Accuracy Kappa Research Articles

Land Use and Land Cover Change in East Java Indonesia from 1972 to 2021: Learning from Landsat

This study analyses land use and land cover (LULC) changes during the last five decades (1970–2021) in East Java Province, Indonesia. The changes are analysed by comparing four maps interpreted from Landsat images (MSS 1972, TM 1997, OLI 2013, and OLI 2021). The main research procedures consist of (1) data collection, (2) field survey, (3) image classification, and (4) LULC change interpretation. The classification uses the maximum likelihood algorithm, achieving an overall kappa accuracy of > 75%. The classification produces eight classes, i.e., built-up land (BU), heterogeneous agricultural land (HAL), bare soil (BS), paddy fields (PF), open water (OW), vegetation (VG), shrubland (SH), and wetlands (WL). The analysis shows a significant shift in land use from 1972 to 2021, with HAL declining by 46%, SH by 91%, and BS by 88%. In contrast, PF has increased by 105%, VG (forest and plantation) by 64%, and built-up areas by a remarkable 366%. These changes show a significant shift from dryland agriculture, shrublands, and barren lands to irrigated regions, vegetated areas, and urban growth. Furthermore, there are increases in water bodies due to the construction of several reservoirs to support water availability. In coastal regions, the development of inland aquaculture leads to a proliferation of wetlands (salt evaporation ponds, fish ponds, and rice-fish integrated farming).

Assessing Land-Cover Change Trends, Patterns, and Transitions in Coalfield Counties of Eastern Kentucky, USA

Surface coal mining and reclamation have greatly reshaped eastern Kentucky’s landscape affecting its socioeconomic, environmental and climatic aspects. This study examined the land-cover changes, trends and patterns in Floyd, Knott, Letcher, Magoffin, Martin, Perry, and Pike counties from 2004 to 2019. Using a random forest classifier, land cover was categorized into seven major classes, i.e., water, barren land, developed land, forest, shrubland, herbaceous, and planted/cultivated, majorly based on Landsat images. The Kappa accuracy ranged from 75 to 89%. The results showed a notable increase in forest area from 5052 sq km to 5305 sq km accompanied by a substantial decrease in barren land from 179 sq km to 91 sq km from 2004 to 2019. These findings demonstrated that reclamation activities positively impacted the forest expansion and reduced the barren land of the study area. Key land-cover transitions included barren land to shrubland/herbaceous, forest to shrubland, and shrubland to forest, indicating vegetation growth from 2004 to 2019. An autocorrelation analysis indicated similar land-cover types clustered together, showing effective forest restoration efforts. As surface coal mining and reclamation significantly influenced the landscapes of the coalfield counties in eastern Kentucky, this study provides a holistic perspective for understanding the repercussions of these transformations, including their effects on humans, society, and environmental health.

Assessing Urbanization Dynamics Through Land Use/Land Cover Change Detection in Zunheboto Sadar, Nagaland, India

Providing a systematic overview of the environment is crucial for identifying environmental changes and understanding human-related factors and feedback in research on land use and land cover (LULC) changes and the incorporation of LULC data facilitates pinpointing particular areas undergoing changes, discerning the character of these changes, and gaining insights into the evolving nature of the land. Utilizing Remote Sensing & Geographic Information System (GIS) and methodologies, this study aims to examine variation LULC patterns between 2014 and 2023 in Zunheboto Sadar, the highest populated region in the Zunheboto district of Nagaland with area covering about 123 Km2 and altitude encompassing from 642 m to 1989 m above mean sea level. Landsat-8 OLI data was employed to produce a map depicting LULC in the study region. The application of Supervised classification method, specifically the Maximum Likelihood Classifier (MLC), were utilised in producing LULC maps. Through this comprehensive analysis, the research aims to identify specific areas undergoing alterations, discern the nature and extent of these changes, and explore the underlying causes and human responses. Eight classes were delineated namely Jhum cultivation, Moderately Dense Forest, Open Forest, Settlements, Shrub, Terrace Cultivation, Very Dense Forest, and Water Bodies. In the year 2023, Jhum cultivation has decreased by 68.36%, Very Dense Forest by 16.96%, Water Bodies by 6.53%, Shrub by 52.36%, and Terrace Cultivation by 32.83%. Conversely, there is an increase in Moderately Dense Forest by 21.39%, Open Forest by 111.25%, and Settlements by 38.41%. The Kappa accuracy assessment technique was chosen for evaluating the accuracy of the categorized map. The categorized map of 2023 exhibits an Overall Classification Accuracy of 86.50%, coupled with a Kappa Coefficient of 0.84, indicating an almost perfect classification. Similarly, for 2014, the Overall Classification Accuracy stands at 78.50%, with a Kappa Coefficient of 0.75, also falling into the Substantial classification category.

