Environmental Quality Monitoring Using Remote Sensing Ecological Index (RSEI) in Can Tho City, Vietnam.

The significance of green spaces and ecological priorities in urban areas–two major concepts and advancements in Vietnam's recent urban growth, particularly in Can Tho City–is addressed by urban planning. An ecological assessment necessitates participating in urban planning and the eco-environment protection system. This study evaluated the change in the ecological quality of Can Tho City in the period 2015 - 2020 by approaching remote sensing indicators and statistical methods. Using Landsat 8 OLI satellite data from 2015 to 2020, four indicators were calculated in this study: heat (LST), dryness (NDBSI), wetness (WET), and greenness (NDVI). These indices were then applied using the Principal Component Analysis (PCA) method to estimate the Remote Sensing Ecological Index (RSEI). According to the data, the RSEI mean was 3.66 in 2015 and 3.60 in 2020. In 2015, areas with good and very good ecological quality (EQ) accounted for 88.78% of the total natural area; by 2020, it was 84.75%. The areas with low quality increased from 1.97% in 2015 to 3.49% in 2020. This indicates a decreasing trend in ecological quality within the area. The declining trend in urban ecological quality (UEQ) at Can Tho City between 2015 and 2020 can be objectively reflected by the RSEI and the shifting proportions of the ecological level. The results of RSEI provide fundamental data on the ecological quality of the urban environment to support policymakers, managers, and authorities in implementing and developing sustainable approaches to urban ecological conservation and environmental quality preservation.


Introduction
Ecological quality for the environment implies the ecological parameters in an ecosystem that modify the development and sustainability of life for all organisms.It is defined within a specific spatial and temporal range [1].This issue is highly revealed in the sustainability of environmental issues to human life and socio-economic [2].Changes in land use and cover are the leading cause of urbanization (LUCC), which has a significant effect on the natural environment [3], [4].Urbanization has become irreversible during the development phase, and it is unlikely that ecological environment strains would force this issue to be reduced shortly.For this reason, the expansion of ecological environment diversity in urban and urbanization plans to contribute to socio-economic growth and the capacity research of the regional environmental problems, which have appropriate fundamental topics on urban ecology nowadays [5], [3], [6].
Examining the environment's quality reveals a significant change in the environment's agents.Several countries and international organizations initially developed a comprehensive rating system to assess environmental quality [7].One critical factor that substantially influences the life quality indices in urban situations is urban ecological quality (UEQ).UEQ measures how receptive the urban IOP Publishing doi:10.1088/1755-1315/1345/1/012018 2 environment is to people [8].Assessing ecological quality in the urban environment is a mandatory field in urban ecological research and the fundamental management of urban planning and ecological components.The urbanization process has provided people with a better standard of living in the cities, promoted the development process, and brought economic advances leading to the consequences of disrupting ecological balances within nature, environment, and society, such as natural resource depletion, economic structural imbalances, and environmental pollution, posing a significant challenge to urban ecological quality control [9].In 2015, the United Nations (UN) member states adopted the Sustainable Development Goals (SDGs) until 2030.Urbanization aims to improve people's living conditions, facilitate sustainable development, and deliver economic benefits, which has adjusted the balance and imbalance in human activity and nature in development and brought significant challenges to UEQ.
Cities are growing at a rapid rate, which increases urbanization and harms the environment.Urbanization converts natural lands into built-up areas, losing green spaces, wildlife habitats, and other ecosystems.It can cause soil erosion, air and water pollution, and the release of greenhouse gases, contributing to climate change.The growing demand for energy, food, and other resources puts additional strain on the environment and can lead to resource depletion and environmental degradation.Numerous studies have been conducted and presented in urban ecology to evaluate diverse viewpoints and methodological approaches.It is now in use because remote sensing technology collects highaccuracy real-time data in various places and is widely applicable in monitoring and processing ecological environment parameters.Two categories of ecological environment assessment exist singleindex and multi-index models [10].In reaction to extreme climatic occurrences in Southwest Vietnam, these models extract vegetation from Landsat OLI data using the Normalized Difference Vegetation Index (NDVI) [11] and analyse the influence of ground temperature on forest above-ground biomass.Due to the richness and complex unpredictability of natural systems, it isn't easy to precisely represent the ecological condition using a single ecological index, even though a single indicator can examine some links between causes and events that impact some amount [12] [13], which revised the remote sensing ecological index (RSEI) in 2013.The four environmental indicators of the RSEI are heat, dryness, wetness, and greenness.These indicators can swiftly, thoroughly, and objectively represent the region's ecological environment level [14].It allows for a quantitative analysis of the current conditions and changes in the Yangtze River's ecological environment.Principal Component Analysis (PCA) reduced the uncertainty contributions inside the automatically weighted computation to recommend these parameters.PCA was established and proved in a variety of ecosystem diversity (e.g., cities, woods, cropland, upland crops, wetlands) [15], [16], [17], [18], [19], [20], [21], [22].
Given its strategic location on the southern bank of the Hau River and Vietnam's rapid urbanization, Can Tho City serves as the hub of the Mekong Delta region.Can Tho is rapidly changing to become a modern, civilized, ecologically conscious city by 2030, infused with the distinctive Mekong Delta culture.By 2050, Can Tho have developed into a contemporary, sophisticated, ecologically aware metropolis alongside other cities in the Lower Mekong Delta, qualifying it for membership in the Club of Developed Nations in Asian cities. Can Tho also envision improving urban quality, adhering to a sustainable development model, and preserving the natural biological environment on land, in water, and the surrounding environment.The application of remote sensing data for quantitative monitoring and ecological environmental quality evaluation has received very little scholarly attention.Remote sensing data offers an overview of ground-level data for a long-term study of the ecological environment quality using valuable methods.
Based on the above reality, the study was conducted to evaluate the quality of Can Tho City's ecological environment based on the remote sensing ecological index (RSEI) approach.The study is expected to (1) provide an overview of Can Tho City's ecological quality in 2015 and 2020; (2) assess changes in ecological quality in the period 2015 -2020.

