CC BY
1
УДК 528.92
DOI 10.52575/2712-7443-2023-47-4-593 -606
Using GIS Tools to Detect the Land Use/Land Cover Changes in Ha Nam province, Vietnam
Bui B. Thien
Southern Federal University, 40 Zorge St, Rostov-on-Don 344090, Russia E-mail: buibaothienha@gmail.com
Abstract. Land is a crucial natural resource for any country. The study of land use and land cover (LULC) change has been instrumental in various areas such as natural resource management, monitoring, land planning, landslides, erosion, and addressing global change issues. In this study, geographic information systems (GIS) and remote sensing (RS) techniques were used to monitor LULC changes in Ha Nam province, Vietnam from 1992 to 2022. The supervised classification method in ArcGIS 10.8 software was applied to Landsat satellite data (Landsat 5-TM for 1992 and 2003, and Landsat 8-OLI/TIRS for 2022) to detect and classify five main LULC types: agricultural land, barren land, built-up, forest, and waterbodies. The classification accuracy was evaluated using kappa coefficients, which were 0.886, 0.905, and 0.933 for 1992, 2003, and 2022, respectively. During the period of 1992-2022, the agricultural land, forest, and waterbodies classes areas decreased by 102.85 km2, 48.57 km2, and 5.25 km2, respectively. Meanwhile, the built-up and barren land classes areas increased by 150.08 km2 and 6.59 km2, respectively. Population growth, urbanization, urban planning policies, and the transition from an agricultural to an industrial economy have contributed to the expansion of built-up areas and the reduction of agricultural land, forests, and waterbodies in Ha Nam province. Moreover, we utilized the Normalized Difference Vegetation Index (NDVI) and Normalized Difference Built-up Index (NDBI) to rapidly evaluate LULC changes, and we observed that their trends aligned with the results obtained from supervised classification. The environment faces substantial risks due to these LULC changes, and the outcomes of this study can provide valuable insights for upcoming land management and planning initiatives in the area.
Key words: Landsat, GIS, land use/land cover change, supervised classification, Ha Nam province
For citation: Thien B.B. 2023. Using GIS Tools to Detect the Land Use/Land Cover Changes in Ha Nam province, Vietnam. Regional Geosystems, 47(4): 593-606. DOI: 10.52575/2712-7443-2023-47-4-593-606
Использование инструментов ГИС для обнаружения изменений в использовании и покрытии земли в провинции Ханам, Вьетнам
Тхиен Б.Б.
Южный федеральный университет, Россия, 344090, г. Ростов-на-Дону, ул. Зорге, 40 E-mail: buibaothienha@gmail.com
Аннотация. Почва - это важный природный ресурс для любой страны. Изучение изменений в использовании и покрытии земли (LULC) играет важную роль в различных областях, таких как управление природными ресурсами, мониторинг, земельное планирование и решение проблем глобальных изменений. В этом исследовании использовались географические информационные системы (ГИС) и технологии дистанционного зондирования (ДЗ) для мониторинга изменений LULC в провинции Ханам, Вьетнам с 1992 по 2022 год. Метод классификации с обучением в программном обеспечении ArcGIS 10.8 применялся к данным спутника Landsat (Landsat 5-TM для 1992 и 2003 годов, и Landsat 8-OLI/TIRS для 2022 года) для выявления и классификации пяти основных типов LULC: сельскохозяйственные угодья, бесплодные угодья, застроенные территории, лес и водоемы. Точность классификации оценивалась с использованием
коэффициентов каппа, которые составили 0,886, 0,905 и 0,933 для 1992, 2003 и 2022 годов, соответственно. За период с 1992 по 2022 годы площади классов сельскохозяйственных угодий, леса и водоемов уменьшились на 102,85 км2, 48,57 км2 и 5,25 км2, соответственно. В то время как площади классов застроенных территорий и бесплодных угодий увеличились на 150,08 км2 и 6,59 км2, соответственно. Рост населения, урбанизация, политика городского планирования и переход от аграрной к индустриальной экономике способствовали расширению застроенных территорий и сокращению сельскохозяйственных земель, лесов и водоемов в провинции Ханам. Кроме того, мы использовали индекс нормализованной разницы вегетационного покрытия (NDVI) и индекс нормализованной разницы застроенных территорий (NDBI) для быстрой оценки изменений в LULC и обнаружили, что их тенденции соответствуют результатам, полученным с помощью надзорной классификации. Эти изменения в LULC представляют серьезные риски для окружающей среды, и результаты этого исследования могут предоставить ценные исследовательские данные для предстоящих инициатив по управлению и планированию земельных ресурсов в этом регионе.
