The Ideal Strategy of Carbon-Neutral for Park Landscape Design: A Proposal for a Rapid Detection Method

Abstract

The primary objective of this study is to investigate the carbon footprint, resilience levels, and optimal landscape area ratios of various parks. Additionally, it explores the relationships between landscape element proportions (LEP), the normalized difference vegetation index (NDVI), resilience indicators (RI), and the carbon reduction benefits associated with carbon neutrality (CN). Six parks were assessed for resilience, NDVI, LEP, and CN values, with Pearson correlation analysis conducted. The results revealed that parks with or without waterbodies exhibited ideal LEP area ratios of 6.5:2:1.5 (Softscape:Waterbody:Hardscape) and 8.3:1.7 (Softscape:Hardscape), respectively. Enhanced Softscape and reduced Hardscape proportions in parks correlated with increased NDVI and CN. NDVI exhibited a positive correlation with Softscape percentage and a negative correlation with Hardscape percentage. Conversely, CN demonstrated a negative correlation with Hardscape percentage and a positive correlation with Softscape percentage. Suggesting Softscape should constitute over 65%, and Hardscape should be under 15% in parks with water bodies. Waterless parks are advised to maintain a Softscape ratio exceeding 83% and a Hardscape ratio below 17%. Finally, the study extended to assess the LEP of 22 additional parks, validating the suitability of the ideal LEP area ratio.

1. Introduction

According to statistics, carbon dioxide (CO₂) accounts for 80% of all greenhouse gases (GHGs) emitted by the European Union [1]. The establishment of comprehensive emissions reduction policies and the implementation of effective Carbon Sequestration methods are essential for reducing GHG emissions [2]. Due to the significant carbon emissions generated during the construction of parks in Taichung City, which have severely impacted the surrounding environment, the Taichung City government has decided to promote low-carbon sustainable development. The Shuinan Economic and Trade Park, including its park areas, has been designated as a key carbon reduction demonstration zone, with the aim of leading other regions’ parks in jointly advancing carbon reduction initiatives [3]. Within urban settings, the most significant public spaces for residents are green areas [4,5], where vegetation plays a critical role in offsetting CO₂ emissions. Through the process of photosynthesis during growth, plants store carbon [6]. Previous research has confirmed a correlation between carbon sequestration and urban parks [7]. The greenhouse effect poses severe environmental and disaster-related challenges for cities. Based on statistical data from 1998 to 2017, the two most frequent types of disasters were floods (43.4%) and storms (28.2%) [8]. Enhancing park resilience provides advantages such as increasing soil carbon storage capacity [9] and improving plant diversity [10]. Therefore, enhancing both the carbon sequestration capacity and the resilience of parks not only aids in absorbing CO₂ but also strengthens the disaster-prevention capabilities of parks, fostering a healthy ecological environment. These complementary strategies are pivotal in achieving the goal of carbon neutrality for parks.

“Landscape elements” encompass a multitude of natural and human-made features within a landscape, including flora, water features, structures, terrain, paving, artworks, and more. The study of these elements can notably enhance urban aesthetics. Investigating landscape components significantly amplifies a city’s visual appeal [11,12,13]. The harsh urban pavement prevalent in roads and squares negatively impacts both residents and parks, whereas the inclusion of expansive lawns, water features, and diverse vegetation can foster their well-being [13,14]. “Softscape” refers to vegetation integrated within a landscape, such as shrubs and flowers. Conversely, “Hardscape” denotes structural components incorporated into the landscape, including elements like fountains, benches, or gazebos [15,16,17]. The NDVI is defined by its primary calculation of the difference between red and near-infrared reflectance. This index is highly effective in assessing chlorophyll and other pigment contents and serves as an excellent proxy for vegetation density parameters such as Leaf Area Index (LAI) and Fractional Vegetation Cover (FVC) [18,19,20,21]. Remote sensing through the NDVI is a valuable tool to evaluate and monitor vegetation cover [21,22]. Leveraging NDVI-generated remote sensing data offers benefits through time-series insights, broad coverage, and high spatial resolution [23]. The study’s additional focal point concerns the “resilience indicators” (RI), designed to evaluate the resilience of urban parks. These indicators are categorized into indices for flood resilience and wind resilience [24]. This paper will reference the RI formulated in the aforementioned study for several reasons: The system has been applied to Taichung City in Taiwan and has been used to assess numerous parks, affirming its feasibility. RI serves as a vital tool for evaluating the resilience of parks. Initially, the RI employs the Fuzzy Delphi Method to filter indicators, subsequently utilizing the Analytic Network Process to calculate the respective weights of each indicator. Carbon neutrality (CN) involves achieving a balance between carbon emissions and carbon absorption, with the goal of reaching net-zero emissions by using carbon sink mechanisms to absorb atmospheric carbon. Specifically, the process of removing CO₂ from the atmosphere and storing it long-term is known as carbon sequestration. To achieve global net-zero emissions, all greenhouse gas emissions must be offset through carbon sequestration and similar methods to mitigate the impacts of climate change [25]. CN refers to reducing or offsetting carbon emissions to bring net carbon emissions to zero as a means of combating climate change and reducing the release of greenhouse gases [26,27,28]. In summary, this study categorizes prominent landscape elements in urban parks into three groups: Softscape (tree cover and green space), Waterbody, and Hardscape (paving, buildings, pavilions, and trellises). These landscape component ratios, referred to as the LEP, will be determined later in the study. Monitoring changes in NDVI allows us to assess plant cover conditions. Employing RI, we will evaluate park resilience and estimate CN potential by assessing carbon sequestration capability.

