Greening improvement scenarios for assessing the potential of urban carbon sinks and runoff reduction

Abstract

Two scenarios (SC1 and SC2) were developed using i-Tree Eco software to estimate carbon dioxide absorption (CO_2abs) by trees and avoided runoff (AR) in improved urban green areas (increasing the sizes of these areas and planting additional trees) in Thailand for a 50-year forecast (2024–2074). Tree annual mortality rates of 1 and 3% were assigned to SC1 and SC2, respectively. The results indicated the high potential of the country’s green areas as urban C sinks if proper greening improvements are implemented. CO_2abs in SC1 increased from 0.32 million metric tons-CO₂ equivalent (Mmt-CO₂e) in 2024 to 9.24 Mmt-CO₂e (a 2787.5% increase) in 2074 and is projected to increase beyond this year. In SC2, it increased from 0.32 Mmt-CO₂e in 2024 to 3.66 Mmt-CO₂e (a 1043.8% increase) in 2059 and is projected to gradually decline. The maximum AR (76.07 Mm³) was detected in SC1 at the end of the simulation period, but the peak AR (28.15 Mm³) in SC2 was detected in 2067, with a subsequent decline projected. Locally, CO_2abs increased from the observed amounts (mainly not more than 5000 mt-CO₂e) from 2019 to 2023 to more than 50,000 mt-CO₂e in 65 (84.4%) of 77 provinces in Thailand in 2065.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-06366-2.

Introduction

Cities are sources of carbon (C) emissions from various urban activities, residential and business (commercial and industrial) areas, vehicles, and transportation. Among dense buildings and other infrastructure components, green areas are appealing urban features as C sinks and other ecosystem service providers. In addition to C sequestration (C_seq) through the process of carbon dioxide absorption (CO_2abs), urban green areas (e.g., parks, street trees, landscape boulevards, greenways, and gardens) provide other important ecosystem services simultaneously, such as air pollutant removal, runoff reduction, heat island phenomenon mitigation, habitats for other living creatures, biodiversity aggregation, and human relaxation/recreation spaces¹. These ecosystem services increase the quality of urban living and the ability of city ecosystems to support life according to Sustainable Development Goal (SDG) 11 (sustainable cities and communities) and SDG 13 (climate action) of the United Nations².

Urban green areas are currently receiving much interest from researchers, environmental managers, and policy makers as nature-based solutions (NbS) to mitigate C emissions through CO_2abs. A large urban green area could be a potential C sink similar to a forest, e.g., the large and old Chapultepec Park in Mexico City³. However, unlike forest ecosystems, most green areas in a city are public manmade assets (e.g., public parks) with space limitations and specific functions for serving people⁴. These green areas are commonly scattered across locations, vary in size, and have different capacities as C sinks. Consequently, enhancing C sinks in a city is challenging in urban sustainability research and requires a considerable budget, proper policies/measures, resource mobilization, and public cooperation. Notable strategies to improve the urban C sink potential proposed in previous studies include urban growth control and green area protection⁵, forest city construction to optimize energy consumption and reduce C emissions⁶, and the establishment of an ecological network framework that couples the green space structure with the C sink function⁷.

Although Thailand has released less than 1% of the total amount of greenhouse gases (GHG) emitted worldwide, it ranks ninth among the countries at risk for climate change impacts. Recently, at the Conference of the Parties (COP 29) in Baku, Azerbaijan, on 11–12 November 2024, the Thai government committed to reducing more than 270 million metric tons-CO₂ equivalent (Mmt-CO₂e) of GHG emissions by 2035 and increasing the country’s GHG sink capacity to 120 Mmt-CO₂e by 2037. For the long-term commitment to GHG reduction, Thailand aimed to achieve carbon neutrality by 2050 and net-zero GHG emissions by 2065^8,9. To make these commitments possible, both policy and action measures for reducing GHG emissions at sources, particularly from five major sectors (energy, transport, waste, industrial process and product utilization, and agriculture) have been implemented in Thailand⁸. Simultaneously, enhancing the sinks of C, whose compounds (CO₂) contribute considerably to the GHG composition (76% CO₂, 19% CH₄, and 5% N₂O) in the country¹⁰, can be achieved through four major NbS systems: forest, wetland, agriculture, and urban green areas.

