City-wide greenhouse gas emissions of communities nearby the world heritage site of Ayutthaya, Thailand

Climate change has emerged one of the greatest threats to sustainable development. Cities are a major contributor to high carbon dioxide levels. This research aimed to quantify city-wide GHG emissions and investigate the potential for climate change mitigation in communities near the World Heritage Site (WHS) of Ayutthaya, Thailand via the multi-criteria analytical hierarchy process (AHP). The total city-wide GHG emission of Ayutthaya Municipality in 2018 was 99,137.04 tCO2eq (1.93 tCO2eq per capita). Energy and waste sectors were the two largest emitters. Pratuchai, the most populated subdistrict and the WHS location, was the largest source of GHGs. However, the cultural heritage site emitted only 0.2% of total GHGs. Based on the IPCC2013 LCA method, residential sector accounted for the largest share (74%), while the WHS contributed only < 1% of total energy-related CO2 emissions. If all the Thailand’s Nationally Determined Contribution (NDC) Roadmap are fully implemented in the residential sector, total GHGs would be reduced by 9735.47% tCO2eq and 6846.86 tCO2eq in 2030. Based on expert interviews, AHP pairwise comparison showed that energy-saving strategies were more preferable than renewable energy technologies. For climate policy initiative, ‘feasibility of implementation’ had the highest AHP weight (0.45) followed by ‘policy feasibility’ (0.39), and ‘environmental performance’ (0.16).


Scientific Reports
| (2022) 12:9787 | https://doi.org/10.1038/s41598-022-14036-w www.nature.com/scientificreports/ could be more effective than emission intensity where multi-faceted factors are systematically considered 6 . From a policy viewpoint, the concept of low carbon society is probably a major opportunity to achieve climate resilience and promote a more sustainable development. For instance, a study conducted by 7 addressed the low carbon initiative framework in Thailand as a showcase of Southeast Asian emerging economies because Thailand is considered one of the brightest newly industrialized economies, with vibrant energy-and carbon-emission-intensive growth. Therefore, according to the United Nations Framework Convention on Climate Change (UNFCCC), Thailand's Intended Nationally Determined Contribution (NDC) aims to reduce GHG emissions by about 20% from the projected business-as-usual (BAU) level by 2030. In summary, it is clearly apparent that cities must be well-managed, more resilient, and reduce their carbon footprint 2 . Despite its growing importance, however, there are only a few systematic studies in Thailand that focus on estimation of GHG emissions at the city level. City-scale GHG inventory and relevant activity data, particularly in cultural heritage areas, are lacking. Many studies have put more emphasis on emission inventories for specific years at the national scale and sectoral assessment of GHG emissions. Further, most relevant case studies are of provincial capital cities. Therefore, the objectives of this study were to estimate city-scale GHG emission and highlight some mitigation potentials in the WHS of Ayutthaya, Thailand based on a multi-criteria AHP approach. The ultimate expected outcome of this study is to help city policy makers gain a thorough understanding of the important role cities can play in limiting global climate change.

Methodology
Case study. Based on Thailand's national economy and the National Economic and Social Development Board report, the five provinces with the highest income (GPP per capita) are Rayong, Chonburi, Bangkok, Phra Nakhorn Si Ayuthaya, and Chachoengsao. This study focuses on the fourth richest city (with a GPP of 14,381.97 USD per capita), Phra Nakhorn Si Ayutthaya, which is an ancient capital in the central region of Thailand. Phra Nakhorn Si Ayutthaya Municipality, which is located approximately 70 km north of Bangkok and has an area of about 14.8 km 2 , was chosen for this research case study. The geomorphological structure of Phra Nakhorn Si Ayutthaya, brought about by the development of towns in the past, has made it into an island surrounded by three rivers-Chao Phraya, Pa sak, and Lop Buri. Demographically, Phra Nakhorn Si Ayutthaya Municipality has a population exceeding 50,000 with approximately 20,220 households (in 2019). It comprises 10 sub-districts: Pratuchai (6344 households), Hua Ro (4312 households), Ho Rattanachai (4185 households), Tha Wasukee (3193 households), Kamung (659 households), Klong Suan Phlu (457 households), Ban Kao (448 households), Huntra (415 households), Klong Sa Bua (116 households), and Khao Rain (91 households). The total area can be divided into the following six land-use types: settlement, governmental area, waterbodies, road/railway, agricultural area, and others. The highest proportion of land (60.5%) is devoted for settlement. As depicted in Fig. 1, by using ArcGIS Pro 2.8.0, approximately 40% (289 ha) of the total island area is protected by the United Nations Educational, Scientific and Cultural Organization (UNESCO) World Heritage Site (WHS), with several cultural heritage aspects including temples, archaeological sites, cultural landscapes, museums, and historical landmarks 8, 9 . In 2018, Phra Nakhorn Si Ayutthaya hosted nearly 6.1 million domestic and international travelers.