Identification of mangrove forests using Sentinel-1 Synthetic Aperture Radar (SAR) satellite imagery in Medan Belawan District, Medan City

Mangroves are a forest ecosystem tolerant to salt (halophytic) and influenced by sea tides. Mangrove forests are producers of ecosystem services that all living things need. The use of remote sensing can be used to identify mangrove forests in large areas in a short time. Sentinel-1 is a satellite product widely used to monitor land cover, including mangrove forests. The research objective was to identify mangrove forests in the Medan Belawan District using Sentinel-1 satellite imagery. Land cover classification in digital images was done using supervised classification with the maximum likelihood and random forest method. The results showed that random forest methods are more accurate than the maximum likelihood of identifying mangrove forests in Medan Belawan. The kappa accuracy value of the digital satellite image classification using the random forest method is 85.09%, and the maximum likelihood method is 77.4%. Based on the random forest method, the mangrove forest area is 1,325.81 hectares or 43.64% of the Belawan District area.

Land cover change analysis in Majalengka Regency using the pan-sharpening method and random forest machine learning algorithm

The ever-increasing population has accelerated the need for housing and supporting facilities. Further, this growing number of residences and life support facilities resulted in changes in land cover. For instance, the construction of the West Java International Airport and the increase in the population resulted in land cover changes in Majalengka Regency, Indonesia. This study aims to analyze changes in land cover during 2014, 2018, and 2022 in the Majalengka district using Landsat 8 satellite imagery. We used the pan-sharpening research method, while for the closure classification, we used machine-learning random forest algorithms on the google earth engine platform. The land cover classification classes adopted in this study were natural vegetation, cultivated vegetation, open land, built-up land, and bodies of water. The obtained land cover classification results suggested overall accuracy values of 0.96, 0.94, and 0.93 in 2014, 2018, and 2022 respectively, with kappa index values of 0.950, 0.925, and 0.91667. The results indicate a trend of changes in the land cover in the Majalengka district. From 2014-2022, the trend of increasing land cover area was observed in open land and built-up land, while the decreasing land cover area was found in natural vegetation and cultivated vegetation areas. Using both the pan-sharpening method and the machine learning random forest algorithm, we established images with a more detailed appearance with an outstanding kappa accuracy value (above 0.85). Therefore, the developed algorithm can be used in land cover change mapping analysis.

Examining the Use of the Watershed Algorithm for Segmenting Crown Closure on a Dry Land Forest

This paper uses a watershed algorithm to detect canopy cover in dryland forests. The study at to determine the best parameters of the watershed segmentation algorithm to obtain information on crown closure from filtered and unfiltered high and very high-resolution images. The best performance of each parameter combination of tolerance value (T), mean value (M), and variance value (V), which is written as C:[T]-[M]-[V], is determined based on the level of accuracy. This study uses Pleiades-1B and SPOT-6 images as primary digital data. The results showed that the low-pass filtered Pleiades-1B image showed the best performance with a combination of parameters C6-MF:[10]-[0.7]-[0.3], had an overall accuracy (OA) of 91.0% and an accuracy Kappa (KA) by 83.2%. While the low-pass filtered Spot-6 image shows the combination of parameters C7-MF:[10]-[0.8]-[0.2], which has an accuracy of 90.6% OA and 65.4% KA. This study concludes that the filtered image with a low-pass filter always gives more accurate results than the original data (without filter), both for Pleiades-1B and SPOT-6 images. The very high spatial resolution provides better accuracy than the high spatial resolution