Previous studies can be classified into four main categories: assessing green spaces’ carbon footprint in parks for carbon neutrality, relevant calculation tools and formulas, strategies to enhance parks’ carbon neutrality through design, and the proportion of green spaces, plantings, and water bodies. The first category encompasses research on the carbon footprint of landscape design, including nursery trees [29], school campuses [30], community landscapes [31], and landscape engineering perspectives [32]. Some studies quantify carbon footprints and evaluate reduction effects. The second category focuses on calculation tools and formulas, such as those for tree carbon uptake, urban green space carbon storage [33], and life cycle carbon footprints [32]. The third category offers strategies for improvement and design, suggesting native plant selection, enhancing three-dimensional greenery, planning rational plant configurations, and reducing construction emissions [34,35,36]. Lastly, the fourth category examines the ratio of park green space, plantings, and water bodies. For instance, Tainan City’s building coverage rate is 12% to 15% [37], and water bodies range from 0% to 52.3% with at least 30% green coverage [38]. Unfortunately, Taichung City has not yet issued specific policies or regulations regarding the proportions of green spaces, vegetation, and water bodies in parks. In the future, this study can only reference relevant policies from other cities in Taiwan as a guide for park design and improvement. Scholars generally employ life cycle assessment for urban park carbon footprint analysis and offer design solutions, with some proposing landscape element proportions. Thus, additional research should explore other attributes of carbon-neutral parks.

This study finds that the five stages proposed by scholar Yang offer greater feasibility and comprehensiveness in the research on park carbon footprints [6]. Notably, it encompasses routine park maintenance and restoration, enhancing the analysis of urban park carbon footprints. Yang’s research yielded the publication of “Carbon Footprint of the Construction Industry—Carbon Management Strategies for Architecture, Landscape, and Interior Decoration” [24]. This book outlines calculation methods, project carbon emission coefficients, and carbon reduction advantages in total carbon footprints and carbon neutrality for landscaping projects. It elucidates the benefits of carbon reduction in achieving carbon neutrality and offers insights into quantifying carbon footprints in landscaping projects. In contrast to other studies, this research evaluates urban park carbon footprints, NDVI, and resilience.

Extensive scholarly research has previously been conducted on carbon sequestration in urban parks. Approaches thus far have included utilizing developed quantitative models for each tree species to estimate the carbon sequestering capabilities of trees in parks [39], adopting established botanical carbon sequestration formulas to assess carbon fixation capacity [1], and measuring various tree attributes—such as tree height, diameter at breast height, aboveground biomass, and age—as well as estimating the carbon fixation amount of turfgrass, thereby calculating the annual carbon sequestration rates of parks [40]. Upon measuring tree characteristics, Wang et al. applied formulas for plant carbon sequestration from the literature to assess carbon fixation capability and referred to the allometric growth equations of biomass cited in references to estimate aboveground biomass [7]. As indicated by the methods employed in the aforementioned literature, field surveys of parks are typically conducted to gather fundamental data on various tree species, attributes of turf grass, areas covered by trees, water bodies, and so forth. This is followed by the application of established formulas or models for plant carbon sequestration and the calculation of the aboveground biomass of the vegetation. These steps ultimately lead to an estimation of the total carbon sequestration capacity of the park. To accurately calculate the carbon emissions over the entire lifecycle of a park, it is critical to incorporate a LCA. For instance, research conducted by scholar Nicese quantifies the carbon emissions during the planting, maintenance, and waste management phases of park development. Following the establishment of a Life Cycle Inventory (LCI), carbon emissions for each component are computed. For example, during the tree planting stage, emissions are calculated for the production and transport of seeds and fertilizers, as well as for the energy consumed by construction equipment (tractors and forklifts). Utilizing this method allows for not only an assessment of environmental impact but also an effective control over the operational costs of park management [5]. The aforementioned analysis details the meticulous calculation of carbon sequestration for each plant species by scholars studying the LCA of urban parks and carbon sequestration. The results, estimated with precision, also showcase certain disadvantages. Firstly, when calculating the aboveground biomass of individual plants, researchers can refer to existing literature for biomass equations; however, they frequently encounter the absence of species-specific equations. While some scholars suggest employing generalized equations as a substitute, such equations carry inherent uncertainties that lack extensive experimental validation. Secondly, following the methods detailed in the literature requires substantial time for data retrieval and calculations [41,42]. Consequently, this study aims to discover a straightforward and rapid assessment method by disregarding the specific traits of plants. In the practice of landscape architecture, a simple and expedient evaluation of the carbon emissions over a park’s full lifecycle becomes a critical step towards achieving carbon neutrality in parks. Crucially, for landscape designers and managers, the ability to estimate carbon emissions at the early stages of park construction or enhancement facilitates more frequent and effortless computation. This, in turn, enables the swift formulation of carbon reduction and design strategies and the rapid monitoring of the progress towards carbon neutrality goals, thus enhancing the resilience of parks.

In summary, this study aims to achieve four objectives: (1) quantify carbon footprints of diverse parks, (2) assess park resilience, (3) propose optimal LEP area ratios, and (4) explore links between LEP, NDVI, and CN. Through these objectives, the study aims to assess parks’ potential for carbon neutrality and resilience while offering recommendations to attain carbon neutrality. Through the fulfillment of these objectives, this study aims to provide landscape architects and managers with a simple and rapid tool for evaluating the carbon neutrality and resilience of parks.

2. Materials and Methods

2.1. Research Process

The study is divided into eight steps, which will be detailed below:

Step 1: Six parks in Taichung City, Taiwan, were selected, varying in size from 0.1 to 6 hectares. These parks include Xindu Ecological Park, Lixin Park, Maple Garden, Yongquan Park, Litchi Old Tree Park, and Dayuan Park. Among these, only the first four include water bodies, enabling a comparison between park types. The locations of the parks are shown in Figure 1. By selecting parks with diverse characteristics, it is possible to compare the performance of different types of parks in terms of CN, ensuring that the study results are representative and scientifically valid.