Sources and boundaries of city-level GHG emissions. The Global Protocol for Community-Scale
Greenhouse Gas Emission Inventories (GPC) 10 was applied to estimate city-level GHG emissions in this research. Basically, the GPC allows meaningful benchmarking and greater consistency in GHG accounting 11 . By categorizing sources of emissions, the GPC defines three scopes for city inventories. Scope 1 covers GHG emissions from sources within the city boundary, Scope 2 covers GHG emissions attributable to grid-supplied energy, heat, steam and/or cooling within the city boundary, and Scope 3 refers to all other emissions that occur outside the geographical boundary of the city (i.e., out-of-boundary waste and wastewater transportation). The GPC classifies overall GHG emissions into five main sectors: (i) Stationary energy, (ii) Transportation, (iii) Waste, (iv) Industrial Process and Product Use (IPPU), and v) Agriculture, Forestry, and Other Land Use (AFOLU) ( Table 1). The geographic boundary of Ayutthaya Municipality served as the boundary for GHG inventory at the city level. All GHG emissions within scopes 1-3 of the GPC were reported (details of scope of GHG emissions are available in supplementary information, Table S1). In terms of source sector, emissions from stationary energy, transportation, AFOLU, and waste in 2018 were reported and expressed as CO 2 equivalent (CO 2 eq). Meanwhile, emissions from IPPU were excluded as this source does not exist within the case city.
Estimation of GHG emissions. Energy. Stationary energy emissions are primarily due to fuel consumption for energy use in residential, commercial, and institutional buildings and facilities (commercial and public buildings such as schools, hospitals, government offices, temples, highway street lighting, and other public facilities) in the city. Specifically, emissions from fuel combustion (Scope 1) and grid-supplied energy consumed (Scope 2) within the city boundary were included. To explore the electricity dataset, real consumption data were obtained from the Provincial Electricity Authority (PEA). A scaled-down method using population indicator 10 was employed to represent city-wide energy consumption. Equation 1 can be used to estimate CO 2 emissions from energy use in this sector.
where energy used is total consumption of electricity (kilowatt-hours; kWh) and each type (i, j) of fuel (liter) in the city, and EF is Emissions Factor of each type of energy (i, j) (Table S2) www.nature.com/scientificreports/ In addition to the application of the GPC, life cycle impact assessment (LCA) was employed to quantify energy-related GHG emissions using a functional unit of '1 year of electricity consumption' 12 in the Ayutthaya Municipality. Using SimaPro 8.3.0, the IPCC 2013 GWP 100a method was followed to evaluate climate change impact as CO 2 eq. Transport. Ayutthaya Municipality does not have railways, waterways or air transport. Therefore, only 'onroad' mode was considered and represented as in-boundary transportation in this research. Following a bottomup approach, the fuel sold method was employed to estimate GHG emissions from transportation within the city. As shown in Eq. (2), total emissions were quantified via multiplication of quantity of fuel sold by countryspecific EFs for each fuel type 13 . Emission factor specific to each type of pollutant was used for estimating transport emission of CH 4 , CO 2 , and N 2 O. Total GHGs was expressed as CO 2 eq.  where Fuel j,k = Fuel sales volume or consumption of fuel (j) for transport type (k); EFi,j = Emission factor for compound (i) emitted from fuel (j).