Ключевые слова: Landsat, ГИС, изменение использования земли/покрытия земли, классификация с обучением, провинция Ханам
Для цитирования: Тхиен Б.Б. 2023. Использование инструментов ГИС для обнаружения изменений в использовании и покрытии земли в провинции Ханам, Вьетнам. Региональные геосистемы, 47(4): 593-606. DOI: 10.52575/2712-7443-2023-47-4-593-606
Introduction
In recent years, studies on global environmental changes have increasingly focused on the issue of land use and land cover (LULC) changes [Msofe et al., 2019]. Researchers recognize LULC as primary factors that impact ecosystems and landscape values. As a result, LULC changes has gained significant attention in the global academic community [Tariq et al., 2022]. The research has primarily concentrated on comparing methodologies using GIS methods, fuzzy sets, and landscape metrics. Additionally, there has been an effort to develop a new approach that combines ecological, geographical, and social anthropological data in LULC changes studies [Vadrevu et al., 2019]. A study was presented by scientists, employing a high-resolution land use change model to downscale land use changes from macro-scale models to the landscape level [Aghsaei et al., 2020]. This approach aims to provide valuable insights for future land use change analysis. Understanding the relations and interactions between anthropogenic factors and the natural environment is crucial in comprehending LULC changes [Zadbagher et al., 2018; Thien et al., 2022]. Both these factors influence LULC changes to varying degrees. However, the current trends indicate a general degradation of the environment and significant fragmentation of the landscape. Numerous studies conducted globally highlight the rapid pace of LULC changes resulting from population growth, intensive land use, and the loss of natural areas [Thien, Phuong, 2023].
Studying urban dimensions, including LULC mapping, urban density, urban modeling, and the environmental effects of urban development over time intervals, can be effectively carried out using powerful tools such as the geographic information system (GIS) and remote sensing (RS) [Mehdi et al., 2016; Majeed et al., 2021]. RS data offers timely, reliable, and accurate information on degraded lands during specific time periods in a cost-effective manner. By utilizing GIS technology, spatial data can be managed and analyzed according to the requirements of the case study [Phuong, Thien, 2023b]. RS data proves valuable for conducting LULC inventory and mapping. Landsat sensors like Landsat-5 Thematic Mapper (TM), Landsat-7 Enhanced TM Plus (ETM+), and Landsat-8 Operational Land Imager/Thermal Infrared Sensor (OLI/TIRS) provide a range of satellite data that plays a crucial role in detecting changes in Normalized Difference Vegetation Index (NDVI), Normalized Difference Built-up Index (NDBI) and LULC for LULC planners [Zheng et al., 2021; Florim et al., 2021; Dash et al., 2023]. Change detection involves quantitatively analyzing the previous effects of an occurrence
using RS information, thereby assisting in identifying changes related to LULC properties with reference to various satellite datasets. Supervised classifications rely on prior knowledge of the scene regions, areas containing materials of interest, and training sites, which are stored and delineated for use in the supervised classification algorithm.
This study used GIS and RS technology to monitor LULC changes from 1992 to 2022 in Ha Nam province, Vietnam. The purpose of our research is (1) to identify and classify LULC types and to quantitatively analyze LULC changes from 1992 to 2022; (2) then conduct NDVI and NDBI change detection, mapping and analysis using satellite data; and (3) to evaluate the factors affecting the change of LULC in the study area in the period 1992-2022.
Materials and methods
Study area
Ha Nam province, located in the North of Vietnam, is situated at geographical
coordinates with latitude 20°23'N - 20°52'N and longitude 105°49'E - 106°25'E, covering a total
2 1
area of 861.92 km (Fig. 1) . The landscape of Ha Nam province is diverse, encompassing plains, hills, and river valleys. The majority of the province's land area consists of plains, while hilly terrain and rivers make up the remaining portions. The province experiences a humid subtropical climate, characterized by four distinct seasons: spring, summer, autumn, and winter. The average annual temperature ranges from 23 °C to 25 °C, with hot and humid summers and relatively cold winters. As of the latest available data, the population of Ha Nam province was approximately 878,306 people in 2022. The urban population accounts for around 45 % of the total population, residing in cities and towns, while the remaining 55 % represents the rural population. The province has been witnessing ongoing urbanization and economic development. Ha Nam province's economy is diverse, with agriculture, industry, and services being the main sectors. The province is known for its agricultural production, particularly in rice cultivation. Industrial activities are also prominent, with various industrial zones attracting investments and contributing to the economic growth of the province. Additionally, the services sector, including commerce, healthcare, and education, plays a vital role in the local economy.
1НГК
106°0'Ü"E
lllfr ICVII' l-
iSl Cliina r • w Taiwan
Laos С / ;СЫП5<' Hoang Sa , Islands
Thailand "s
_ ) , >\VIETNAM si
..A Cambodia \ ' w -¿'l-i Truong Sa Island^ > '.<"* " Г . W T*Hi]jpp*iiie|
Legend
Country boundaries ^(|ProvinGe boundaries Ha Nam province ROB
■■Red: Rand_7 H Green: Band_5 iH Blue: Band 3
■ Km
и it nin и u r. um IIMI L у ~ щ
Fig. 1. Study area map of Ha Nam province Рис. 1. Карта территории исследования провинции Ханам
1 General Statistics Office. 2022. Statistical Yearbook of Viet Nam 2022. Statistical Publishing House. URL: https://www.gso.gov.vn/wp-content/uploads/2023/06/Sach-Nien-giam-TK-2022-update-21.7_file-nen-Water.pdf
Data collection
We used satellite images obtained from the USGS Glovis website (https://glovis.usgs.gov) to map LULC in Ha Nam province and assess LULC changes from 1992 to 2022. In this study, we employed Landsat 5-TM images for the years 1992 and 2003, while Landsat 8-OLI/TIRS images were utilized for 2022. To evaluate the accuracy of the LULC classification map, we collected point data, consisting of 300 points per year. For 1992 and 2003, Google Earth Pro software was employed to collect these points, while for 2022, we conducted field surveys and used GPS devices. Throughout the study, we utilized ArcGIS 10.8 and Microsoft Excel 2016 software. A comprehensive data summary can be found in Table 1.