Among the four NbS systems, urban green areas in Thailand currently have the lowest C sink capacity because of their small size and tree planting constraints. Although expanding green areas is a straightforward method to increase urban C sinks, it is difficult in practice, particularly in crowded areas, where spaces are limited and allocated for multiple urban activities. According to the relevant measures proposed by the Thai government’s Office of Natural Resources and Environmental Policy and Planning (ONEP), the proportion of green areas should be at least 10% of urban and built-up (U) areas across Thailand for serving people, with 10 m² or more of the green areas per capita by 2027 [¹¹: p.12]. The six types of green areas defined by the ONEP [¹¹: pp.10–11] are as follows.

• Public green areas, encompassing all types of public parks, botanical gardens, and playgrounds;

• Utility green areas, such as the green areas of private sector businesses, institutions (e.g., governmental offices, schools, universities, and historical sites), and public utility areas (e.g., landfills, wastewater treatment plants, cemeteries, and airports);

• Strip-like green areas, such as roadsides, railway areas, riversides, irrigation canals, and the strips along public utility routes;

• Green areas associated with the community economy, such as green areas in or around farms, orchards, and aquaculture;

• Natural green areas, such as hills, wetlands, and riparian areas; and.

• Undeveloped and abandoned green areas.

However, the area of the greening sector (including all types of green areas above) across Thailand remains far below the baseline proportion (10% of U areas). Greening improvement (increasing green areas and planting additional trees) and urban tree mortality rates (which vary on the basis of biophysical variables, such as the species and age of trees, environmental conditions, maintenance, and pests) are two key variables that shape the urban C sink potential^1,4,12. Relevant studies indicated that efficient greening improvement would maximize not only urban C sinks but also other ecosystem services, such as air purification, floodwater regulation, and urban heath island mitigation^1,4,13–20. In Thailand, studies were conducted only at the local level, e.g., in Bangkok^1,14,19,20 and Chiang Mai¹⁸, whereas there have been no studies on the mortality rates of urban trees in the country. The first objective of this study was therefore to determine how the above two key variables influence CO_2abs (where CO_2abs = C_seq × CO₂ molecular mass/C atomic mass) in urban green areas across Thailand (at the national level) through two modeling scenarios (SC). The second objective was to evaluate the vital cobenefit of urban C sink improvement in terms of avoided runoff (AR), which could mitigate urban flooding in many at-risk provinces in the country, particularly in the central region, during the rainy season.

Materials and methods

Study area

Thailand is located slightly above the equator (between 5° 37′ N, 97° 22′ E and 20° 27′ N, 105° 37′ E) and covers an area of 513,115 km². The country comprises 76 provinces and Bangkok—the local special administrative area and the country’s capital. For the purposes of land use analysis and monitoring, the Land Development Department (LDD²¹) categorizes the country’s provinces into five regions: the northern (17 provinces), northeastern (20 provinces), central (18 provinces, including Bangkok), eastern (7 provinces), and southern (14 provinces) regions (Fig. 1). Owing to its geographic location and tropical monsoon climate, Thailand is a sunny country year round without frost. Southwest (from mid-May to mid-October) and northeast (from mid-October to mid-February) monsoon winds influence the regional climate across the three seasons. These include summer (February–April), the rainy season (May–October), and winter (November–January). The country’s annual mean temperature in 2024 was 28.5 °C. In the same year, the lowest temperature (8.1 °C) was detected in Tak Province in winter (January), whereas the highest temperature (44.2 °C) was detected in Lam Pang Province in summer (April). These provinces are located in northern Thailand. The country’s cumulative rainfall in 2024 was 1704.4 millimeters (mm) on average. In the same year, the highest cumulative rainfall (more than 5000 mm) was observed in the Khong Yai District of Trat Province (located in the eastern region), and the lowest cumulative amount (720.1 mm) was observed in the Mueang District of Tak Province²².

Fig. 1.

Thailand comprising 77 provinces (including Bangkok) in five regions, its urban and built-up areas observed from 2019–2023 by the Land Development Department²¹, and tree locations observed during the same five-year period by the Department of Climate Change and Environment²³. The maps were produced by the author using ArcGIS Pro version 3.3 (https://www.esrith.com/products/arcgis-pro/).