Step 2: The orthophoto images of each park were obtained using drones and Landsat 8-OLI satellite imagery. This step significantly influences the selection of study sites, as the choice of parks is largely constrained by the availability of construction budget documents and design plans, which are essential for carbon footprint calculations. Due to the difficulty in obtaining such documents, the study is currently limited to analyzing and processing data from the six parks for which these materials were available. Effectively utilizing tools to obtain orthophotos ensures the provision of precise geographic information and detailed surface features, which are crucial for subsequent carbon footprint and NDVI calculations. Additionally, selecting parks with complete construction budgets and design plans ensures the accuracy of carbon footprint calculations, as these documents form the foundation for conducting carbon footprint analysis.

Step 3: Carbon footprints were determined based on construction materials and elements outlined in the compiled Parks Project Estimates, incorporating cited references for CN calculations. The calculation of the carbon footprint is a crucial step in assessing carbon neutrality efficiency. By conducting an in-depth analysis of the various construction activities involved in park development, the total carbon emissions of the park can be accurately quantified, providing a solid data foundation for subsequent correlation analysis.

Step 4: NDVI values for each park were computed using ENVI 5.3 software between March and April 2023. Calculating NDVI is an indispensable step in assessing a park’s ecological benefits and carbon neutrality potential, as it directly illustrates the critical role of vegetation in the carbon neutrality process.

Step 5: This step is based on the park resilience indicators established in Pan et al.’s study on park resilience [24]. The study divides the indicators into two categories: flood resilience indicators (ratio of lawn to water permeability, drainage capacity grade of different soil types, detention basin grade, and permeable pavement grade) and wind resilience indicators (bearing capacity grade of trunk and root systems, tree height and wind resistance grade, wind resistance grade of tree species, and configuration level of multilayer planting). This park resilience evaluation system is used to assess the resilience levels of the parks. A park’s ability to adapt and withstand floods and windstorms has a profound impact on its long-term carbon neutrality efficiency. Therefore, it is essential to conduct a comprehensive and systematic evaluation of the park’s ecological stability and disaster resilience.

Step 6: LEP percentages for each park were determined, offering insights into the most effective components for achieving carbon-neutral parks. The purpose of calculating LEP is to quantify the contribution of different landscape elements to carbon neutrality efficiency. This provides a basis for optimizing park design and enhancing carbon neutrality performance.

Step 7: Pearson correlation analysis was conducted to confirm interrelationships among LEP, NDVI, and CN variables within the parks. This step helps identify key factors that significantly influence carbon neutrality efficiency, offering guidance for optimizing park design.

Step 8: To validate ideal LEP ratios and extend insights, the LEP of 22 additional parks was calculated alongside the study’s chosen parks. By increasing the number of research samples through this step, the validity of the ideal LEP ratio can be further verified, providing valuable insights for a broader range of parks.

Step 9: Based on the findings from each park, recommendations for enhancing parks’ carbon neutrality were formulated, addressing deficiencies where carbon neutrality standards were not met. Based on improvement recommendations for each park, practical guidelines can be provided for the construction and renovation of future parks, thereby enhancing their carbon neutrality efficiency and achieving the goal of park carbon neutrality.

2.2. Data Collection and Processing

2.2.1. Landscape Engineering’s Carbon Footprint and Carbon Neutrality’s Carbon Savings

This study calculates the total carbon footprint of landscape engineering for each park by referencing the various items listed in the park project budget and consulting major carbon coefficient websites in Taiwan. Given the complexity of the budget items, the study categorizes all work items into four main categories: earthwork, construction materials, labor, and planting. The carbon footprint for each of these categories is then calculated separately to determine the total carbon footprint. Excel software was used to organize and calculate the carbon footprints, yielding the total carbon footprint and CN value for each park. Based on the carbon footprint calculation references and literature mentioned in the introduction, only a few key formulas are listed in

Table 1. The primary formula for calculating carbon neutrality.

Item	Formula	Source
Landscape engineering total carbon footprint	TCF = CFmc + CFrm + CFeu + CFdw − Cfa − CFo	[5,40,43,44]
	TCF: Total Carbon Footprint of Landscape Works (kgCO2e)
	CFmc: Carbon Footprint of Materials and Construction of New Construction (kgCO2e)
	CFrm: Carbon Footprint of Defect Repair and Renewal (kgCO2e)
	CFeu: Carbon Footprint of Daily Use (kgCO2e)
	CFdw: Carbon Footprint of Dismantling and Disposing of Wastes (kgCO2e)
	CFa: Carbon Sequestration Equivalent of Planting (kgCO2e)
	CFo: Carbon reduction from low-carbon building materials and equipment (kgCO2e)
Total Carbon Footprint of Cut and Fill Volumes (kgCO2e)	(Fuel Consumption of Equipment × Carbon Emission Factor of Fuel) + (Transportation Distance × Carbon Emission Factor of Transportation Process) + (Total Filling Volume × Carbon Emission Factor of Filling) + (Total Excavation Volume × Carbon Emission Factor of Excavation)	[40,43,44]
Total Carbon Footprint of Construction Materials (kgCO2e)	${\displaystyle {\displaystyle \sum }_{i=1}^n}\left(\mathrm{Quantity} \mathrm{of} \mathrm{material}\times \mathrm{carbon} \mathrm{emission} \mathrm{factor} \mathrm{of} \mathrm{material}\right)$	[5,40,43,44]
Workers’ Total Carbon Footprint (kgCO2e)	${\displaystyle {\displaystyle \sum }_{i=1}^n}\left(\mathrm{Number} \mathrm{of} \mathrm{hours} \mathrm{worked} \mathrm{by} \mathrm{workers} \times \mathrm{carbon} \mathrm{emission} \mathrm{factor} \mathrm{of} \mathrm{work} \mathrm{activities}\right)$	[43,44]
Calculating the Total Carbon Footprint of Planting Vegetation (kgCO2e)	Arbor (diameter <10 cm): ∑(number of plants × total carbon emission 4.30)	[1,39,43,44]
	Trees (diameter 10–15 cm): ∑(Number of plants × Total carbon emission 5.60)
	Trees (diameter 15–20 cm): ∑(Number of plants × Total carbon emission 8.0)
	Trees (diameter 20–25 cm): ∑(Number of plants × Total carbon emission 11.25)
	Shrubs, grasses, bagged seedlings (1–3 inches): ∑(Number of plants × Total carbon emission 0.01)
	Shrubs, grasses, bagged seedlings (4–7 inches): ∑(Number of plants × Total carbon emission 0.02)
	Lawn (m2): ∑(Area × Total carbon emission 0.11)
Carbon Sequestration of Planting (CFa)	CFa (general parks and green spaces) = 30.0 × Nt × LC	[43,44]
	CFa: total carbon sequestration of tree greening (kgCO2e)
	Nt: number of trees (plants)
	LC: landscape engineering life cycle standard (yr)
Efficiency of Carbon Neutrality (kgCO2e)	$\mathrm{CN}={\displaystyle \frac{\mathrm{Total} \mathrm{Carbon} \mathrm{Sequestration}-\mathrm{Total} \mathrm{Carbon} \mathrm{Emissions}}{\mathrm{Total} \mathrm{Carbon} \mathrm{Emissions}}}\times 100\%$	[43,44]

