in the decline in SGIs, especially in the long term under SSP5-8.5. Water availability will decrease during this century under SSP3-7.0 and SSP5-8.5, as reflected in reduced land class inflows and increased outflows. These findings emphasize the need for stakeholders to implement integrated water governance for sustainability in the Ziz watershed.

Engineering Applications of Urban Green Space Planning in Mountainous Areas: An Improved Structure-based RS Land Class Information Extraction Method for U-Net Networks

Engineering Applications of Urban Green Space Planning in Mountainous Areas: An Improved Structure-based RS Land Class Information Extraction Method for U-Net Networks

Land suitability assessment for rapeseed potential cultivation in upper Tigris basin of Turkiye comparing fuzzy and boolean logic

Assessment of land suitability is a prerequisite for the conservation and maintenance of land productivity and the improvement of land use and management systems. This study assessed land suitability for rapeseed (Brassica napus L.) production using topography, climate, and soil data by analytical hierarchy process (AHP) and the Mamdani Fuzzy Inference System (MFIS). The study area covers 3737 km2 of land in the Diyarbakir province of southeastern Turkiye. The weights of topography, soil and climate factors in AHP were determined by expert opinions and the information in related literature. They were included in the whole process, mainly membership functions and rule base stages in the MFIS. The highest weighted factor was slope (0.264), followed by altitude (0.121), annual average temperature (0.114) and soil texture (0.112). The MFIS-based land suitability assessment indicated that the proportions of moderately (S2), marginally (S3) and currently not suitable (N1) land classes in the study area were 71.35%, 18.75% and 9.9%, respectively. The AHP results showed that 98.94% of the land was S3, and 1.06% was N1. The compatibility of AHP and MFIS methods in N1 land units was 96.05%, while the agreement for S2 and S3 land classes was not sufficiently high. The suitability of rapeseed cultivation has been more sensitively assessed by the fuzzy continuous classification obtained by the MFIS method.

Mapping surface water dynamics (1985–2021) in the Hudson Bay Lowlands, Canada using sub-pixel Landsat analysis

The fate of the carbon stored in frozen peatlands in Canada's Hudson Bay Lowlands (HBL) depends strongly on water, with wetter conditions favoring long-term storage. Dynamic surface water products generated from historical Landsat data exist to inform surface water trends in the HBL based on binary classifications of land vs water at 30 m spatial resolution. However, the HBL contains many water features smaller than 30 m, including streams and patterned fens that require a sub-pixel mapping approach to fully capture their dynamics. In this paper, we leverage an existing binary dynamic surface water product to generate a spatially and temporally comprehensive Landsat sub-pixel surface water time-series over the HBL. We first generate Landsat Red, NIR and SWIR composites from 1985 to 2021 and evaluate data-driven machine learning and physical models to derive sub-30 m water fractions across the entire mapping domain. Data-driven models were calibrated on 30 m water fractions obtained from 16 of 17 WorldView scene water fraction maps and assessed on the 17th held-out scene in a jackknife validation. Data-driven models were also calibrated by relating scaled Landsat reflectance to water fractions obtained from binary water masks at corresponding scales from 90 to 270 m. Physical models involved simple linear unmixing using a new approach that samples land endmembers in the vicinity of each pixel to represent land reflectance, and a global water endmember sampled beneath permanent water in each image. The assumption of linear scaling was verified at sub-30 m to 270 m resolution for NIR, SWIR and NDVI, which were all highly correlated with water fraction and had no significant differences among water fraction – spectral feature relationships across the full range of scales. Each of 11 different methods and input spectral features were evaluated against WorldView water fractions and ranked based on nine criteria. We found that machine learning using all spectral features including local NIR land endmembers performed best among data-driven approaches, and linear unmixing of the NIR band using local land endmembers performed best among physical. Both best-performing methods were applied to the entire time-series and surface water area trends were calculated and compared with trends obtained from binary water masks. All three methods agree on the direction of surface water change towards wetting, with sub-pixel inter-annual variation agreeing with stream discharge noted in other studies.