Urban land use and green areas

Of the five major land use groups (U areas, agriculture, forest, water bodies, and miscellaneous areas) classified by the LDD, the U areas from the latest monitoring data series (2019–2023) was 31,765.4 km², or only 6.19% of the total area of the country. The subgroups of U areas include municipal and commercial areas, developed lands, villages, government and private institutions, airports, train and bus stations, industrial areas, recreation areas, golf courses, and all other built-up areas²¹. Likewise, urban green areas across Thailand are apparently small, although they contain high tree species diversity. According to a combination of field observations and satellite image analyses conducted by the Department of Climate Change and Environment (DCCE) from 2019 to 2023, the estimated green area in Thailand was 127.91 km² (0.4% of the country’s U areas), with a CO_2abs capacity of 247,236.75 mt-CO₂e and containing 637 tree species²³. These included 422, 316, 584, 190, and 380 species observed in the northern, northeastern, central, eastern, and southern regions of Thailand, respectively (Table 1). The observed tree species are listed by region in Appendix A (Table A1). The U areas and tree locations were mapped (Fig. 1). The values of U areas, observed green areas, and observed numbers of tree species in the green areas are summarized by region in Table 1.

Table 1 .

Urban and built-up (U) areas, observed green areas (Obs.ga), numbers of tree species (Obs.spe), and CO_2abs in the obs.ga by region in Thailand during the 5-year period (2019–2023^21,23).

Region	U areas (km2)	Obs.ga (km2)	Obs.spe (no.)	CO2abs (mt-CO2e)
Northern	7045.9	35.35	422	22,183.79
Northeastern	10,336.3	49.19	316	112,629.65
Central	7544.0	23.03	584	59,956.16
Eastern	3189.3	10.29	190	18,078.58
Southern	3649.9	10.05	380	34,388.57
Total	31,765.4	127.91	637	247,236.75

Modeling scenario development

Two SC for forecasting tree CO_2abs in urban green areas across Thailand were developed according to two key variables: greening improvement and tree mortality rates. The modeling conditions were established as follows. First, only the DCCE observed trees²³ that had a diameter at breast height (DBH) of at least 1.3 m aboveground in the urban green areas were included in the initial dataset for simulating the two SC in each province. The trees in the bamboo and palm groups were excluded from the SC because of their low CO_2abs potential⁴. Second, the simulated green areas in each province were increased to 10% of the U areas per province for additional tree planting at a density of 1 tree/36 m², or approximately 44 trees/1600 m², which is appropriate for growing medium and large trees²⁴. Third, annual mortality rates (AMR) of 1% (SC1) and 3% (SC2) of the trees were assigned to the simulated green areas in each province. SC1 represented the best case, in which urban trees grew well without severe threats. For example, major public parks in Bangkok with good maintenance usually have tree mortality rates that do not exceed 1% annually¹. SC2 represented the worst case, in which urban trees received poor maintenance or faced severe threats, e.g., floods, drought, or pest outbreaks. The AMR of the trees assigned to SC1–SC2 in this study were within the same range as the observed AMR (1.5–2.9%) for most tree species in urban tree-planting programs in Florida, which has a climate similar to that in Thailand^1,25.

i-Tree Species software version 2.3.0²⁶ was used to select a tree species with at least 50% suitability for additional planting in the green areas of Thailand. The software computed the percentage suitability (0–100%) of a tree species for planting based on the local weather conditions in the planting areas and the functional abilities of the required trees. For this study, the top 10–50% of tree species for planting in each province obtained at least a score of 8 (10 was the best score) for (1) the removal of CO, NO₂, PM, and SO₂; (2) low emission of volatile organic compounds; (3) carbon storage; and (4) reduction in ultraviolet radiation. Seventy-three species of local trees that satisfied all of the above criteria were selected for additional planting in the northern (51 species), northeastern (40 species), central (68 species), eastern (33 species), and southern (49 species) regions of Thailand. The list of the suggested tree species is provided in Appendix A (Table A2), whereas the U areas, simulated green areas, and number of simulated species and relevant individuals of trees are listed by province in Appendix A (Table A3).

Each SC was simulated using i-Tree Eco software version 6.0.38²⁷ for a 50-year forecast starting from year 0 (2024) and ending in 2074 (the 50th year). The same number of trees, with an initial DBH of 2 cm for young trees, among the selected tree species were assigned, if possible, to achieve a density of 1 tree/36 m² via additional planting in the simulated green areas in each province. Owing to the country’s tropical climate without frost, the trees grew year round (i.e., 365 days) each forecasted year. To estimate CO_2abs by trees in urban green areas, i-Tree Eco computes the C_seq by trees from the estimated aboveground and belowground biomasses of the modeled tree species using relevant allometric equations²⁷. The software computes the AR of trees on the basis of the area of tree cover, the impervious cover beneath trees, including the trees’ root systems, and the weathering data for the simulated area or nearby proximity. The unobserved variables (tree cover area, impervious cover beneath trees, gas exchange, and particulate matter interception by trees) were estimated with the relevant equations and characteristic values of the input tree species in i-Tree Eco software (see details in^28,29).