Table 1 Таблица 1
Detailed data summary of satellite imagery used in the study Подробные данные спутниковых снимков, использованных в исследовании
Satellite image Sensor Acquisition data Path/row Landsat scene ID
Landsat 5 TM 01/12/1992 126/046 LT51260461992336BJC00
21/10/1992 127/046 LT51270461992295BJC02
Landsat 5 TM 16/12/2003 126/046 LT51260462003350BJC00
23/12/2003 127/046 LT51270462003357BKT01
Landsat 8 OLI/TIRS 17/10/2022 126/046 LC81260462022290LGN00
16/10/2022 127/046 LC81270462022289LGN01
Image pre-processing and supervised classification
We combined distinct bands from Landsat 5-TM and Landsat 8-OLI/TIRS through layer stacking to create a comprehensive image of the study area. During the subset setup process, we delineated the desired study area using extract by mask tools within ArcGIS 10.8 [Kumari et al., 2019; Phuong, Thien, 2023 a]. Referring to the scheme proposed by Anderson et al. [1976] and verifying it through field surveys, we identified five primary LULC categories in the study area: agricultural land, barren land, built-up areas, forest, and waterbodies (Table 2). Using ArcGIS 10.8 software, we drew polygons around pixels with similar reflectance values for each category, forming training samples [Verma et al., 2020]. Pixels enclosed by these polygons in each Landsat image were marked to extract spectral signatures for different LULC classes [Thien et al., 2023]. Next, we applied a maximum likelihood classification algorithm to classify LULC based on these spectral signatures [Verma et al., 2020; Isma'il et al., 2023]. Fig. 2 provides a detailed illustration of the methodology employed in this research.
Table 2 Таблица 2
Classes delineated from field survey Классы, выделенные на основе полевых исследований
Class Description
Agricultural land Cropland and paddy field
Barren land Fallow land, sands and earth dumps
Built-up Residential, industrial, roads and other manmade structures
Forest Natural forest, plantations and mixed forest lands
Waterbodies Reservoirs, rivers and lakes
Fig. 2. Flow chart for methodology Рис. 2. Блок-схема методологии
Classification accuracy assessment
During the LULC classification process, it is important to evaluate the accuracy to account for potential misclassifications between pixels [Zadbagher et al., 2018; Thien, Phuong, 2023 a]. To assess the accuracy of the classification results, we utilized an error matrix and compared the results with 150 reference data points collected for each year. We calculated various metrics based on the error matrix for each respective year, including user's accuracy, producer's accuracy, overall accuracy, and kappa coefficient [Vadrevu et al., 2019; Phuong, Thien, 2023a]. These metrics provide valuable insights into the reliability of the classification results by measuring the agreement between the predicted and actual classifications. Formulas (1), (2), (3), and (4) were employed to calculate the user's accuracy, producer's accuracy, overall accuracy, and kappa coefficient, respectively.
User's accuracy =
Producer's accuracy =
Number of correctly classified pixels in each category Total number of reference pixels in each category ( row total ) Number of correctly classified pixels in each category
Total number of reference pixels in each category (column total)
x100 (1)
x100 (2)
Overall accuracy =
x 100
(3)
Number of sampling classes classified correctly Number of reference sampling classes
Po - Pe Kappa = „ (4)
1 - Pe
where Po is the agreement ratio between the predicted classification results and the actual classification results. Pe is the random agreement ratio between the predicted classification results and the actual classification results.
Estimation and correlation between NDVI and NDBI
By utilizing satellite imagery, we can estimate the NDVI and NDBI, which provide valuable information for monitoring vegetation health and urbanization processes [Florim et al., 2021]. The NDVI serves as a vegetation index, utilizing the near-infrared (NIR) and red (RED) bands of satellite images to distinguish vegetation [Florim et al., 2021]. As vegetation cover ex-
pands, the NDVI value increases, while it decreases with diminishing vegetation cover. On the other hand, the NDBI serves as an urban index, utilizing the shortwave infrared (SWIR) and near-infrared (NIR) bands of satellite images to detect built-up areas [Zheng et al., 2021]. The NDBI value increases as built-up areas expand and decreases with a reduction in built-up areas. We calculate the NDVI and NDBI indices using formulas (5) and (6), respectively.