.

2.2.2. Calculation of NDVI Value of Each Park

The normalized difference vegetation index (NDVI) is the most commonly used spectral indicator in ecology and agriculture for assessing vegetation conditions [44,45,46]. NDVI values are categorized as follows: 0.60–1.00 indicates the highest density, 0.45–0.60 is high, 0.30–0.45 is medium, 0.15–0.30 is low, 0.00–0.15 is the lowest, and −1 to 0 may represent water [47]. Utilizing ENVI 5.3 software, standardized NDVI values are computed through remote sensing data processing and NDVI value extraction [48,49].

In this study, Landsat 8-OLI satellite images, obtained from the U.S. Geological Survey and covering the entire study area with less than 6% cloud cover [50], were processed using ENVI 5.3 software. First, atmospheric correction was applied to the images, converting the original Digital Number (DN) values to radiance and then to surface reflectance. Each image was subsequently cropped to the appropriate size and extent and extracted using the study area mask in ESRI ArcGIS ArcMap 10.5 software. Based on the Landsat 8-OLI band configuration, bands 2 to 5, corresponding to blue, green, red, and near-infrared (NIR), were selected for spatial subsetting. Vegetation indices for each image were then calculated using the band math tool in ENVI 5.3. The NDVI was computed according to the standard formula (NIR-Red)/(NIR + Red), representing the difference in reflectance between the NIR and red bands [51].

2.2.3. Evaluation of Resilience of Each Park

This study aims to assess the resilience levels of various parks, not only to serve as temporary shelters but also to mitigate the impact of disasters on urban areas. Based on the research by Pan et al., a park resilience index system was established, comprising indicators for flood resilience and wind resilience, each further divided into four sub-indicators. Each indicator is rated on a scale of 1 to 5. After assigning resilience scores to the parks, these scores are multiplied by the respective indicator weights to calculate the total RI [24], as detailed in

Table 2. The resilience index system of flood and wind disasters in the park.

Overall Objective	Guidelines	Weight	Indicators	Weight
The resilience index system of flood and wind disasters in the park	Flood resilience index (A)	0.75	Lawn and water permeability ratio (A1)	0.25
			Drainage capacity grade of different types of soil (A2)	0.2
			Grade of detention pond (A3)	0.18
			Permeable pavement grade (A4)	0.3
	Wind disaster resilience index (B)	0.25	Bearing capacity grade of trunk and root (B1)	0.07
			Tree Height and Wind Resistance grade (B2)	0.07
			Wind Resistance grade of Tree Species (B3)	0.06
			Configuration grade for multilayer planting (B4)	0.05

. This resilience index system has also been applied to evaluate the resilience of 81 parks in Taichung City, Taiwan, providing targeted recommendations for parks with lower resilience levels.

2.2.4. Calculation of the Proportion of LEP in Each Park

AutoCAD 2022 software was utilized to compute tree cover, green space (grass, shrubs), water bodies, and rugged terrain (pavement, structures, pavilions, scaffolding) in each park’s orthophoto. While this approach does not yield absolute precision, it provides relatively accurate data for determining item proportions.

3. Results

3.1. Calculation Result of LEP Proportion of the Park

Based on the LEP distribution and proportions presented in Figure 2 and

Table 3. Landscape Element Proportion.

No.	Park Name	Landscape Element Proportion (%)
		Softscape		Waterbody	Hardscape
		Tree Cover	Green Space
1	Xindu Park	35.77	43.19	5.85	15.19
2	Lixin Park	25.19	42.38	6.54	25.89
3	Maple Garden	13.03	19.2	11.43	56.34
4	Yongquan Park	9.48	10.5	36.05	43.97
5	Litchi Old Tree Park	63.89	22.23	-	13.88
6	Dayuan Park	13.17	57.77	-	29.06

, it is evident that among the four parks featuring water bodies, Xindu Ecological Park boasts the largest Softscape portion at 78.96%, while Yongquan Park has the smallest at 19.98%. For parks with water bodies, Yongquan Park holds the highest water proportion at 36.05%. Comparatively, among the two parks lacking water bodies, Litchi Old Tree Park has the larger Softscape portion, making up 86.12%.

3.2. Calculation Result of NDVI Value

As illustrated in Figure 3 and

Table 4. NDVI index of each park.