Machine learning for IoT-based smart farming

Agriculture balances food requirements for mankind, and the supply of essential raw materials for many industries is the fundamental occupation in India. Smart farming allows analyzing the growth of crops and the parameters which influence crop growth and supports farmers in their activities, it is more profitable and reduces irrigation wastages. The proposed model is a smart farming system that analyzes the influence of parameters on crop growth and predicts the soil condition using a machine learning algorithm. Temperature, Ph, humidity, gas, and water level are the few most essential parameters to determine the quantity of water required and to find hazardous gas in any agriculture field. This system comprises temperature, pH, humidity, smoke detector, and water level sensor, deployed in an agricultural field, sends data through a microprocessor, developing an IoT device with cloud. In this study, we present a model that predicts soil series with regard to land type and, in accordance with the prediction, suggests appropriate crops. For soil land classification and crop prediction application is developed using KNN algorithms. Three steps are necessary for its implementation: the first is data collecting using sensors placed in an agricultural field, the second is data cleaning and storage, and the third is predictive processing utilizing the ML technique. The results obtained through the algorithms are sent to the cloud, which helps in decision-making in advance.

Appraisal of irrigation land suitability of semi-arid climatic zone of Rajasthan, India

Abstract Planning how to use the land resources that are available for irrigation is crucial for reducing the problem of food security. Analytical hierarchy process methodologies were used to undertake an analysis of the appropriateness of land using the GIS approach. Each parameter's weight was assessed using a 6 × 6 pairwise comparison matrix because six parameters, including slope, distance to the nearest road, land use land cover, population density, soil texture and rainfall deficit, are taken into account. After all, the weighted overlay approach of the GIS application was used to create the ultimate land suitability map for irrigation. Using the natural break classification approach, the final map was divided into land suitability classes. Of the entire region, 12.9% fell into the highly suitable (S1) class, 85.14% fell into the moderately suitable (S2) class, and 2.57% fell into the least suitable (S3) class. The finding of the study will play a significant role in performing irrigation taking into account the input parameters and ultimately defining suitable land classes and lands that may be used in accordance with current irrigation technology.

Environmental land use conflicts and ecosystem services: a paper review

The topic of potential conflicts related to land use involving human activities in a watershed is an important matter to be discussed. The background of this research is that conditions in mountainous watersheds, and agricultural land use cause changes in ecosystem services, with trade-offs between crop production and erosion regulation. Watershed management with an environmental concept often faces problems with different interests among stakeholders. Although several studies have initiated the mapping of land-use conflicts between human activities and conservation, the spatial assessment of land-use conflicts on environmental issues and trade-offs of ecosystem services in agricultural areas has not been fully considered. The purpose of this study is to map land use, with indicators of measuring conflicts in the value of conservation and agricultural development, through scenarios of the level of erosion hazard on agricultural land. This study provides input to decision-makers regarding watershed conservation efforts that still consider aspects of the economic needs of agricultural land. The objective of the reviewers is to understand how the concept of mapping the potential conflict of land use in the Haean watershed in South Korea can be applied in Indonesia. The methodology used is to apply the agricultural land suitability index based on various analytical criteria to estimate the spatial preferences of agricultural activities. To predict erosion, using the Revised Universal Soil Loss Equation (RUSLE) method and the classification of agricultural land in the watershed is divided into four levels of land use conflict (lowest, low, high, and highest).

Mapping the spatial heterogeneity of global land use and land cover from 2020 to 2100 at a 1 km resolution

A fine global future land use/land cover (LULC) is critical for demonstrating the geographic heterogeneity of earth system dynamics and human-earth interaction. In this study, we produced a 1 km global future LULC dataset that takes into account future climate and socio-economic changes as well as the impact of simulated results of the former year on temporally adjacent periods. By incorporating the variations in climatic and socio-economic factors, we differentiated LULC suitability probabilities for historical and future periods across representative SSP-RCP scenarios. Then, by using an improved cellular automata model-PLUS to simulate the patch-level changes of various land classes, we iteratively downscaled water-basin-level LULC demands in various future scenarios to a spatial resolution of 1 km. Our dataset achieves a high degree of simulation accuracy (Kappa = 0.94, OA = 0.97, FoM = 0.10) and precisely captures the spatial-temporal heterogeneity of global LULC changes under the combined effects of climate change and socio-economic development. This robust and fine-scale LULC dataset provides valuable spatially-explicit information essential for earth system modeling and intricate dynamics between anthropogenic activities and the environment.