NDVI =
NDBI =
NIR - RED NIR + RED
SWIR - NIR SWIR + NIR
(5)
(6)
We utilized regression analysis to quantify the correlation between NDVI and NDBI in Ha Nam province for the years 1992, 2003, and 2022. The regression analysis yielded correlation coefficient values within the range of -1 to +1 [Pal, Ziaul, 2017]. To conduct the regression analysis, we employed the random point generator feature within ArcGIS 10.8 software to generate 200 random point data within the study area boundaries. The extract multi values to points tool facilitated the extraction of a value for each point data from the NDVI and NDBI pixels. Subsequently, we exported these values to Microsoft Excel 2016 software (Microsoft, USA) to estimate the regression equation between NDVI and NDBI.
Results and discussion
Land use/land cover classification
The map of LULC status in Ha Nam province in the three years 1992, 2003 and 2022 is shown in Fig. 3. Table 3 shows the area and proportions of each LULC type respectively. From Fig. 3 and the data in Table 3, it can be seen that significant changes have occurred in agricultural land and built-up classes over the 30 years in the study area.
(a) 1992
(b)2003
(c) 2022
ilil^J ' ЩШш
Legend
Agricultural land Barren land ■■ Built-up IH Forest ■ Waterbodies
s 1«
2«
JO
¡Km 4«
106°0'0"E
Fig. 3. Land use/land cover maps for Ha Nam province in 1992 (a), 2003 (b), and 2022 (c) Рис. 3. Карты землепользования/земельного покрова провинции Ханам в 1992 г. (a), 2003 г. (b) и 2022 г. (c)
Based on the LULC classification results in Table 3, in the year 1992, agricultural land class
accounted for the largest area in Ha Nam province, accounted for 78.17 % (673.75 km ) of the total
area. Forest class area accounted for 15.88 % (136.85 km2), waterbodies class area accounted for 22 2.80 % (24.16 km ), barren land class area accounted for 2.05 % (17.71 km ), and built-up class area
had the smallest coverage at only 1.10 % (9.45 km2) (Table 3). By 2003, the areas of forest and
22 waterbodies classes had decreased to 11.47 % (98.85 km ), and 2.03 % (17.53 km ), respectively
(Table 3). In contrast, the areas of agricultural land, barren land, and built-up classes had increased to
81.12 % (699.21 km2), 3.15 % (27.12 km2), and 2.23 % (19.21 km2), respectively (Table 3). By 2022,
the area of built-up class had further increased and accounted for 18.51 % (159.53 km2). Additionally,
the area of waterbodies class had continued to increase and accounted for 2.19 % (18.91 km ) in 2022.
Meanwhile, the areas of agricultural land, barren land, and forest classes had decreased to 66.24 %
(570.90 km2), 2.82 % (24.30 km2), and 10.24 % (88.28 km2), respectively (Table 3).
Table 3 Таблица 3
The land use/land cover area distribution from 1992 to 2022 in Ha Nam province Распределение землепользования/земельного покрова с 1992 по 2022 год в провинции Ханам
Class 1992 2003 2022
Area (km2) % Area (km2) % Area (km2) %
Agricultural land 673.75 78.17 699.21 81.12 570.90 66.24
Barren land 17.71 2.05 27.12 3.15 24.30 2.82
Built-up 9.45 1.10 19.21 2.23 159.53 18.51
Forest 136.85 15.88 98.85 11.47 88.28 10.24
Waterbodies 24.16 2.80 17.53 2.03 18.91 2.19
Total 861.92 100.00 861.92 100.00 861.92 100.00
The assessment of the post-classification accuracy in this study was performed by comparing the classified LULC classes with the reference data [Islami et al., 2022; Thien, Phuong, 2023 a]. The results of the classification evaluation showed that the overall accuracy of the years 1992, 2003, and 2022 was 91.58 %, 93.00 %, and 95.00 %, respectively (Table 4). Overall, the producer's accuracy and the user's accuracy for each soil layer in all 3 years were above 80 % (Table 4). The kappa coefficient values in 1992, 2003 and 2022 in the study area were recorded as 0.886, 0.905, and 0.933, respectively (Table 4). Kappa coefficients ranging from 0.81 to 1.00 are considered almost perfect in LULC classification [Regasa et al., 2021; Wahla et al., 2023]. These results show reliable land cover classification and good consistency between referenced and classified maps.