Improving the accuracy of mangrove forest mapping using sentinel image and developed classification algorithm based on machine learning

ABSTRACT The national mangrove rehabilitation required mangrove spatial information with very high accuracy. This study successfully developed the classification algorithm measurable for mapping mangrove forest change based on machine learning. We also modified the recursive feature elimination (RFE) method to obtain the most optimal feature importance (FI). The results showed that the developed classification algorithm was built based on the five most important features including digital elevation model (DEM), near-infrared (NIR), normalized difference moisture index (NDMI), normalized difference water index (NDWI), and distances from brackish water river (DBWR). The developed algorithm increased overall and kappa accuracy by 0.48% and 0.01, respectively.

Application of Sentinel-2A Images for Land Cover Classification Using NDVI in Jember Regency

The advancement of remote sensing technology has led to the development of sophisticated image processing methods that yield highly accurate land cover classification information, minimizing misinterpretations. The Normalized Difference Vegetation Index (NDVI) is a widely utilized method in remote sensing for measuring green vegetation. A significant portion of the Jember Regency area is covered by vegetation. This study aimed to identify various land cover types in the Jember Regency area, quantify the area for each classification, and establish the NDVI value ranges for each type of cover. Sentinel-2 was employed as the primary data source, and the NDVI method was utilized for land cover classification in the Jember Regency. The region exhibited diverse land cover types. Data from Sentinel-2A captured in June and October 2019 were chosen due to their accessibility, open-source nature, and adequate spectral, spatial, and temporal resolution. The classification in this study encompassed five classes: water bodies, settlements, dry fields, irrigated paddy fields, and forests. Error analysis was conducted using a confusion matrix with the Overall and Kappa algorithms. The accuracy results for June indicated a Kappa Accuracy of 37.7% and Overall Accuracy of 54.5%. In October, the Kappa Accuracy increased to 39.9%, and the Overall Accuracy reached 56.5%. In conclusion, the NDVI method did not meet the criteria for accurately interpreting land cover classification.

LinkNet-Spectral-Spatial-Temporal Transformer Based on Few-Shot Learning for Mangrove Loss Detection with Small Dataset

Mangroves grow in intertidal zones in tropical and subtropical regions, offering numerous advantages to humans and ecosystems. Mangrove monitoring is one of the important tasks to understand the current status of mangrove forests regarding their loss issues, including deforestation and degradation. Currently, satellite imagery is widely employed to monitor mangrove ecosystems. Sentinel-2 is an optical satellite imagery whose data are available for free, and which provides satellite imagery at a 5-day temporal resolution. Analyzing satellite images before and after loss can enhance our ability to detect mangrove loss. This paper introduces a LSST-Former model that considers the situation before and after mangrove loss to categorize non-mangrove areas, intact mangroves, and mangrove loss categories using Sentinel-2 images for a limited number of labels. The LSST-Former model was developed by integrating a fully convolutional network (FCN) and a transformer base with few-shot learning algorithms to extract information from spectral-spatial-temporal Sentinel-2 images. The attention mechanism in the transformer algorithm may effectively mitigate the issue of limited labeled samples and enhance the accuracy of learning correlations between samples, resulting in more successful classification. The experimental findings demonstrate that the LSST-Former model achieves an overall accuracy of 99.59% and an Intersection-over-Union (IoU) score of 98.84% for detecting mangrove loss, and the validation of universal applicability achieves an overall accuracy of more than 92% and a kappa accuracy of more than 89%. LSST-Former demonstrates superior performance compared to state-of-the-art deep-learning models such as random forest, Support Vector Machine, U-Net, LinkNet, Vision Transformer, SpectralFormer, MDPrePost-Net, and SST-Former, as evidenced by the experimental results and accuracy metrics.