i-Tree Eco software has been used worldwide for forecasting the growth and relevant ecosystem services of urban trees^{1,14,15,30–33}. In Thailand, Intasen et al.¹⁴ used i-Tree Eco version 5.0.6 to assess the urban forest in Bangkok and validated the forecasted results against those obtained from their field observations between January and May 2013. They concluded that the forecasting performance was good, but software version 5.0.6 had important limitations, e.g., it could not estimate the amount of stormwater runoff and the monetary values for CO₂ storage and sequestration, O₂ production, and PM_2.5 removal. However, these limitations of the early software versions were resolved. The latter versions of i-Tree Eco can estimate all the variables indicated by Intasen et al.¹⁴ and more. For example, Singkran¹ successfully used i-Tree Eco version 6 to evaluate three ecosystem services of 25 major parks in Bangkok, including C_seq, AR, and four air pollutant removals (CO, NO₂, PM₁₀, and PM_2.5), and estimate the monetary values of these variables.

Although some modeling variances might occur, such as physiological variations for the same tree species in different climatic regions^1,14,15, the software is considered highly reliable as long as the local input data are sufficient for modeling^30,31,34. In fact, modeling results obtained from a model developed using i-Tree Eco rely on the objectives of a study and data collection methods to answer certain research questions. Consequently, the variations in the modeling results, even in the same study area but different study periods, are likely the results of differences in the characteristics of the input data used to develop a model^31,35. These differences are due mainly to (1) the variability of the input tree-related variables (.e.g., species diversity, abundance, size, growth, and mortality) collected in different study periods and (2) differences in the study objectives that lead to different study designs (e.g., selection of tree species and quantity in the study and field data collection methods).

Results

CO_2abs and AR by country

For the entire country, the CO_2abs by both existing trees and additional planted trees in the increased green areas (10% U areas) considerably improved over the 50-year forecasts for both SC compared with the observed amount from 2019 to 2023 (0.25 Mmt-CO₂e²³) and the forecasted amounts in the initial year of simulation (year 0 or 2024). For SC1 (1% tree AMR), CO_2abs increased from 0.32 Mmt-CO₂e in 2024 to 9.24 Mmt-CO₂e (a 2787.5% increase) in 2074 (the 50th year of the simulation) and tended to increase beyond the last simulation year (Fig. 2a). SC2 (3% tree AMR) resulted in a lower amount of CO_2abs, with a peak before the simulation period ended. The forecasted CO_2abs in SC2 increased from 0.32 Mmt-CO₂e in 2024 to 3.66 Mmt-CO₂e (a 1043.8% increase) in 2059 (the 35th year of the simulation) and gradually declined thereafter (Fig. 2a). Similarly, the maximum AR (76.07 Mm³) was detected in SC1 at the end of the simulation year, but the peak AR (28.15 Mm³) in SC2 was detected in 2067 (the 43rd year) and subsequently declined (Fig. 2b). The total cobenefit value (of CO_2abs and AR) increased from 44.64 MUSD in 2024 to 1235.71 MUSD (a 2768.2% increase) in SC1 and 535.5 MUSD (a 1199.6% increase) in SC2 in 2065–the year aimed at achieving net-zero GHG emissions in Thailand. These estimated benefits were based on the current conversion rates for CO_2abs (130 USD/mt-CO₂e) and AR (2.36 USD/m³) of the i-Tree software team²⁷. For example, the total value in SC1 for northern Thailand in 2037 was 0.66 Mmt-CO₂e × 130 USD/mt-CO₂e + (1.50 Mm³ AR × 2.36 USD/m³) = 89.34 MUSD. The forecasted amounts of CO_2abs and AR and their estimated monetary values are summarized by region and period (2037, 2050, and 2065) of Thailand’s nationally determined contribution (NDC) in Table 2.

Fig. 2.

Carbon dioxide absorption (CO_2abs) in million metric ton-CO₂ equivalent (Mmt-CO₂e) unit observed in urban green areas from 2019–2023²³ and annual simulated scenarios (SC) from 2024–2074 after green area improvement with assigned tree annual mortality rates (AMR) of 1% (SC1) and 3% (SC2) across Thailand (a). The annual simulation of avoided runoff (AR) over the same period shown in (b); observed AR data were unavailable (na).