Table 4 Таблица 4
Accuracy assessments for classified maps Оценка точности классифицированных карт
LULC classes 1992 2003 2022
Producer's accuracy (%) User's accuracy (%) Producer's accuracy (%) User's accuracy (%) Producer's accuracy (%) User's accuracy (%)
Agricultural land 94.87 91.36 94.87 90.24 95.83 95.83
Barren land 90.91 83.33 91.67 84.62 93.33 82.35
Built-up 83.33 88.24 88.89 92.31 96.61 98.28
Forest 90.57 96.00 93.75 97.83 93.33 96.55
Waterbodies 90.32 93.33 91.43 96.97 91.67 91.67
Overall accuracy (%) 91.58 93.00 95.00
Kappa Coefficient 0.886 0.905 0.933
Land use/land cover change
Fig. 4 illustrates the specific changes in each LULC class during the period of 1992-2022
in Ha Nam province. The analysis of area changes for each LULC class in different periods
(1992-2003, 2003-2022, and 1992-2022) is also presented in Table 5. During the period of
1992-2003, the forest area experienced the highest decrease by 4.41 % (38.00 km2) compared to
the initial area. The decrease in forest area could be attributed to illegal logging, conversion of
forests to agricultural or built-up land, and urban expansion. The waterbodies also decreased by
0.77 % (6.63 km ). The reduction in waterbodies could be due to drought, water
mismanagement, and climate change. The decrease in waterbodies may impact water availability
for domestic use and agricultural irrigation. In contrast, the agricultural land, barren land, and
built-up areas increased by 2.95 % (25.46 km2), 1.09 % (9.41 km2), and 1.13 % (9.76 km2),
respectively. The increase in agricultural land could be a result of agricultural expansion to meet
the growing demands for food production and population growth. Meanwhile, the built-up area
has increased possibly due to urban development, infrastructure construction and other
construction projects to accommodate population and economic growth. Examining the LULC
change model during the period of 2003-2022, a significant decrease of 14.89 % (128.31 km )
was observed in agricultural land. The decrease in agricultural land could be due to the
conversion of agricultural land to built-up or other land types, shifting economic structure with
increased industrial and service activities. Additionally, the barren land and forest areas
22 decreased by 0.33 % (2.82 km ) and 1.23 % (10.57 km ) respectively. The reasons for the
decrease in barren land might include climate change, unsustainable resource exploitation, and
the expansion of other activities on barren land. Meanwhile, the built-up area continued to
increase during the 2003-2022 period, reaching 16.28 % (140.32 km2) (Table 5). The increase in
built-up areas could be a result of urban development, expansion of industrial zones, and other
construction projects. Furthermore, in the period of 2003-2022, the waterbodies experienced a
slight increase, with a total increase of 0.16 % (1.38 km2) (Table 5). The slight increase in
waterbodies could be attributed to improved water resource management, environmental
restoration efforts, and conservation measures.
Table 5 Таблица 5
The land use/land cover change analysis from 1992 to 2022 in Ha Nam province Анализ землепользования/изменения земельного покрова с 1992 по 2022 год в провинции Ханам
Class 1992-2003 2003-2022 1992-2022
Area (km2) % Area (km2) % Area (km2) %
Agricultural land 25.46 2.95 -128.31 -14.89 -102.85 -11.93
Barren land 9.41 1.09 -2.82 -0.33 6.59 0.76
Built-up 9.76 1.13 140.32 16.28 150.08 17.41
Forest -38.00 -4.41 -10.57 -1.23 -48.57 -5.64
Waterbodies -6.63 -0.77 1.38 0.16 -5.25 -0.61
In general, over the past 30 years (1992-2022) in the study area, there have been significant changes in LULC. Fig. 4 and Table 5 show that the built-up area has continuously increased, with a total increase of 17.41 % (150.08 km2), and the highest increase occurred during the period of 2003-2022. There are several factors contributing to the expansion of built-up areas, with the main factors being population growth, urbanization, and urban planning policies [Waiyasusri, 2021; Singh et al., 2022]. The population growth during this period has led to an increasing demand for housing, driving the expansion and development of urban areas and residential areas in Ha Nam province [Niu et al., 2022].
Fig. 4. Land use/land cover changes map for Ha Nam province from 1992 to 2022 Рис. 4. Карта изменений землепользования/земного покрова провинции Ханам с 1992 по 2022 год
Additionally, the process of urbanization and the expansion of urban areas have resulted in the development of new urban areas, the renovation and expansion of urban infrastructure, and the construction of urban areas outside existing urban centers [Wang et al., 2021]. During the period of 1992-2022, due to rapid urbanization, a significant portion of agricultural land, forests, and water bodies had to be converted for construction purposes [Getu Engida et al., 2021; Herrera Arango et al., 2022]. The total area of agricultural land, forest, and waterbodies decreased by 11.93 % (102.85 km2), 5.64 % (48.57 km2), and 0.61 % (5.25 km2), respectively (Table 5). Furthermore, previously, the main economic activity in the study area was rice cultivation, but water shortages for irrigation due to drought conditions led the local authorities to gradually shift towards an industrial economy. Meanwhile, the area of barren land increased by 0.76 % (6.59 km2) during the period of 1992-2022, mainly resulting from housing and industrial construction projects. Overall, the increase in built-up areas in Ha Nam province from 1992 to the present can be attributed to population growth, urbanization, and urban planning policies. The conversion of agricultural land, forests, and waterbodies for construction purposes was driven by the rapid urbanization process. Additionally, the shift from an agricultural to an industrial economy and the presence of abandoned land from housing and industrial projects also contributed to the expansion of built-up areas [Thekkeyil et al., 2023; Thien et al., 2023].