No.	Park Name	NDVI
1	Xindu Park	0.62
2	Lixin Park	0.54
3	Maple Garden	0.34
4	Yongquan Park	0.21
5	Litchi Tree Old Park	0.65
6	Dayuan Park	0.54

, Xindu Ecological Park exhibits the highest NDVI, attributed to consistent plant maintenance and minimal human disturbance, promoting lush vegetation. The park’s abundant trees, lawns, and substantial vegetation coverage contribute to its elevated NDVI. In contrast, Yongquan Park has the lowest NDVI due to extensive rigid pavement hindering plant growth. Furthermore, limited plant coverage results from a small Softscape and expansive Hardscape. Across all parks, the average NDVI stands at 0.48, signifying substantial vegetation density.

3.3. Calculation Result of RI Value

Based on the assessment of park resilience presented in the table, Xindu Ecological Park emerges with the highest RI, attributed to its extensive lawns, less intricate pavement, increased permeable surfaces, sizable flood detention capacity, and robust wind resistance. In contrast, Dayuan Park exhibits the lowest RI due to limited permeable surfaces and the absence of a flood detention basin. Moreover, the park’s recent establishment leads to wind-sensitive saplings, as indicated in

Table 5. The resilience index system of flood and wind disasters for six parks.

No.	Park Name	The Score of Each Indicator								The Total Score after the Weight
		A1	A2	A3	A4	B1	B2	B3	B4
	Index Weigh (%)	24.8	19.93	17.73	12.54	6.72	6.6	6.39	5.28
1	Xindu Ecological Park	5	4	3	4	4	3	4	5	4.06
2	Lixin Park	4	4	4	3	3	4	3	3	3.69
3	Maple Garden	2	5	5	4	4	3	3	2	3.64
4	Yongquan Park	2	5	5	1	3	4	3	2	3.27
5	Litchi Old Tree Park	5	5	1	4	4	5	4	2	3.88
6	Dayuan Park	5	4	1	3	2	1	2	1	2.97

.

3.4. Results of CN Value Calculation

Utilizing the reference materials and calculation formulas provided earlier, calculate the budgetary items for each park’s engineering proposal. The CN values for each park are presented in

Table 6. Statistical table of the total carbon footprint of each park landscape project.

No.	Park Name	Cut and Fill	Construction Materials	Worker	Planting	Total Carbon Emissions of the Park	Carbon Sequestration of Plantings	Carbon Neutrality
1	Xindu Park	36,314	34,941	365	2152	73,773	89,100	20.78%
2	Lixin Park	237,588	75,946	-	3198	316,733	389,700	23.04%
3	Maple Garden	186,181	108,188	3263	1194	298,826	98,100	−67.17%
4	Yongquan Park	1,220,008	1,919,717	356	6199	3,146,280	514,800	−83.64%
5	Litchi Old Tree Park	36,549	34,500	265	1154	72,468	88,200	21.71%
6	Dayuan Park	57,567	43,301	1705	875	103,448	100,800	−2.56%

Note: Unit (kg CO₂e).

. Notably, the dominant contributors to carbon emissions across parks are the extent of fill, excavation, and construction materials, with worker count and vegetation playing minor roles. Xindu Ecological Park, Lixin Park, and Litchi Old Tree Park meet the carbon-neutral criteria, attributed to limited fill, excavation, low carbon emissions from construction materials, abundant plantings and trees, and significant carbon sequestration by the vegetation, surpassing the park landscape engineering’s overall carbon footprint. The CN values for Maple Garden, Yongquan Park, and Dayuan Park are negative. A common factor among these parks is that the carbon emissions generated from Cut and Fill operations and Construction materials are significantly high, while the carbon sequestration from vegetation is insufficient to offset these emissions.

3.5. Analysis of LEP, NDVI, RI, and CN Values of All Parks

Based on the comprehensive analysis of LEP, NDVI, RI, and CN values presented in

Table 7. Statistical table of LEP, NDVI, RI, and CN.

No.	Park Name	Landscape Element Proportion (%)			NDVI	RI	CN (%)
		Softscape	Waterbody	Hardscape
1	Xindu Park	78.96	5.85	15.19	0.62	4.06	20.78
2	Lixin Park	67.57	6.54	25.89	0.54	3.69	23.04
3	Maple Garden	32.23	11.43	56.34	0.34	3.64	−67.17
4	Yongquan Park	19.98	36.05	43.97	0.21	3.27	−83.64
5	Litchi Old Tree Park	86.12	-	13.88	0.65	3.88	21.71
6	Dayuan Park	70.94	-	29.06	0.54	2.97	−2.56

, a prevailing pattern emerges. Parks with a Softscape value exceeding 65% tend to exhibit a reasonable NDVI value and a positive CN value, signifying the potential to offset construction-related carbon emissions through increased planting and vegetation. The NDVI demonstrates an upward trajectory as Softscape percentages rise, with NDVI and RI largely unaffected by water body proportions. Decreasing Hardscape proportions correspond to increased NDVI values. In synthesis, the optimal LEP area ratio (Softscape:Waterbody:Hardscape) for water-containing parks is 6.5:2:1.5, while for waterless parks, it is 8.3:1.7. This ratio, founded on observed trends from referenced data, can serve as both the baseline for carbon-neutral parks’ LEP area ratio and a guideline to ensure satisfactory NDVI and CN values by maintaining Softscape above 65% and Hardscape below 15%.

3.6. Results of Pearson Correlation Analysis

In this investigation, Pearson correlation analysis was carried out using IBM SPSS Statistics 26 software to establish relationships among Softscape, Hardscape, NDVI, and CN, as detailed in

Table 8. Pearson correlation analysis table.