Study area
Can Tho City boasts a 1,401 km 2 natural environment and a riverside of over 65 km.It's situated on the left bank of the Hau River.It is the largest city in the level delta of the Mekong River and a central industrial hub.The nine administrative units consist of five counties: Ninh Kieu, Binh Thuy, Cai Rang, O Mon, and Thot Not.The four districts are Vinh Thanh, Co Do, Thoi Lai, and Phong Dien.Each of the nine counties has five towns, forty-two wards, and thirty-six communes.The population is around 1.4 million people with 3 main ethnicities: Kinh, Khmer, and Hoa.The temperature is pleasant and somewhat muggy.26-270 degrees Celsius is the typical monthly temperature.With 1,157 km of complex canals and rivers, including the three major rivers in the area-Hau, Cai Lon, and Can Tho-Can Tho is considered a river city.Regarded as Vietnam's urban architectural heritage from the French colonial era, the city is a municipality comparable to the nation's provinces; by 2020, it will be an industrialized metropolis, serving as the socio-economic, health, education and training, science and technology, and cultural center of the Mekong Delta region.It will also be a significant hub for local and foreign traffic (Fig. 1).

Remote sensing data and Image Preprocessing
This study explores, analyzes, and evaluates the ecological quality of the environment in Can Tho City using Landsat images as a data source.Landsat 8 OLI images with 30m spatial resolution, WGS 84 zone 48N reference system, and 16-day image repetition period were collected on the Google Earth Engine (GEE) platform in 2015 and 2020 in the study area via Catalog (ID="LANDSAT/LC08/C02/T1_L2").
Under the condition of cloud cover, a total of 15 images were collected in 2015 and 11 images were collected in 2020.
The first step in pre-processing Landsat 8 images is reflectance calibration.Next, the Top-Of-Atmosphere (TOA) Level-1C orthoimage outputs are preprocessed with the FLAASH model to apply the atmospheric correction.Next, cloud data was eliminated using the maskL8SR function, which was constructed based on the band "QA Pixel" at the study area.The clip() command on GEE is used to limit the study area to eliminate redundant areas, reducing the probability of misclassifying data because Landsat remote sensing images cover a wider area than the study area.