Table 2.

The simulated urban green area (Sim.ga), forecasts of CO₂ absorption (CO_2abs) and avoided runoff (AR), and value by region of each forecasted variable in the initial simulated year (2024) and three specific periods of thailand’s nationally determined contribution. The values of CO_2abs and AR were estimated at rates of 130 USD/mt-CO₂e and 2.36 USD/m³, respectively²⁷.

Forecast by region	Sim.ga (km2)	CO2abs (Mmt-CO2e)				AR (Mm3)				Value of CO2abs (MUSD)				Value of AR (MUSD)				Total value (MUSD)
		2024	2037	2050	2065	2024	2037	2050	2065	2024	2037	2050	2065	2024	2037	2050	2065	2024	2037	2050	2065
Northern	704.59	0.07				0.17				9.10				0.40				9.50
1% AMR			0.66	1.38	1.92		1.50	4.72	8.58		85.80	179.40	249.60		3.54	11.14	20.25		89.34	190.54	269.85
3% AMR			0.51	0.81	0.83		1.15	2.78	3.72		66.30	105.30	107.90		2.71	6.56	8.78		69.01	111.86	116.68
Northeastern	1033.63	0.10				0.43				13.00				1.01				14.01
1% AMR			0.90	1.89	2.73		3.71	11.60	21.61		117.00	245.70	354.90		8.76	27.38	51.00		125.76	273.08	405.90
3% AMR			0.69	1.11	1.18		2.84	6.82	9.36		89.70	144.30	153.40		6.70	16.10	22.09		96.40	160.40	175.49
Central	754.4	0.08				0.21				10.40				0.50				10.90
1% AMR			0.68	1.39	1.80		1.66	5.06	9.22		88.40	180.70	234.00		3.92	11.94	21.76		92.32	192.64	255.76
3% AMR			0.53	0.82	0.78		1.28	2.98	4.00		68.90	106.60	101.40		3.02	7.03	9.44		71.92	113.63	110.84
Eastern	318.93	0.03				0.22				3.90				0.52				4.42
1% AMR			0.29	0.62	0.92		2.10	6.76	12.69		37.70	80.60	119.60		4.96	15.95	29.95		42.66	96.55	149.55
3% AMR			0.22	0.36	0.40		1.61	3.98	5.49		28.60	46.80	52.00		3.80	9.39	12.96		32.40	56.19	64.96
Southern	364.99	0.04				0.26				5.20				0.61				5.81
1% AMR			0.34	0.71	0.96		2.25	7.05	12.65		44.20	92.30	124.80		5.31	16.64	29.85		49.51	108.94	154.65
3% AMR			0.26	0.42	0.42		1.73	4.15	5.48		33.80	54.60	54.60		4.08	9.79	12.93		37.88	64.39	67.53
Total	3176.54	0.32								41.60				3.04				44.64
1% AMR			2.87	5.99	8.33		11.22	35.19	64.75		373.10	778.70	1082.90		26.49	83.05	152.81		399.59	861.75	1235.71
3% AMR			2.21	3.52	3.61		8.61	20.71	28.05		287.30	457.60	469.30		20.31	48.87	66.20		307.61	506.47	535.50

CO_2abs and AR by region and NDC period

The forecasted amounts of CO_2abs and AR by region varied with the simulated sizes of urban green areas for additional tree planting; i.e., the larger the increase in green areas and the greater the number of trees planted were, the larger the values of CO_2abs and AR (Table 2). However, although CO_2abs and AR in every region increased from the estimated baseline values in 2024 (year 0 of the simulation), their rates obviously declined over time. In the three NDC periods, for SC1, the increased rates of CO_2abs ranged from 750–866.7% from 2024 to 2037 (Thailand aims to achieve a GHG sink capacity of 120 Mmt-CO₂e), 104.4–113.8% from 2037 to 2050 (Thailand aims to achieve carbon neutrality), and 29.5–48.4% from 2050 to 2065 (Thailand aims to achieve net-zero GHG emissions). The increase rates of AR were 690.5–854.5%, 204.8–221.9%, and 79.4–87.7% for the three consecutive periods. For SC2, the rates of increase in CO_2abs and AR were 562.5–633.3% and 509.5–631.8%, respectively, from 2024 to 2037, 54.7–63.6% and 132.8–147.2%, respectively, from 2037 to 2050, and  −4.9–11.1 and 32–37.9%, respectively, from 2050 to 2065.