	NDVI	CN	Softscape	Hardscape	RI
NDVI	1
CN	0.965 **	1
Softscape	0.994 **	0.967 **	1
Hardscape	−0.880 *	−0.903 *	−0.912 *	1
RI	0.473	0.443	0.407	−0.438	1

Note: **. p < 0.01; *. p < 0.05.

. This analysis aimed to validate the earlier conclusions. The results of the correlation analysis revealed the following relationships:

• NDVI and CN exhibited a strong positive correlation, with a coefficient of 0.965 and p < 0.01, implying that CN increases in tandem with NDVI growth.
• NDVI and Hardscape were inversely correlated, with a coefficient of −0.88 and p < 0.05, indicating that Hardscape proportion decreases as NDVI rises.
• NDVI and Softscape were highly correlated, with a coefficient of 0.994 and p < 0.05, revealing that Softscape proportion increases with NDVI enhancement.
• CN and Hardscape were inversely related, with a coefficient of −0.903 and p < 0.05, illustrating that Hardscape proportion negatively influences CN.
• CN and Softscape were positively correlated, with a coefficient of 0.967 and p < 0.01, signifying that CN is positively associated with Softscape.
• Hardscape and Softscape were inversely correlated, with a coefficient of −0.912 and p < 0.05, indicating that higher Softscape percentages lead to reduced Hardscape proportions.

3.7. Estimating the Carbon Neutrality Capacity of Additional Parks

Given the limited number of parks analyzed, a 5% extrapolation was conducted to assess the feasibility of attaining carbon neutrality. As data on parks’ carbon emissions was scarce, an estimation approach was adopted. This involved estimating 5% of existing parks’ carbon emissions and applying this estimation to other parks based on existing data. This approach aids in managing carbon emissions for parks under construction, potentially reducing emissions for new parks. Nonetheless, the estimation method introduces uncertainty, as it relies on park-specific data for accuracy. For parks with waterbodies, Lixin Park’s area is 4.7 hm² and its 5% carbon emission is 15,836.65 kgCO₂e, while Maple Garden’s area is 3 hm² and its 5% carbon emission is 14,941.3 kgCO₂e. The CN value of each park depends on park size, construction carbon emissions, and other factors. Also, for parks without waterbodies, Litchi Old Tree Park has an area of 0.54 hm² and a 5% carbon emission of 3623.4 kgCO₂e, whereas Dayuan Park has 0.6 hm² and 5172.4 kgCO₂e. To recap, Lixin Park and Maple Garden were selected as reference points for parks featuring water bodies, given their moderate sizes. Conversely, Yongquan Park was deemed unsuitable due to its substantial carbon emissions linked to its sizable water body. For parks without water bodies, Litchi Old Tree Park and Dayuan Park were employed. The data from Litchi Old Tree Park and Dayuan Park served as representative examples for such parks. This study encompassed the evaluation of LEP in 22 Taichung parks, combining green spaces and water bodies, to ascertain their potential for achieving carbon neutrality.

Table 9. Landscape element proportion of 10 parks with waterbody in the validation stage.

No.	Park Name	Park Area (hm2)	Landscape Element Proportion (%)
			Softscape		Waterbody	Hardscape
			Tree Cover	Green Space
1	Jiushe Park	4	59.17	18.29	1.00	21.54
2	Beitun Children’s Park	4.62	35.81	19.1	9.91	35.18
3	Dunhua Park	2.7	30.29	45.66	3.02	21.03
4	Fengle Sculpture Park	6	24.32	27.08	15.15	33.45
5	Nanxing Park	6.6	22.62	47.07	1.7	28.61
6	Taichung Park	10.5	37.9	19	9.95	33.15
7	Pinglin Forest Park	11.4	24	37.32	2.44	36.24
8	Buzi Park	5	18.78	38.6	3.12	39.5
9	Solanaceous Tree Park	0.6	65.34	11.65	7.84	15.17
10	Jingxian Park	1.03	67.49	14.97	0.93	16.61

shows ten parks with water bodies, ranging from 0.6 to 11.4 hm², while

Table 10. Landscape element proportion of 12 parks without waterbody in the validation stage.

No.	Park Name	Park Area (hm2)	Landscape Element Proportion (%)
			Softscape		Hardscape
			Tree Cover	Green Space
1	823 Artillery Battle Memorial Park	2.6	34.95	42.57	22.48
2	Sanfenpu Park	1.75	60.64	34.39	4.99
3	Renhe Park	0.65	8.14	69.6	22.26
4	Shiquan Park	0.6	29.98	31.93	38.09
5	Ruilian Children’s Park	0.6	23.23	66.24	10.53
6	wenhua park	0.65	43.46	31.02	25.52
7	Guangchang Park	0.64	56	33.86	10.14
8	Gangxing Park	0.64	32.65	33.66	33.69
9	Lecheng Park	2.9	51.57	38.74	9.69
10	Dongying Park	2.4	46.39	29.74	23.87
11	Junfu Park	1.4	40.17	31.17	28.66
12	Environmental Park	1.5	61.61	23.41	14.98

shows twelve parks without water bodies, ranging from 0.6 to 2.9 hm².

The obtained data reveal distinct Softscape, Waterbody, and Hardscape ratios among the parks studied. Among the parks with water bodies, Jingxian Park boasts the highest Softscape ratio (82.46%), while Fengle Sculpture Park records the lowest (51.4%). For Waterbody percentages, Fengle Sculpture Park leads (15.15%), while Jingxian Park has the lowest (0.93%). Regarding Hardscape ratios, Solanaceous Tree Park ranks lowest (15.17%), with Buzi Park having the highest (39.5%). The average ratios across these parks are approximately 6.6:0.6:2.8 (Softscape:Waterbody:Hardscape). In parks without water bodies, Sanfenpu Park exhibits the most substantial Softscape ratio (95.03%) and Shiquan Park the lowest (61.91%), with the highest Hardscape ratio at 38.09%. The Softscape to Hardscape average ratio here is 8:2.