A dual spatial analysis method based on recreation opportunity spectrum and analytical hierarchy process for outdoor recreation site suitability

Spatial planning tools are required to conduct outdoor recreational activities in harmony with the natural environment. This study introduces a dual spatial analysis method in which the Recreation Opportunity Spectrum (ROS) and Analytic Hierarchy Process (AHP) are run sequentially to evaluate different criteria groups. This study was conducted in the coastal region of the Enez district of Edirne, one of Turkey's northwestern provinces. With ROS, remoteness, and evidence of humans criteria were considered unweighted, and four land classes, urban, rural, natural with road, and primitive-semiprimitive, and activities suitable for these land classes were determined in the GIS environment. These are six terrestrial activities including nature walk, cycling, horse riding, picnic, camping, and wildlife observation, and four water-based activities including canoeing, swimming, diving, and handline fishing. With the AHP, nine natural criteria were analyzed as weighted and four suitability levels were determined for each activity in the GIS environment, as not suitable, marginally suitable, moderately suitable, and highly suitable. Sensitivity analysis by changing the priorities of the criteria showed that the analysis using the AHP results was reliable especially for determining highly suitable areas, and could be helpful for outdoor recreation planning. Finally, the suitability maps matched the ROS land classes, and the verified final maps were created. This study provides reliable results, showing that planners and decision-makers can adapt to similar areas with different administrative or geographical boundaries. Simultaneously, it contains a series of information that will contribute to planning components such as exploring, developing, diversifying, and presenting outdoor recreation opportunities and restricting them when necessary.

Assessment of remotely sensed inventories for land cover classification of public grasslands in Manitoba, Canada

AbstractLand cover classification is one of the most common applications of remote sensing and is used for developing and modifying land management policies on agricultural landscapes to achieve conservation and economic goals, such as reducing grassland degradation and improving livestock and crop production. In this study, the grassland classification of the crown lands (public grasslands in Canada) from a newly developed remotely sensed dataset in the Prairie Province of Manitoba (i.e., the Manitoba Grassland Inventory, MGI) was assessed in terms of accuracy by comparison to non‐spatial government records. The analysis consisted of (i) converting non‐spatial records from the provincial crown land database to spatially‐defined parcels by performing parcel delineations using geographic information system (GIS) and R programming tools, (ii) summarising the MGI classification at the same spatial scale, and (iii) comparing the agreement between MGI and the crown land database. The most common land cover types identified were: forest (30%) and shrubland (25%), followed by native (10%) and tame (9%) grasslands. However, the class agreements between woody (i.e., forests and shrublands) and grassy (i.e., native and tame grasslands) vegetation classes were low between these datasets because of their spectral similarities. Based on these results, we suggest additional refinements on both sensor and ground data to improve the classification agreement between these datasets. This study is one of the first attempts to compare ground‐collected government records against a remotely sensed product in Manitoba.

Salt Land classification based on pansharpening images using Random Forest algorithm

ABSTRACT The analysis and policymaking of national salt production require quality Salt Land maps with more detailed object information. This study successfully classified Salt Land into Active and Inactive based on pansharpening images using random forest algorithm. The results show: 1) the best pansharpening method is UNB PanSharp; 2) the texture variables have more significant contributions to enhance classification accuracy; 3) scheme 5 is the most efficient and recommended scheme in research, achieves an OA of 66.699% and a Kappa coefficient of 39.4%. This study can provide a good reference for developing Salt Land mapping and policies for practitioners or stakeholders.