Remote Sensing Ecological Index Estimation
RSEI is a composite index based on Remote Sensing data and principal component analysis (PCA).It includes four indicatorsgreenness, wetness, heat, and dryness, which are related to physiological factors of climate and land surface that are perceived by humans and are closely linked to the Environment Ecological [23].Besides, applying principal component analysis (PCA) helps eliminate subjective human analysis factors in determining weights for data [24].RSEI is a method used in regional ecological environment assessment and monitoring because of the comprehensiveness, reliability, and accessibility of the data, which is clearly shown in previous research [25,26,27,28].
Using the Remote Sensing Ecological Index (RSEI) allows for accurately determining Can Tho's required degree of eco-environmental quality.The four indications of heat, moisture, dryness, and greenness are combined to create the RSEI, which can only be determined with remote sensing data.Each shows additional elements of the natural environment, such as temperature, soil moisture, greenery, and built-up area, as represented by equations (1) [13].RSEI = f(Greenness, Wetness, Heat, Dryness) The greenness index is commonly used to monitor vegetation growth and vegetation cover, which is used to represent the vegetation index and a comprehensive response to the growth of surface vegetation with the most widely using the Normalized Difference Vegetation Index (NDVI).The wetness index represents the moisture component (WET) is used to represent the moisture index, which reflects the surface water content and the regional water-heat balance.The heat index represents temperature quantified by Land Surface Temperature (LST) which reflects the energy flow and material exchange of the soil-vegetation-atmosphere system, which reflects the condition of the uncovered surface and surface damage.The dryness index mainly refers to bare land and built-up area which can be expressed by the Normalized Difference Bare Soil Index (NDBSI).The Index-based Built-up Index (IBI) and Soil Index (SI) are used as mean values to represent the dryness index to observe the thermal pollution of the area, and to assess the possibility of natural hazards at the site.An index value can be calculated using the formula below: The components of vegetation, often known as the greenness index, are represented by the Normalized Difference Vegetation Index (NDVI), which may be calculated as (3) [29]: in which NIR is a near-infrared band, and RED is the red band in the Landsat 8 images.
After calculating the four indicators of NDVI, Wet, NDSI, and LST, the dimensions of the four indices are different.Each indicator should be normalized so that the four values are mapped between 0 and 1. Equation reveals the formula (10).
In which: Xij is the initial value of pixel j in index i and Yij is the normalized value of pixel j in index i; Maxi and Mini are the largest and smallest values of index i, respectively.
After the four component indexes have been processed, the RSEI indicators are combined using the principal component analysis method.Principal component analysis, or PCA, has the advantage of automatically and impartially determining each indicator weight by examining each index's contribution to the first principle component (PC1) [31].Equation (11) illustrates that the first output component is the initial remote sensing ecological index (RSEI PC1): RSEI0 = PC1[f(NDVI, LST,WET, NDBSI)] (11) RSEI0 must also be normalized for the computed RSEI value during the measurement procedure to enable comparability.The normalized value ranges from 0 to 1, where the high value shows the ecological difference, the low value shows the ecological good.When it is nearer 1, the ecological environment's quality is higher.Higher RSEI values were deducted from 1, typically used to indicate more ecological situations.The formula for normalizing is displayed in equation ( 12) [32].(12) where the remote sensing ecological index, or RSEI, is utilized.The starting value of the remote sensing ecological index is denoted as RSEI0.The initial variable RSEI0 has two values, RSEI0 max and RSEI0min, representing its minimum and maximum, respectively.
The standardized RSEI value is divided into five levels of ecological quality according to Table 1 [13].

Spatial analysis and map creation
The data analysis method and map editing established a layer attribute table of land use and ecological quality, calculated the area of each quality level, and created a map design using the Group Start and Editor tool on QGIS 3.26 software.

Characteristics of changes in component variables in the RSEI index
When the greenness score is high, it indicates a high rate of vegetation coverage and a higher quality ecological environment [33].On the other hand, low greenness values were associated with lower levels of vegetation cover and lower ecological environment quality standards ( [34], [35]).The greenness was represented by the normalized difference vegetation index (NDVI) in 2015 and 2020 which was calculated and normalized to the value range [0-1].The overall distribution characteristics of NDVI in 02 years are the same.High NDVI values (blue) are distributed throughout the study area, areas with low NDVI values are mainly distributed in the Southeast.In 2020, the range of low values in the southeast increased significantly (light brown), showing that the region's vegetation cover was decreasing.
The higher the wetness value and the higher the surface water content, the better the quality of the ecological environment and vice versa [36].The wetness index (WET) in 2015 and 2020 was calculated and standardized to the value range [0-1].The research area's humidity dropped between 2015 and 2020.In 2015 and 2020, the wetness index had an average value of 0.44 and 0.31, respectively.In 2015, the WET value decreased significantly compared to 2020 (Table 2).The land area that has experienced degradation and lower ecological and environmental quality is more adversely affected by a more significant dryness (NDBSI) score.However, low NDBSI values are associated with good ecological quality [37].The dryness index values are calculated and normalized to the value range [0-1] in 2015 and 2020.The average value of the NDBSI index for the period was 0.19 and 0.27, respectively.As a result, residential land construction in the area increased, and land quality deteriorated sharply as economic development and urbanization sped up.It was shown that some places had more dryness indices than others, which may have contributed to the deterioration of the biological environment.