Spatial Analysis of the Growth of the Bengawan Solo Delta, Gresik Regency, East Java Province, Indonesia 1995-2023

The objective of this research is to conduct a spatial analysis of the development of the Bengawan Solo River Delta, Gresik Regency, East Java Province from 1995 - 2023. This research method uses both re-mote sensing technology and Geographic Information Systems (GIS), such as supervised classification and overlay techniques, kappa accuracy, and direct field verification (ground check). The results of this research indicate that deltaic growth has occurred in the Bengawan Solo delta region. This is demonstrated by the increasing amount of land advancing towards the sea, also known as oloran land. Analysis was done from 1995 - 2000, 2000 - 2005, 2005 - 2010, 2010 - 2015, and 2015 - 2023 to find the development rate of this difference. Quite high growth occurred in the range of 1995 - 2000 which was an area of 472.33 ha. and in 2010 - 2015 which was an area of 423.48 ha. While the range of 2000 - 2005 is an area of 253.82 ha. and 2015 - 2020 which is an area of 259.94 ha experienced growth of no more than the growth in 1995 - 2000 and in 2010 - 2015. This area experienced land erosion from 2005 - 2010, with eroded soil spanning an area of -278.38 ha. Then it will grow once more from 2010 - 2023, totaling 517.63 hectares. The new land utilization of Delta Bengawan Solo is mostly grown from mangroves naturally and planting results (44%), some are used by the surrounding community for the expansion of land (33%), and some are still empty land (19%) as well as for other provision (1%).

Minimum Noise Fraction and Long Short-Term Memory Model for Hyperspectral Imaging

In recent years, deep learning techniques have presented a major role in hyperspectral image (HSI) classification. Most commonly Convolutional Neural Network (CNN) and Long Short-Term Memory (LSTM) has greatly advanced the accuracy of hyperspectral image classification, making it powerful tool for remote sensing applications. Deep structure learning, which involves multiple layers of neural network, has shown promising results in effectively addressing nonlinear problems and improving classification accuracy and reduce execution time. The exact categorization of ground topographies from hyperspectral data is a crucial and current research topic that has gotten a lot of attention. This research work focuses on hyperspectral image categorization utilizing several machine learning approaches such as support vector machine (SVM), K-Nearest Neighbour (KNN), CNN and LSTM. To reduce the number of superfluous and noisy bands in the dataset, Principal Component Analysis (PCA) and Minimum Noise Fraction (MNF) were utilized. Different performance evaluation measures like time taken for testing, classification accuracy, kappa accuracy, precision, recall, specificity, F1_score, and Gmean have been taken to prove the efficacy of the models. Based on the simulation results, it is observed that the LSTM model outperforms the other models in terms of accuracy percentage and time consumption, making it the most effective model for classifying hyperspectral imaging datasets.

Using UAV multispectral photography to discriminate plant species in a seep wetland of the Fynbos Biome

Wetlands harbour a wide range of vital ecosystems. Hence, mapping wetlands is essential to conserving the ecosystems that depend on them. However, the physical nature of wetlands makes fieldwork difficult and potentially erroneous. This study used multispectral UAV aerial photography to map ten wetland plant species in the Fynbos Biome in the Steenbras Nature Reserve. We developed a methodology that used K-Nearest Neighbour (KNN), Support Vector Machine (SVM), and Random Forest (RF) machine learning algorithms to classify ten wetland plant species using the preselected bands and spectral indices. The study identified Normalized green red difference index (NGRDI), Red Green (RG) index, Green, Log Red Edge (LogRE), Normalized Difference Red-Edge (NDRE), Chlorophyll Index Red-Edge (CIRE), Green Ratio Vegetation Index (GRVI), Normalized Difference Water Index (NDWI), Green Normalized Difference Vegetation Index (GNDVI) and Red as pertinent bands and indices for classifying wetland plant species in the Proteaceae, Iridaceae, Restionaceae, Ericaceae, Asteraceae and Cyperaceae families. The classification had an overall accuracy of 87.4% and kappa accuracy of 0.85. Thus, the findings are pertinent to understanding the spectral characteristics of these endemic species. The study demonstrates the potential for UAV-based remote sensing of these endemic species.