For water body parks, an effective LEP area ratio is approximately 6.5:2:1.5 (Softscape:Waterbody:Hardscape), with Softscape above 65%, Waterbody exceeding 5.5%, and Hardscape kept under 28%. Waterless parks may maintain an LEP ratio of 8.3:1.7 (Softscape:Hardscape) for feasibility. However, the applicability of these findings might be constrained to the context of Taichung City, Taiwan. Parks with water bodies can leverage carbon sinks and aquatic life to achieve carbon neutrality. Specific recommendations include enhancing Softscape in Beitun Children’s Park, introducing vertical greenery and permeable paving, adapting design strategies for Taichung Park and Feng Le Sculpture Park to incorporate more trees and reduce high-carbon facilities, promoting tree planting in Shiquan Park, integrating permeable pavement in Gangxing Park, and considering greening options for Shiquan Park to counterbalance its Hardscape dominance.

4. Discussion

4.1. Interpretation of Findings

The findings of this study underscore a positive relationship between NDVI and CN, wherein higher NDVI corresponds to increased CN. Moreover, CN is positively linked to the proportion of Softscape and inversely correlated with the Hardscape ratio. The study demonstrates that a high NDVI reflects a substantial Softscape percentage and a lesser Hardscape proportion. For instance, Xindu Ecological Park features 78.96% Softscape, an NDVI of 0.62, and 15.19% Hardscape.

Notably, factors like fill, excavation, and carbon-intensive building materials significantly impact CN, with Softscape-rich parks having fewer structures and pavement, resulting in reduced fill, excavation, and material carbon emissions. The role of water bodies warrants separate consideration, as their impact on CN surpasses their impact on resilience due to their association with extensive fill and excavation. While water bodies offer flood mitigation, their construction contributes to CN increase.

The correlation analysis suggests that RI is not strongly linked to LEP, NDVI, and CN. This study advises increasing the Softscape-to-Hardscape ratio and incorporating permeable pavement to enhance CN. Although water bodies play a secondary role, balancing them with tree planting is vital for CN improvement. New parks face challenges in wind resistance due to young trees’ susceptibility and ornamental species’ lack of wind resilience.

The data demonstrates that parks with over 65% Softscape and under 15% Hardscape achieve carbon savings. Previous research aligns with these results, indicating high NDVI in smaller green parks and a negative correlation between NDVI and urban development [52,53,54]. Park size may also influence tree density [55,56]. Parks with irrigation and maintenance tend to have higher NDVI, showcasing well-maintained, shrub-covered parks’ higher NDVI values [52,57]. This study underscores the positive link between Softscape proportion and NDVI magnitude, echoing the idea that lush vegetation is associated with high NDVI. Plant biomass, a key carbon fixer, strongly correlates with NDVI.

According to the statistical results presented in Table 7, four parks have Softscape (Tree cover + Green space) exceeding 65% and Hardscape below 30%, and all of these parks demonstrate positive carbon reduction benefits. This indicates that larger Softscape areas and smaller proportions of built-up surfaces result in lower carbon emissions during park construction, thereby enhancing carbon reduction efficiency. The findings of this study align with previous research. For instance, Dronova et al.’s study on NDVI and local spatial clustering of sensor surface brightness temperature found that small green parks are characterized by high tree cover and high green cover (>69%) with higher NDVI values [58]. Bino et al. explained 84% of the variation in bird species richness through stepwise multiple regression, revealing a strong negative correlation between NDVI and the percentage of roads and built-up areas. Areas with higher paving and building densities tend to have fewer plants [58,59,60]. The findings of Dronova et al.’s study are closely aligned with the data in this study, showing that higher NDVI is associated with lower Hardscape proportions, consistent with the results of Bino et al. and Dronova et al.

According to the site survey and research findings in this study, Yongquan Park covers an area of 6 hectares with 572 trees and an NDVI of 0.21, while Lixin Park spans 4.7 hectares with 433 trees and an NDVI of 0.54. These results do not entirely align with previous research. It is generally assumed that larger parks would have more trees and biomass [61,62]. This study suggests that, aside from the size of the park and the number of trees planted, the density of tree planting may also play a role. Additionally, Bino et al. observed that urban parks tend to have higher NDVI values due to year-round maintenance and irrigation [60]. Consistent with this study’s findings, parks with higher NDVI values are regularly maintained and have more shrubs planted periodically. The results of this study show that higher NDVI correlates with a higher proportion of Softscape, in line with previous research findings. NDVI represents the greenness of a park; a high NDVI indicates a greater amount of vegetation. NDVI is closely related to plant biomass and the carbon fixed through photosynthesis [63,64,65,66].

4.2. Inspiration from Previous Research

Huang et al.’s research focuses on quantifying the carbon footprint of green spaces within parks, and they establish formulae for carbon storage in urban green areas. However, their formulae primarily involve basic engineering calculations and fail to encompass additional carbon footprint factors like demolition, repair, and renewal [19]. This study asserts that while the formulae in Huang et al.’s [33] work lack the comprehensiveness of Yang’s [65] more intricate and time-consuming calculations, they still offer a swift means to estimate a park’s carbon footprint, accounting for both pre-construction and post-construction phases. This attribute makes these formulae useful as a reference for quick estimations in the future.