Principal component analysis of environmental factors
The contribution ratio of the eigenvalue of the first principal component of RSEI was above 60%, and the contribution ratio of each index has the same positive and negative distribution in the first principal component.Specifically, the contribution ratio of NDVI (representing vegetation cover) and WET (representing environmental wetness) in the first principal component was positive.LST (representing ground surface temperature) and NDBSI (representing impermeability of buildings and bare soil) were negative values (Table 3).This result indicates how important WET and NDVI were in improving the promotion of the natural environment.The eigenvalue percentage (%) on PC1 was anticipated to be 67.67% in 2015 and 65.30% in 2020, respectively, indicating that LST and NDBSI only partially support the ecological environment.
From 2015 to 2020, PC1's dryness index (NDBSI) contributed -0.165 and -0.293, respectively.The dryness index shows changes in construction land based on urban planning in Can Tho city, which has been shown to significantly impact the quality of the urban ecological environment.Urban expansion has resulted from the development of the building process, which has grown to be a significant factor impacting the ecological environment quality in the city.

Ecological environment quality in Can Tho City in the period of 2015 -2020
The RSEI spatial distribution of the 2 years of the study, comprehensively illustrates the ecological conditions of the city and the spatial changes of RSEI during the study period.The red patches represent areas of low to very low ecological conditions.Green -Blue patches show moderate to very good ecological conditions.

Ecological environment quality in Can Tho city in 2015
The ecological quality results in 2015 for Can Tho City were categorized into four groups: very good, good, moderate and low (Fig. 3).In particular, very good and good ecological quality was mainly distributed in suburban areas such as Vinh Thanh, Thot Not, Co Do, O Mon, Thoi Lai, Phong Dien districts and part of the area in 02 central districts Binh Thuy and Cai Rang; Moderate and low ecological quality was concentrated in 03 central districts of the City (Ninh Kieu, Binh Thuy, Cai Rang), and part of the area was concentrated in the central area of suburban districts, industrial parks, and along traffic routes.4), and the RSEI areas and percentages are identified in five levels (I to V) from very low to very good levels in 2015 and 2020 shown on Table 5.The areas of ecological quality levels from III to V counted more than 95%, and the remaining two levels (I and II) reported values less than 5%.Those statistical results reveal that Can Tho City still had an very good ecological environment quality from 2015 to 2020.The average ecological quality value for each district of Can Tho City from 2015 to 2020 is displayed in Table 6.In central districts such as Ninh Kieu, Binh Thuy, and Cai Rang, RSEI values were at levels II -Low and III -Moderate.This has shown that in central districts, urban ecological quality was at a relative level.In suburban districts such as Phong Dien, O Mon, Thot Not, Co Do, Thoi Lai, and Vinh Thanh, the average RSEI level was in the range greater than IV -Good.This has shown that although there were industrial activities occurring in the suburbs, overall, the ecological quality in the suburban districts was still considered to be at an acceptable level (Table 6).The spatial distribution of RSEI (Fig 3 and Fig 4) has shown that areas with good ecological environments were mainly distributed in urban fringe areas around central districts.These areas have a relatively low economic base and the dominant land use was arable land and woodland, with rich vegetation and high levels of biodiversity.Areas with low ecological quality were concentrated in 03 districts of Ninh Kieu, Binh Thuy, and Cai Rang, and in some administrative centers of the remaining administrative units.Ninh Kieu serves as the hub and is home to many universities, colleges, and businesses, as well as a dense population and commercial activity.Tra Noc 1 Industrial Park and Hung Phu 1 Industrial Park are two examples of industrial parks and non-industrial clusters active in the Binh Thuy district.Cai Rang's location in the city center draws people engaged in growing commercial and service sectors, industrial production, and other economic activities that immediately influence the environment.