The NDVI and NDBI
The high NDVI index values indicate denser and healthier vegetation, while lower values correspond to sparse or no vegetation [Florim et al., 2021]. In 1992, the NDVI value ranged from -0.51 to +0.65 (Fig. 5a); in 2003, NDVI values ranged from -0.99 to +0.97 (Fig. 5b); and in 2022, NDVI values ranged from -0.17 to +0.55 (Fig. 5c). Significant spatial changes in vegetation cover and green area were observed between the lowest and highest NDVI values recorded in 2003, along with improved agricultural productivity in areas such as forests and vegetation cover (Figs. 5 a, b, c). The NDBI index is used to assess the level of urban development in the study area, the NDBI values increase as the built-up area increases and decreases when the built-up area decreases [Degerli, £etin, 2022]. In 1992, the NDBI value ranged from -0.96 to +0.78 (Fig. 5d); in 2003, NDBI values ranged from -0.97 to +0.98 (Fig. 5e); and in 2022, the NDBI values ranged from -0.41 to +0.31 (Fig. 5f) The red areas in Figs. 5d, 5e, and 5f show minimal vegetation cover, such as built-up and barren land.
Fig. 5. NDVI and NDBI maps for Ha Nam province in 1992, 2003, and 2022 Рис. 5. Карты NDVI и NDBI провинции Ханам в 1992, 2003 и 2022 годах
> _
Q
g -0.40
1992
0.40 |
eW*i* .
-0.20 •<)'»&% JTjo i
-0.2(Р»
I» •
% -0.40 , -0.60
n -lit ••
0.60
у =-0.087 lx +0.058 R2 .= 0.0067
NDBI
> -0.40
Q
К
2003
0.40 Ш
-0.20 • он -0.2»
-(?40
-0.60
-0.80
30
0.4(f
0.60
у = -0.2826x +0.1297 R2 = 0.0501
NDBI
2022
o.io
0.60
у --0.3552X- 0.0466 R2 = 0.2860
NDBI
Fig. 6. Regression analyses between NDVI and NDBI in Ha Nam province Рис. 6. Регрессионный анализ между NDVI и NDBI в провинции Ханам
A linear regression analysis was conducted to demonstrate the relationship between two indices (NDVI and NDBI) [Florim et al., 2021]. The changes in NDBI values related to land use were assessed by evaluating the variations in land use intensity within the LULC units through regression analysis (R ) [Majeed et al., 2021]. Furthermore, a negative correlation between NDVI and NDBI was identified. Specifically, correlation coefficients of R2 = 0.0067 for 1992, R2 = 0.0501 for 2003, and R2 = 0.2860 for 2022 were depicted in Fig. 6. As observed in Fig. 6, this illustrates the relationship between the vegetation index (NDVI) and the integrated component derived from NDBI. In addition, linear regression analysis shows that R values have gradually increased, which indicates that urbanization has negatively affected vegetation covers. The regression analysis also revealed that the highest NDBI values corresponded to areas with the lowest NDVI values, and vice versa. This clearly indicates that the increase in built-up areas and barren land leads to a decrease in vegetation coverage.
Conclusion
The study highlights that the combined use of GIS and RS techniques provides valuable insights into LULC changes. Ha Nam province has undergone significant LULC changes from 1992 to 2022, the study area has experienced a decline in agricultural land, forest, and waterbodies classes, with area decreases of 11.93 %, 5.64 %, and 0.61 %, respectively. In addition, there has been a significant increase in the areas of built-up and barren land classes from 1992 to 2022, with a total increase of 17.41 % and 0.76 % respectively. These changes are primarily driven by rapid urbanization, leading to deforestation, and conversion of agricultural land. In general, the effects of urbanization, population growth, and climate change create negative trends in land use, which causes a number of medical, economic, and environmental problems for research area. The NDVI and NDBI indices were also employed to evaluate changes
in land cover characteristics, revealing a strong correlation between impervious surfaces and vegetation cover. The data also indicated that there are limited natural resources and significant environmental areas that authorities could designate as hotspots for conservation or mitigation. The study emphasizes the importance of understanding these changes for effective conservation and mitigation efforts, utilizing spatial-temporal analysis enabled by GIS and RS technologies. The findings have implications for future policies promoting sustainable land use practices in Ha Nam province.