Application of NDVI Sentinel-2Ac Imagery for Agricultural Land Development in Food Buffer Area of New Capital City Indonesia

Paser Regency is one of the food buffer areas for the New Capital City (IKN) of Indonesia in East Kalimantan Province. The relocation of the capital to IKN increases the number of populations and food demand as well. To meet food needs in IKN in the future, the development of agricultural land is needed. This study aimed to identify the type and direction of land use in Paser Regency to ensure food self-sufficiency. Sentinel-2 satellite Imagery with the Maximum Likelihood Classification (MLC) method was used in this study. The sentinel data from 2018 to 2022 was analyzed to identify the existing land use types over the past five years. The land use direction was determined based on the existing land use type, soil type, and land slope class. The results with an accuracy value for the overall test of 91.68% and kappa accuracy of 0.86 showed a change in land use from 2018 to 2022. Eight types of main land used were identified such as non-swamp rice fields, swamp rice fields, oil palm plantations, moors, shrubs, forests, settlements, and water bodies. The application of NDVI Sentinel-2Ac Imagery can identify a potential area of 13,249.7 hectares for rice field development and can significantly increase rice production to ensure food self-sufficiency in IKN.

Comprehensive Land Use and Land Cover Classification in Chhatrapati Sambhajinagar City, Maharashtra, India, Using Satellite Imagery and Machine Learning Techniques

Abstract: In This research presents an extensive analysis of land use and land cover (LULC) classification in Chhatrapati Sambhajinagar City, a rapidly expanding urban centerin Maharashtra, India, highly susceptible to climate change. Using remote sensing and GIS techniques, the research evaluates the impact of LULC on ecosystem service values (ESV) over a nine- year period (2013-2021). Satellite-based data were utilized to quantify LULC changes, encompassing Green Area, Buildup Area, Road Area, WaterBodies, and Barren Land. The research employs five distinct classification methods and investigates the effectiveness of pre-processing techniques in improving classification accuracy during selected years (2013, 2016, 2019, and 2021). The performance of Mahalanobis Distance, Maximum Likelihood, Parallelepiped, and Support Vector Machine methods is assessed under both preprocessed and non-preprocessed conditions, using overall accuracy and Kappa accuracy as evaluation metrics. The results reveal consistent enhancements in overall accuracy, with Mahalanobis Distance, Maximum Likelihood, and Support Vector Machine methods demonstrating percentage improvements ranging from 0.43% to 5.08% after preprocessing. Particularly noteworthy are the substantial percentage increases in Parallelepiped Classification, ranging from 23.74% to 53.34%, highlighting the transformative impact of preprocessing on its accuracy. These findings emphasize the necessity for customized preprocessing strategies to refine the precision of LULC classification models, offering valuable contributions to the realms of remote sensing and geospatial analysis. Moreover, the insights gleaned from this research provide crucial guidance for urban growth management and climate resilience strategies in rapidly developing urban areas, making it a valuable resource for sustainable urban planning

Land Use and Land Cover (LULC) Change in Eastern Areas of East Java From 1972 To 2021: Learning From Landsat Image