Discussion

Increasing the size of urban green areas to 10% U areas and additional tree planting to achieve a density of 1 tree/36 m² resulted in a higher CO_2abs potential of the urban green areas in many provinces of Thailand. The forecasted CO_2abs increased from the observed amounts (mainly not more than 5000 mt-CO₂e) in the urban green areas across the country from 2019 to 2023²³ to more than 50,000 mt-CO₂e in 65 (84.4%) of 77 provinces in 2065 (Fig. 3). This finding revealed the high potential of the country’s green areas as urban C sinks if proper greening improvements^1,4 are implemented. The major constraint of urban greening improvement is the small size of U areas for green area allocation, particularly in provinces with dense urban activities or industrial layouts. In the best-case SC (SC1), 12 provinces still had a low potential for CO_2abs (and AR—the cobenefit considered in this study), not more than 50,000 mt-CO₂e, after reaching the 41st year of the simulation in 2065 (Fig. 3). Among these 12 provinces, Samut Songkhram, located in the central region, had the lowest CO_2abs potential of 14,950.8 mt-CO₂e in 2065, linked to its small size (416.7 km² in total and only 63.5 km² of U areas), followed by Ranong, located in the south, with U areas of 86.9 km² and a CO_2abs potential of 23,384.8 CO₂e in 2065. The remaining 10 provinces, with U areas in the range of 105–189.1 km², are projected to have CO_2abs potentials in the range of 28,651.3–49,768.32 mt-CO₂e in 2065. These provinces include Mae Hong Son in the north; Amnat Charoen in the northeast; Ang Thong and Sing Buri in the central region; Trat in the east; and Pattani, Phangnga, Phuket, Satun, and Yala in the south (Fig. 3; the locations and names of all the provinces are given in Fig. 1).

Fig. 3 .

Carbon dioxide absorption (CO_2abs) observed from 2019–2023²³ and the forecasted values in 2037, 2050, and 2065 from the best-case scenario of urban greening improvement, with the tree annual mortality rate of 1% in 77 provinces across Thailand (see the province names and regions in Fig. 1). The maps were produced by the author using ArcGIS Pro version 3.3 (https://www.esrith.com/products/arcgis-pro/).

In addition to the greening improvement and assigned tree mortality rates, the decreased rates of tree CO_2abs (and AR) over time for each SC (previously described in the results) reflect the effects of other uncontrolled biophysical variables, e.g., the ages and taxa of trees and local variations among the simulated green areas^1,12,30,31. Kim and Coseo³⁰ reported that density, species diversity, and the percentage of canopy cover of trees are important elements of urban parks to provide ecosystem services. Song et al.³¹ reported the different C_seq capacities and AR of trees among four types of green spaces (public park, protective green space, square green space, and attached green space) as the result of spatial variations in terms of size, health, and canopy coverage of trees observed in those green spaces. In this study, the rates of CO_2abs and AR by trees increased as the trees grew in size (based on DBH) and decreased after the trees passed maturity, as their DBH growth rates gradually decreased. Additionally, the high assigned tree AMR (e.g., 3% for SC2) caused more trees to die before reaching maturity and optimal CO_2abs levels. These findings were consistent with some results from previous studies (e.g^1,4,30,31). The consequence of high tree AMR was clearly detected in SC2, where the forecasted CO_2abs by trees decreased before the 50th year of the simulation period ended, whereas the forecasted values in SC1 (1% tree AMR) still gradually increased beyond the last simulation year (Fig. 2).

Considering regional differences, the maximum CO_2abs by urban trees in SC2 were detected in the 35th, 37th, 30th, 40th, and 33rd years in northern, northeastern, central, eastern, and southern Thailand, respectively, and decreased afterward. The synergetic interactions of the high assigned tree AMR (3% in this study) and the influences of land use, climate, and soil conditions, which varied among the five regions, contributed to the spatiotemporal variations in the urban tree CO_2abs among these regions. Previous studies have indicated that changing land use as a result of urbanization has negative impacts on tree C_seq capacity or C storage in urban ecosystems^36,37. Climatic variations (temperature and precipitation) and soil conditions also affect the potential C_seq of urban trees^38–40. In the case of SC2, to maintain the optimal CO_2abs potential of the urban trees over time, tree replanting (tree rotations) should be considered to replace the dead trees and the trees that passed their maturity^1,4. The tree rotations should be conducted in the year after the maximum CO_2abs by urban trees are attained in each region as indicated above.