References
Aghsaei H., Dinan N.M., Moridi A., Asadolahi Z., Delavar M., Fohrer N., Wagner P.D. 2020. Effects of Dynamic Land Use/Land Cover Change on Water Resources and Sediment Yield in the Anzali Wetland Catchment, Gilan, Iran. Science of the Total Environment, 712: 136449. https: //doi.org/10.1016/j.scitotenv.2019.136449. Anderson J.R., Hardy E.E., Roach J.T., Witmer R.E. 1976. A Land Use and Land Cover Classification System for Use with Remote Sensor Data. In: Geological Survey Professional Paper, U.S. Government Printing Office. Washington DC: 1-28. Dash P., Sanders S.L., Parajuli P., Ouyang Y. 2023. Improving the Accuracy of Land Use and Land Cover Classification of Landsat Data in an Agricultural Watershed. Remote Sensing, 15(16): 4020. https: //doi.org/10.3390/rs 15164020. Degerli B., Cetin M. 2022. Evaluation from Rural to Urban Scale for the Effect of NDVI-NDBI Indices on Land Surface Temperature, in Samsun, Türkiye. Turkish Journal of Agriculture-Food Science and Technology, 10(12): 2446-2452. https://doi.org/10.24925/turjaf.v10i12.2446-2452.5535. Florim I., Albert B., Shpejtim B. 2021. Measuring UHI Using Landsat 8 OLI and TIRS Data with NDVI
and NDBI in Municipality of Prishtina. Disaster Adv, 14: 25-36. Getu Engida T., Nigussie T.A., Aneseyee A.B., Barnabas J. 2021. Land Use/Land Cover Change Impact on Hydrological Process in the Upper Baro Basin, Ethiopia. Applied and Environmental Soil Science, 2021: 1-15. https://doi.org/10.1155/2021/6617541. Herrera Arango J., Senent-De Frutos J.A., Molina E.H. 2022. Murky Waters: The Impact of Privatizing Water Use on Environmental Degradation and the Exclusion of Local Communities in the Caribbean. International Journal of Water Resources Development, 38(1): 152-172. https://doi.org/10.1080/07900627.2021.1931052. Islami F.A., Tarigan S.D., Wahjunie E.D., Dasanto B.D. 2022. Accuracy Assessment of Land Use Change Analysis Using Google Earth in Sadar Watershed Mojokerto Regency. In IOP Conference Series: Earth and Environmental Science, 950(1): 012091. https://doi.org/10.1088/1755-1315/950/1/012091.
Isma'il M., Zubairu S.M., Aliyu A., Ahmed M.H., Ibrahim S., Magaji A., Hassan I.M. 2023. Evaluating the Performance of Machine Learning Algorithms and Maximum Likelihood Classifier for Land-Use and Land-Cover Change Detection in Yola, Nigeria. SLU Journal of Science and Technology, 7(1): 97-115. https://doi.org/10.56471/slujst.v7i.380. Kumari M., Sarma K., Sharma R. 2019. Using Moran's I and GIS to Study the Spatial Pattern of Land Surface Temperature in Relation to Land Use/Cover Around a Thermal Power Plant in Singrauli District, Madhya Pradesh, India. Remote Sensing Applications: Society and Environment, 15: 100239. https://doi.org/10.1016Zj.rsase.2019.100239. Majeed M., Tariq A., Anwar M.M., Khan A.M., Arshad F., Mumtaz F., Farhan M., Zhang L., Zafar A., Aziz M., Abbasi S., Rahman G., Hussain S., Waheed M., Fatima K., Shaukat S. 2021. Monitoring of Land Use-Land Cover Change and Potential Causal Factors of Climate Change in Jhelum District, Punjab, Pakistan, Through GIS and Multi-Temporal Satellite Data. Land, 10(10): 1026. https://doi.org/10.3390/land10101026. Mehdi S.M., Pant N.C., Saini H.S., Mujtaba S.A.I., Pande P. 2016. Identification of Palaeochannel Configuration in the Saraswati River Basin in Parts of Haryana and Rajasthan, India, Through Digital Remote Sensing and GIS. Episodes Journal of International Geoscience, 39(1): 29-38. https://doi.org/10.18814/epiiugs/2016/v39i1/89234.
Msofe N.K., Sheng L., Lyimo J. 2019. Land Use Change Trends and Their Driving Forces in the Kilombero Valley Floodplain, Southeastern Tanzania. Sustainability, 11(2): 505. https://doi.org/10.3390/su11020505. Niu X., Hu Y., Lei Z., Yan H., Ye J., Wang H. 2022. Temporal and Spatial Evolution Characteristics and Its Driving Mechanism of Land Use/Cover in Vietnam from 2000 to 2020. Land, 11(6): 920. https://doi.org/10.3390/land11060920. Pal S., Ziaul S.K. 2017. Detection of Land Use and Land Cover Change and Land Surface Temperature in English Bazar Urban Centre. The Egyptian Journal of Remote Sensing and Space Science, 20(1): 125-145. https://doi.org/ 10.1016/j.ejrs.2016.11.003. Phuong V.T., Thien B.B. 2023a. A Multi-Temporal Landsat Data Analysis for Land-Use/Land-Cover Change in the Northwest Mountains Region of Vietnam Using Remote Sensing Techniques. Forum Geografic, 22 (1): 54-66. https://doi.org/10.5775/fg.2023.030.i. Phuong V.T., Thien B.B. 2023b. Using Landsat Satellite Images to Detect Forest Cover Changes in the Northeast Region of Vietnam. Bulletin of the Transilvania University of Brasov. Series II: Forestry Wood Industry Agricultural Food Engineering, 16(1): 19-36. https://doi.org/10.31926/but.fwiafe.2023.16.65.L2. Regasa M.S., Nones M., Adeba D. 2021. A Review on Land Use and Land Cover Change in Ethiopian