Spatial and temporal changes in the ecological environment.
From 2015 to 2020, the area with good and very good ecological quality decreased, from 88.775% to 84.749%.Throughout the period, the average RSEI shifts from 3.659 to 3.595.Table 6 illustrates the results for 2015 to 2020, which indicate a gradual decline in the mean value of RSEI in Can Tho's suburban districts.This decline in the quality of the natural environment is attributed to the city's economic transformation on the construction of transportation infrastructure, particularly roads built in residential neighborhoods and activity centers.
According to data from the Provincial Statistics Department, the gross regional product (GRDP) of Can Tho City increased at an average annual rate of 7.53% between 2015 and 2020.The city's GRDP per capita reached 97.2 million VND, an increase of 1.65 times from 2015.It represents the lowest rate of poverty in the Mekong Delta.Between 2015 and 2020, the population grew from 1,208.50 to 1,240.73 thousand, resulting in changes to the ecological quality caused by human activities such as deforestation, increased construction, and soil pollution.Rapid economic development, population density, and per capita GDP growth have increased human activities, causing damage to the surface ecological environment.In particular, the impact of rapid urban population growth was related to the reclamation of new land and the destruction of the original land structure to serve construction and development activities of infrastructure, zones -industrial park clusters.This affects soil quality, including surface runoff, soil water infiltration, and soil water content, leading to reduced surface vegetation cover and reduced water infiltration capacity.Developing industrial zones requires a lot of space, consumes a lot of energy, and discharges pollution.

Conclusion
Using the ecological index of remote sensing (RSEI), this research evaluated the eco-environmental quality evaluation on four ecological indicators including NDVI, LST, WET, and NDBSI to assess Can Tho's ecological environment, the quantitative characteristics of these variables are controlled using principal component analysis (PCA) showing relatively good levels from 3.595 to 3.659 in 2015 and 2020 at Can Tho city.The ecological quality area with good and very good levels decreased from 88.775% to 84.749% in the 5-years from 2015 to 2020.
Ecological environment monitoring can raise living standards and maintain social order.It is understanding the intricacies of the ecosystem and the detrimental consequences of human activity on both the environment and human health.This research achieved specific results that have been contributing to ecological and environmental governance for policymakers' references that contribute information to help policymakers monitor ecological environmental quality and plan land use for the future.

Figure 2 .
Figure 2. Index image factors in 2015 and 2020.(a): NDVI in 2015; (b) NDVI in 2020; (c) WET in 2015; (d) WET in 2020; (e) NDBSI in 2015; (f) NDBSI in 2020; (g) LST in 2015; (h) LST in 2020.The blue and red colors in Figure 2 represent areas with low and high heat values.The area's ecological quality decreases with increasing heat value.The average LST values in 02 years are 0.26 and 0.34, respectively.In the distribution of LST in 2015 and 2020, high-temperature areas (yellow-red color) include industrial park areas, industrial park clusters, district administrative centers, and traffic routes, while non-urban and river areas have suitable temperatures (green color).

Figure 3 .
Figure 3. Ecological quality map of Can Tho City in 2015 4.2.2.Ecological Environment Quality in Can Tho City in 2020The ecological quality status of Can Tho City was divided into five categories for 2020: very good, good, moderate, low, and very low (Fig4).).In particular, very good and good levels were distributed in most districts/districts of the city, including: Vinh Thanh, Thot Not, Co Do, O Mon, Thoi Lai, Phong Dien, and a part of the outer area including: Binh Thuy, Cai Rang districts; Moderate ecological quality was mainly distributed on main traffic routes and the center of Cai Rang and Binh Thuy districts.Next,

Figure 4 .
Figure 4. Urban ecological quality map of Can Tho City in 2020The mean RSEI value has a slight change value (0.064) between 2015 and 2020 at Can Tho City (Table4), and the RSEI areas and percentages are identified in five levels (I to V) from very low to very good levels in 2015 and 2020 shown on Table5.The areas of ecological quality levels from III to V counted more than 95%, and the remaining two levels (I and II) reported values less than 5%.Those statistical results reveal that Can Tho City still had an very good ecological environment quality from 2015 to 2020.

Table 1 .
Ecological quality hierarchy table

Table 2 .
The average values of factors in 2015 and 2020

Table 3 .
The Principal Component Analysis results

Table 4 .
Average ecological quality in 2015 and 2020

Table 5 .
Area of ecological quality in 2015 and 2020

Table 6 .
Average RSEI in the period 2015 -2020 in the Districts of Can Tho City