Comparatively, the maximum AR in SC2 was detected before the last simulation year ended, but took longer to reach, as this variable was also governed by local climates, geographic locations, and soil conditions^40–43. The highest AR amounts were derived in the 43rd year in the northern and central regions, 45th year in the northeastern, 44th year in the eastern, and 42nd year in the southern regions. Although the climate across Thailand is characterized by a tropical monsoon, the rainfall patterns vary among the five regions. High rainfall commonly occurs year-round in the southern region, where local rainforest trees are abundant and have high rainfall interception capacity. Consequently, the maximum AR was detected earliest (in the 42nd year) in the southern region, followed by those in the northern and central regions. The terrain of the upper northern region is mountainous, and high-elevation areas lead to fast-flowing runoff downstream through low-lying areas and water sources to the central region. These conditions allowed the surviving trees located in the urban areas downstream in the northern and central regions to quickly absorb the urban runoff with their root systems⁴¹.

The trees located in the eastern region took more time to reach the maximum AR level (in the 44th year). Although the rainfall in this region is relatively abundant, it mainly rains in the coastal zone, where the urban tree density is low. The northeastern region has the lowest rainfall among the five regions causing the urban trees in this region to spend the longest time reaching the maximum AR level (in the 45th year). The variations in soil conditions among the five regions also govern the trees’ runoff interception potentials^42,43. However, owning to their complex characteristics (e.g., structure and compaction, texture, depth, organic matter, and infiltration rates), thorough studies are still rare to draw conclusions about the influences of soil conditions on tree runoff reduction in urban ecosystems.

With respect to the country’s NDC in 2037 (achieving a C sink capacity of 120 Mmt-CO₂e), the forecasted CO_2abs by trees in the improved green areas throughout the country in SC1was apparently small, at 2.88 Mmt-CO₂e (2.4% of the proposed C sink capacity). Consequently, the Thai government must rely on major C sinks, such as forest and wetland ecosystems, to fulfill its goal by 2037. Nevertheless, improvements in urban green areas are still needed not only for increasing the capacity of urban C sinks but also for providing of other ecosystem services, such as increasing the urban quality of living and biodiversity and reducing urban runoff, air pollutants, and urban heat island impacts^{1,15,18–20,30–33,35–38}. These cobenefits are valuable, but it is difficult to quantify some of them, such as aesthetics and biodiversity, in terms of monetary value for enhancing public perception¹. In this study, the estimated cobenefits of CO_2abs and AR in SC1 (1% tree AMR) in terms of monetary value were relatively high (Table 2), particularly the value of CO_2abs as its conversion rate increased from 51.27 to 130 USD/mt-CO₂e in 2025. This current rate was the average cost of one mt-CO₂e in 2020 (120 USD) and 2030 (140 USD) estimated by the US EPA at a discount rate of 2.5%²⁷.

The monetary benefit from AR alone for both SC was also impressive. Over the 50-year forecasting period, the estimated AR values across the country for SC1 and SC2 were 3.04 MUSD in 2024 (year 0 of the simulation) and increased to 26.49 and 20.31 MUSD in 2037, 83.05 and 48.87 USD in 2050, and 152.81 and 66.20 MUSD in 2065, respectively (Table 2). In the case of the megafloods in Thailand from 1995 to 2016 [⁴⁴: Table 2], the average damage cost was approximately 31.9 MUSD/event, excluding the severest damage in 2011, a level reached only once in the 50-year period. The estimated values of AR obtained in both SC in 2050 and afterward were higher than the mean damage costs associated with the historical flood events. This reflected the quantitative cobenefits of urban C sinks and might persuade the government and private sector to devote their budgets and work forces to urban greening improvement. Additionally, to facilitate the expansion of urban C sinks across the country, the Thai government had to update relevant laws/regulations to support green area improvement policies and activities. Raising people’s awareness of the benefits of urban green areas, such as C sinks and other ecosystem service providers, and offering incentives for greening improvement projects are indispensable for efficiently enhancing urban greening.