Numerous scholars have investigated diverse aspects of carbon footprints. Some have computed the entire carbon footprint of construction projects, the carbon footprint associated with landscape upkeep and management, and even the carbon sequestration potential of tree biomass [29]. There exists a choice between calculating carbon sequestration in trees versus plants, sparking a debate on the preferable approach. This study contends that calculating tree biomass aims to precisely evaluate and quantify the mass of biomass stored in trees. Conversely, plant carbon sequestration estimates the carbon accumulated by plants within a specific timeframe. The choice between the two methods hinges on the research’s future objectives, possibly requiring an in-depth exploration of ecosystem dynamics within urban parks. However, it is acknowledged that both approaches face limitations. Tree data scarcity and a restricted sample size are noteworthy constraints identified in the literature review.

4.3. Purpose and Significance of the Study

The current study underscores the significance of assessing the total carbon footprint and resilience as crucial steps toward achieving carbon neutrality in parks. This aids in developing strategies to enhance parks that fall short of carbon neutrality. The study’s four research objectives encompass calculating the carbon footprint of six parks’ landscape engineering, evaluating park resilience, summarizing ideal LEP area ratios, and conducting correlation analysis of LEP, NDVI, and CN. The research successfully fulfilled all these objectives.

Past research on park carbon footprints revealed certain limitations. Strohbach et al. [53] addressed the carbon footprint of entire life cycles but omitted factors like irrigation practices and carbon emissions from certain park facilities. Karuchit et al. [23] focused solely on tree carbon sequestration, neglecting grasslands and other vegetation. Park et al. [42] employed generalized construction material carbon calculations, unlike this study, which localized data for accurate results. Furthermore, while previous studies explored NDVI, RI, and CN in urban parks, none investigated their interconnectedness through field research and data analysis. Thus, this study bridges a crucial research gap.

The study’s significance lies in several aspects: (1) Comprehensive life cycle carbon footprint calculations, encompassing labor, shrubs, and facilities, providing a more holistic assessment. (2) Inclusion of NDVI, RI, and CN, yielding insights into the impact and resilience of carbon-neutral urban parks. (3) Development of ideal LEP ratios tailored to different park types, offering valuable guidance for park construction, enhancement, and carbon neutrality achievement. (4) Provide landscape designers and managers with a tool to assess the carbon neutrality and resilience of parks swiftly and effortlessly, enabling rapid and effective decision-making prior to the design and improvement stages of park projects.

4.4. Research Conclusions

For parks with water bodies, the recommended LEP area ratios are 6.5:2:1.5, and for those without water, the ratios are 8.5:1.5. When the Softscape proportion surpasses 65% and Hardscape remains below 15%, parks with water can achieve satisfactory NDVI and CN values. Similarly, in waterless parks, maintaining a Softscape proportion above 83% and Hardscape under 17% is likely to result in elevated NDVI and CN values. The correlation analysis underscores that NDVI and CN are positively linked, indicating higher NDVI leads to greater CN. Furthermore, a higher NDVI corresponds to a lower Hardscape proportion and a higher Softscape proportion. Elevated CN is associated with reduced Hardscape proportion, while higher Softscape proportion corresponds to higher CN. Conversely, a higher Hardscape proportion is connected to a lower Softscape proportion.

In analyzing the carbon neutrality capabilities of six parks in Taichung City, this study developed a set of evaluation methods for LEP, NDVI, RI, and CN. This method highlights the importance of the interactions among these factors within urban parks and provides actionable recommendations for enhancing park carbon neutrality. However, several key considerations must be addressed to apply and replicate this research framework more broadly.

Firstly, the results of this study indicate a significant correlation between LEP, NDVI, RI, and CN, particularly in parks with higher green space proportions and lower Hardscape ratios. However, these findings are based on a limited sample size and a specific regional context. Therefore, when applying this method to other cities or regions, it is essential to account for differences in local climate conditions, vegetation types, and park management practices.

Secondly, water bodies may play a potentially significant role in park resilience and carbon neutrality. Although this study only included a few parks with water bodies, existing literature suggests that water bodies might make substantial contributions to carbon sequestration. When replicating this research framework, it is recommended to further explore the role of water bodies in various types of parks and consider the integration of facilities such as rainwater management systems. This will contribute to a more comprehensive understanding of the impact of water bodies on park carbon neutrality.

Lastly, this study did not thoroughly investigate the specific effects of different plant species on carbon neutrality efficiency. Future research could involve a detailed assessment of the biomass and carbon sequestration capacity of various tree and shrub species. As research progresses and expands, this study aims to develop more universally applicable methods to provide valuable insights for park planners and other professionals.

4.5. Research Constraints

This study’s scope is restricted to Taichung City, Taiwan. Due to the difficulty in obtaining accurate park project budgets, only six parks were included in the analysis. Future research could gather data from more parks for a more encompassing investigation. Furthermore, as parks were established at different times, comparing newer and older parks will be feasible with a broader dataset. In the calculation of park carbon footprint, this study focuses solely on plant types (tree canopy and shrub/grass coverage). For computational convenience, differences among species, plant cultivation, and related calculations are not considered. Another limitation lies in the seasonal assessment of NDVI, which was only conducted during spring. Data for other seasons is unavailable in this study. This study only included the carbon emissions associated with the construction phase of six newly constructed parks in Taichung City due to the unavailability of data on park maintenance and daily operations, which are typically documented separately. Consequently, the formula for “Landscape Engineering Total Carbon Footprint” includes components like CFrm (maintenance emissions) and CFeu (daily use emissions) that cannot be quantified. Future research will benefit from acquiring detailed data on maintenance and daily operations to provide a more comprehensive assessment of the parks’ carbon footprints.

4.6. Future Research Directions

As previously discussed, achieving carbon neutrality in parks demands a close interrelation among LEP, NDVI, and CN, along with the significance of RI. While this study included a limited number of parks with water bodies, such bodies play a crucial role in both RI and carbon neutrality. When a more extensive park sample is available, incorporating additional water bodies could offer valuable insights. Moreover, future research will concentrate on striking a balance between LEP, NDVI, RI, CN, and the aesthetic quality of parks.