Basins. Land, 10(6): 585. https://doi.org/10.3390/land10060585. Singh P., Sarkar Chaudhuri A., Verma P., Singh V.K., Meena S.R. 2022. Earth Observation Data Sets in Monitoring of Urbanization and Urban Heat Island of Delhi, India. Geomatics, Natural Hazards and Risk, 13(1): 1762-1779. https://doi.org/10.1080/19475705.2022.2097452. Tariq A., Yan J., Mumtaz F., 2022. Land Change Modeler and CA-Markov Chain Analysis for Land Use Land Cover Change Using Satellite Data of Peshawar, Pakistan. Physics and Chemistry of the Earth, Parts A/B/C, 128: 103286. https://doi.org/10.1016Zj.pce.2022.103286. Thekkeyil A., George A., Abdurazak F., Kuriakose G., Nameer P.O., Abhilash P.C., Joseph S. 2023. Land Use Change in Rapidly Developing Economies - a Case Study on Land Use Intensification and Land Fallowing in Kochi, Kerala, India. Environmental Monitoring and Assessment, 195(9): 1089. https://doi.org/10.1007/s10661-023-11731-7. Thien B.B., Huong D.T.V., Liem N.D. 2022. Assessment of Mangroves Forest Change from Satellite Images in Can Gio District, Ho Chi Minh City in Period of 1990-2020. In: Proceeding of the 13th National Conference on Geography Science, Hanoi, 133-140. Thien B.B., Phuong V.T. 2023. Using Landsat Satellite Imagery for Assessment and Monitoring of Long-Term Forest Cover Changes in Dak Nong Province, Vietnam. Geographica Pannonica, 27(1): 6982. https://doi.org/10.5937/gp27-41813 Thien B.B., Yachongtou B., Phuong V.T. 2023. Long-Term Monitoring of Forest Cover Change Resulting in Forest Loss in the Capital of Luang Prabang Province, Lao PDR. Environmental Monitoring and Assessment, 195(8): 1-17. https://doi.org/10.1007/s10661-023-11548-4. Vadrevu K., Heinimann A., Gutman G., Justice C. 2019. Remote Sensing of Land Use/Cover Changes in South and Southeast Asian Countries. International Journal of Digital Earth, 12(10): 1099-1102. https://doi.org/10.1080/17538947.2019.1654274 Verma P., Raghubanshi A., Srivastava P.K., Raghubanshi A.S. 2020. Appraisal of Kappa-Based Metrics and Disagreement Indices of Accuracy Assessment for Parametric and Nonparametric Techniques Used in LULC Classification and Change Detection. Modeling Earth Systems and Environment, 6: 1045-1059. https://doi.org/10.1007/s40808-020-00740-x Wahla S.S., Kazmi J.H., Tariq A. 2023. Mapping and Monitoring of Spatio-Temporal Land Use and Land Cover Changes and Relationship with Normalized Satellite Indices and Driving Factors. Geology, Ecology, and Landscapes: 1-17. https://doi.org/10.1080/24749508.2023.2187567 Waiyasusri K. 2021. Monitoring the Land Cover Changes in Mangrove Areas and Urbanization Using Normalized Difference Vegetation Index and Normalized Difference Built-Up Index in Krabi Estuary Wetland, Krabi province, Thailand. Applied Environmental Research, 43(3): 1-16. https://doi.org/10.35762/AER.2021.43.3.1 Wang S., Bai X., Zhang X., Reis S., Chen D., Xu J., Gu B. 2021. Urbanization can Benefit Agricultural Production with Large-Scale Farming in China. Nature Food, 2(3): 183-191. https://doi.org/10.103 8/s43016-021 -00228-6
Zadbagher E., Becek K., Berberoglu S. 2018. Modeling Land Use/Land Cover Change Using Remote Sensing and Geographic Information Systems: Case Study of the Seyhan Basin, Turkey. Environmental Monitoring and Assessment, 190: 1-15. https://doi.org/10.1007/s10661-018-6877-y Zheng Y., Tang L., Wang H. 2021. An Improved Approach for Monitoring Urban Built-Up Areas by Combining NPP-VIIRS Nighttime Light, NDVI, NDWI, and NDBI. Journal of Cleaner Production, 328: 129488. https://doi.org/10.1016/jjclepro.2021.129488
Received September 18, 2023; Поступила в редакцию 18.09.2023;
Revised October 24, 2023; поступила после рецензирования 24.10.2023;
Accepted October 29, 2023 принята к публикации 29.10.2023
Conflict of interest: no potential conflict of interest related to this article was reported. Конфликт интересов: о потенциальном конфликте интересов не сообщалось.
INFORMATION ABOUT THE AUTHOR
Bui B. Thien, Master's student, Institute of Earth Sciences, Southern Federal University, Rostov-on-Don, Russia
ИНФОРМАЦИЯ ОБ АВТОРЕ
Буй Б. Тхиен, магистрант Института наук о Земле, Южный федеральный университет, Ростов-на-Дону, Россия