Conclusions

The results of this study verified that urban green areas have high potential for CO_2abs and runoff reduction if their size is sufficiently large for growing certain tree species and supporting certain planting densities. In addition to the optimal conditions for maximizing CO_2abs, the tree mortality rate is another key variable influencing the potential of urban C sinks. The synergistic consequences of the two key variables were revealed through SC1 and SC2 of urban greening improvement (increasing the size of urban green areas to 10% U areas and planting additional trees to achieve a density of 1 tree/36 m²), with assigned tree AMR of 1% in SC1 and 3% in SC2. That is, according to Thailand’s three NDC periods, the CO_2abs and AR values obtained in SC1 were 1.3, 1.7, and 2.3 times greater than those obtained in SC2 in 2037 (achieving a GHG sink capacity of 120 Mmt-CO₂e), 2050 (achieving carbon neutrality), and 2065 (achieving net-zero GHG emissions), respectively. With respect to the best-case SC (SC1), the improved green areas across the country with the tree AMR of 1% would have CO_2abs potentials of 2.87, 5.99, and 8.33 Mmt-CO₂e in 2037, 2050, and 2065, respectively, whereas the national AR potentials would be 11.22, 35.19, and 64.75 Mm³ in the three consecutive periods.

In the case of SC2 (3% tree AMR), tree rotations should be considered to replace the trees observed from 2019 to 2023 that reached maturity or died in each region before the 50th year (2074) of the simulation. This occurred for most trees in the northern, northeastern, central, eastern, and southern regions in the 35th (2059), 37th (2061), 30th (2054), 40th (2064), and 33rd (2057) years, respectively. These results reflected the gradual decline over time in the country’s total CO_2abs amount. Thus, the tree rotations should be conducted in the year after the trees in each region reach their maximum CO_2abs level as stated above. Overall, the country’s highest amount of CO_2abs across all five regions was detected in the 35th year (2059), declining thereafter. Thus, tree rotations in the case of SC2 may be conducted at the national level after 2059. In contrast, tree rotations under SC1 (1% tree AMR) may not be necessary before 2074 (the 50th year of the simulation) because the trees’ maximum CO_2abs level in each region was detected at the end of 2074 and tended to increase beyond this year. In addition to the strong influence of the highly assigned tree AMR, which caused more trees to die before reaching maturity and optimal CO_2abs and AR levels, some insightful findings were obtained from SC2. That is, the complex interplay among the inherent characteristics of trees and local variations in urban land use, climate, and soil conditions contributed to the spatiotemporal variations in attaining the maximum CO_2abs and AR in the green areas among the five regions.

Greening improvements in urban ecosystems, however, are difficult due to space limitations while supporting the diverse activities of urban dwellers. To increase urban green areas and public cooperation, the Thai government should implement mandatory policies (such as budget support and relevant law enforcement) and measures (such as public awareness, education, and incentives). Promoting both the environmental and economic benefits of urban green areas is another attractive way to gain support from various urban sectors. In this study, the forecasted results in SC1 revealed the enormous cobenefits of CO_2abs and AR in terms of monetary value over time, reaching 399.59, 861.75, and 1235.71 MUSD in 2037, 2050, and 2065, respectively. Both the environmental and economic advantages of urban green areas may help increase people’s awareness of and motivation for improving urban C sinks. Additionally, the Thai government should provide some incentives, such as tax exemptions/reductions to private sectors and people who dedicate their land for tree planting or donate land to the government’s urban greening improvement projects.

In this study, although i-Tree Eco software was successfully used to develop the two SC for assessing C_seq (to derive the estimated CO_2abs) by trees and AR in urban green areas across Thailand and the relevant monetary values, the forecasting results were based on tree-related data collected by the DCCE from 2019 to 2023. Using this modeling tool, differences in study design, field data collection, and study period can result in different modeling results in the same study area. Therefore, when the field observation-based software is used to develop a relevant model, the developed model should be periodically updated with the latest field collection data, when they are available, to maintain model reliability over time. Additionally, although the current version of i-Tree Eco software covers more features for assessing urban green areas, the software still has some limitations for evaluating abstract ecosystem services, e.g., aesthetics and other social values provided by urban trees/green areas, and converting them to monetary values.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Author contributions

Nuanchan Singkran designed and conducted the study, analyzed the data, wrote and revised the manuscript, and finalize all relevant manuscript submissions.

Funding

This work was part of the main research project supported by Thailand Science Research and Innovation (Contract No. ORG67OI001). Mahidol University supported the article publishing charge.

Data availability

The secondary data (the observed tree and land use data across Thailand) used in this study were obtained from the relevant Thai government agencies, thus they cannot provided by the author. However, some of the data further analyzed by the author are available in the online supplementary information (Appendix A).

Declarations

Competing interests

The author declares no competing interests.