Urban development, population growth, high traffic jams, and intensive disaster events are indicators of changing the landscape of the eastern area of East Java. Investigating these changes is vital for planning and environmental protection in the future. This study examines changes in land use and land cover (LULC) during the past 50 years in the eastern part of East Java from 1972 to 2021. The changes are examined by contrasting four maps derived from Landsat images (1972, 1997, 2013, and 2021). The following are the main study procedures: (1) data inventory, (2) field survey, (3) image processing and classification, and (4) interpretation of LULC changes. With Google Earth Engine, all photos are downloaded (GEE). Landsat image classification was completed using the maximum likelihood algorithm with an overall Kappa accuracy of>85%. Eight (8) major classifications are therefore produced by the classification: (1) the pavement or urban area (PUA); (2) heterogeneous agricultural land (HAL); (3) bare soil (BS); (4) paddy field (PF); (5) open water body (OWB); (6) vegetation/plantation (VG); (7) shrubland (SL); and (8) wetlands (WL). In the areas with rapid development, the LULC change is more pronounced, i.e., Pasuruan, Jember, and Banyuwangi Regencies. LULC change in and near mid-regency and the rural regions comes next. Over the past fifty years, regional growth has resulted in increases in PUA (+4.4 percent), PF (+12.2 percent), and VG (+17.9 percent). On the other hand, the development has decreased SL, BS, and HAL by 5.8 and 15.9 percent, respectively (-13.1 percent ). LULC alterations for human activities have profoundly altered the natural landscape. Keywords: East Java, GEE, Image Processing, Landsat, LULC

Monitoring the Impact of Land Cover Change on Urban Heat Island with Remote Sensing & GIS

The increasing need for land has resulted in a higher rate of land conversion and urbanization, leading to a rise in urban density and the occurrence of an Urban Heat Island (UHI) effect. The application of remote sensing and GIS can serve as a substitute for data collection in monitoring the UHI phenomena. This work utilizes Landsat 8 OLI satellite image data, namely band 10, to analyze Land Surface Temperature (LST). Bands 5 and 4 are employed to assess the distribution of Normalized Difference Vegetation Index (NDVI) in Bekasi Regency during the years 2014 and 2020. The relationship between NDVI and LST is highly correlated as they can effectively forecast the influence of areas with sparse vegetation on temperature. The guided classification approach, employing the maximum likelihood algorithm and kappa validation, is utilized to evaluate alterations in land use. The kappa accuracy test yielded a score of 0.90% for 2014 and 0.99% for 2020. The research conducted between 2014 and 2020 revealed changes in land distribution. Specifically, the built-up land area increased by 99.92 Km2, empty land expanded by 280.82 Km2, bodies of water covered an additional 46.13 Km2, and vegetation expanded by 293.91 Km^2. According to the UHI research, it is evident that there has been a rise in surface temperature in Bekasi Regency from 2014 to 2020. In 2014, the minimum temperature reached 30 °C, and the maximum temperature reached 51 °C. In 2020, the minimum temperature was recorded at 34 °C, while the maximum temperature reached 52 °C.

Data-driven framework for analyze the visual quality of streets: A systematic measurement of street quality through street view images utilizing deep learning algorithms

The study developed a framework that utilizes deep learning algorithms to analyse the visual quality of streets (VQOS) through street view images (SVIs), while overcoming the constraints observed in existing practices. The study consisted of four main stages, including literature reviews and expert discussions, development of deep learning algorithms, testing of developed algorithms, and validation. The study developed both convolutional neural network (CNN) and feed forward neural network (FFNN) algorithms, using 2684 street view images extracted from Google Street View images and Cityscape datasets, rated by an expert panel and the general public. The proposed framework comprises stages such as the collection and rating of street view images, image processing, developing and training the CNN, and testing and validating stages. The proposed framework achieved 90.51% internal validation accuracy using the accuracy metric in Keras and 86.7% external validation accuracy, with an accepted level of kappa accuracy of 80%. Urban planners, designers, architects, and landscape architects can use this framework as a tool for quantitatively measuring and mapping the visual quality of streets and assessing the impact of proposed developments and guidelines on the visual quality of streets. This proposed framework will enhance the effectiveness of their new proposals and designs.

Detecting land use land cover using supervised maximum likelihood algorithm on spatiotemporal imagery in Samarinda, Indonesia

Land use change commonly has an impact on urban areas. The land use and land cover (LULC) model should be used to address land problems. Land conversion is inextricably linked to Samarinda, Indonesia, the administrative center of the province of East Kalimantan. Additionally, this city serves as the new capital city’s Ibu Kota Nusantara-specific buffer zone. The current study aims to find and assess Samarinda City’s LULC. A supervised maximum likelihood technique was used to extract this data from spatiotemporal images. Both spatial and non-spatial data about changes in LULC are used in this study. The imagery data for the Samarinda region includes Landsat 5, 7, and 9 images from 1994 to 2022, along with administrative maps and GPS measurement data enabling on-the-ground checkpoints. Before classification analysis, radiometric and atmospheric correction, cropping, and layer stacking procedures were used to treat the image data. In supervised classification, the maximum likelihood method is applied to the four land classes of uncultivated vegetation, cultivated vegetation, waterbodies, and land with buildings. The research result indicates that the number of uncultivated land classes decreased by 6.38 percent and the percentage of cultivated land decreased by 5.52 percent, according to the study’s findings. On the other hand, the proportion of water bodies increased by 2.39 percent, and the proportion of built-up land classes increased by 9.51 percent. The overall and kappa accuracy test’s average value is 97.67%, higher than the required minimum of 75%.

ENDAU ROMPIN NATIONAL PARK LAND USE LAND COVER CHANGES USING REMOTE SENSING APPROACH

Identification of land use and land cover in forest areas can be challenging due to various land cover types within a forest can be similar, making it hard to differentiate between them using remote sensing techniques. We hypothesized that random forest classification (RF) would outperform maximum likelihood (ML) in the classification of land use and land cover (LULC) in forest areas compared to maximum likelihood (ML). To verify this hypothesis, we conducted a comparative analysis, assessing the accuracy of RF and ML in the classification of LULC within the Taman Negara Endau-Rompin (TNER) region, utilizing Landsat 8 imagery. An accuracy assessment demonstrated that the RF classifier (overall accuracy: 87% kappa coefficient: 0.778, performed better than ML classifying land cover (overall accuracy 77% with kappa accuracy: 0.473) Our results suggest that both methods are able to classify land cover of forest area, but RF is more accurate than ML. From the classification result of RF classification, we calculate the land cover changes of TNER from 2013 to 2022. results showed that there are small changes of forest area were found in TNER. The total forest area decreases from 163250.089 ha to 144765.46 ha during 2013 to 2022. This finding suggests that the effectiveness of the protected area in mitigating deforestation in its surrounding regions may be somewhat limited, as indicated by the observed minor changes.

ANALISIS PERUBAHAN PENGGUNAAN LAHAN DI DAS BATANG TABIR KABUPATEN MERANGIN, PROVINSI JAMBI (Land Use Change Analysis in Batang Tabir Watershed Merangin Regency, Jambi Province)

A continuous land use change in a watershed will affect the hydrological characteristics of the watershed. The Batang Tabir watershed has experienced a decrease in its carrying capacity shown by an increase in the frequency of flooding and the area of flood inundation. This study analyzes the land use changes in the Batang Tabir watershed, Merangin Regency, Jambi Province. The data used are DEM, topographic maps, river network maps, and Landsat imagery. Analysis of land use change used the Maximum Likelihood Classification (MLC) and the accuracy of image interpretation was carried out through the Kappa accuracy test. The land use prediction using the ANN model. The results showed that the rate of land use change increased in secondary forests, built-up areas, agriculture, and open areas; primary forests, paddy fields, and shrubs continued to decline. It is predicted in the year 2030, the trend will be the same; the built-up areas, agriculture, and secondary forests will increase, while primary forests, paddy fields, shrubs, and open areas will decrease.