AFOLU. GHG emissions from AFOLU within the city boundary were estimated based on the IPCC Tier 1 method. In terms of in-boundary emissions, only paddy field farming was accounted for in Ayutthaya Municipality. Methane emissions from rice cultivation was quantified using Eqs. 3 and 6, data on annual harvested area, water regimes during cultivation period, crop cultivation period, and soil amendments obtained from the Office of Agricultural Economics (OAE) and Ayutthaya Agricultural Extension Office (DOAE). As noted, GHG emissions from on-road transportation to and from the locations of agriculture and forestry activities were accounted under the transportation sector.
where CH 4Rice is annual methane emissions from rice cultivation (Gg CH 4 /yr), EF i is adjusted daily emission factor for a particular harvested area (kg CH 4 /ha/day), A is annual harvested area of rice (ha/yr); t is cultivation period of rice (day). In this study, rice cultivation period was 180 days.
where EF c is baseline emission factor for continuously flooded fields without organic amendment (1.30 kg CH 4 / ha/day); SF w is a scaling factor to account for the differences in water regime during the cultivation period (SF w = 0.27 for rain-fed area); SF p is a scaling factor to account for the differences in water regime in the pre-season before the cultivation period. SF p value of 0.68 was used for non-flooded pre-season of more than 180 days. SF o is a scaling factor that should vary for both type and amount of organic amendment applied (Eq. 3). SF s,r is a scaling factor for soil type, rice cultivar, and so on. Due to non-availability of soil type and rice cultivar data, SF s,r was not considered in this study.
where ROA i is the application rate of organic amendment i, in dry weight for straw and fresh weight for others (tonne ha −1 ); CFOA i is the conversion factor for organic amendment. During the land preparation stage, rice straw was incorporated into soil via ploughing and tilling (> 30 days) before cultivation. CFOAi value of 0.29 was chosen in this research. Total amount of rice straw was estimated from average rice yields and straw to grain ratio (SGR) (Eq. 6), which varied with season, location, and cutting height. An average SGR ratio of 0.75 was chosen.
Waste. In this study, CH 4 emissions associated with waste management activities were defined under Scope 3 because all waste generated in the municipality area were treated at a landfill site outside the city boundary. The IPCC default method was used to estimate methane emissions from solid waste disposal. Specifically, using Eq. (7), methane emissions were estimated based mainly on mass of solid waste sent to landfill annually. All the various types of municipal solid waste collected by municipalities were included in the estimation of degradable organic carbon (DOC) or the organic carbon which is accessible to biochemical decomposition (Eq. 8).
where MSW T is mass of solid waste sent to landfill in inventory year; MSW F is fraction of solid waste disposed to disposal sites; MCF is methane correction factor (fraction); DOC is degradable organic carbon in year of deposition (fraction); DOC F is fraction of DOC that is degraded. Tabasaran's theoretical equation of DOC F = 0.014 T + 0.28, where T is temperature is used to determine the value. F is Fraction of methane in landfill gas (IPCC default is 0.5); 16/12 is Stoichiometric ratio between methane and carbon; R is recovered methane; OX is oxidation factor where DOC is degradable organic carbon, A is fraction of municipal solid waste that is paper and textiles, B is fraction of municipal solid waste that is garden waste or other non-food organic putrescibles, C is fraction of municipal solid waste that is food waste, D is fraction of municipal solid waste that is wood or straw.
As previously mentioned, the total emissions of city-wide greenhouse gases of communities nearby the World Heritage Site of Ayutthaya, Thailand were reported in the unit of carbon dioxide-equivalents (CO 2 eq) which is derived by multiplying the total emissions of GHGs by the associated Global Warming Potential (GWP) (i.e., GWP values for 100-year time horizon of IPCC AR4 is 25). www.nature.com/scientificreports/ GHG mitigation options. GHG mitigation scenarios. In this research, emissions projections were evaluated under the following two scenarios: (i) business-as-usual (BAU) and (ii) Nationally Determined Contributions-(NDCs-) mitigation plan. Hypothesizing that the current economic trends of the city would continue without technological or structural change, the baseline (BAU) scenario was forecasted from 2018 to 2030. More specifically, mitigation scenarios indicated in Thailand's NDC were forecasted to the target year (2030) and further compared to the BAU level (2018). Provincial and municipal economic growth rates of 4.13% and 1.12%, respectively, were proposed for forecasting GHG emissions in the case city using Eqs. 9 and 10.
where GPP per capita,n = Gross Provincial Product, n = Total number of years where GPP per capita,n = Gross Provincial Product per capita (year n), Population n = Total population of municipality (year n).
Multi-criteria assessment of climate change mitigation policy. Analytical Hierarchy Process (AHP) was used to prioritize climate change mitigation policy and explore challenges associated with implementing GHG mitigation policy in Ayutthaya Municipality based on multiple criteria analysis. Climate policy analysis employing the AHP involves two steps: Step 1: Selection of climate policy by focusing on mitigation measures indicated in the Thailand's NDC. Further, the following criteria modified from 14 were used to assess the preferences for initiating climate policy in a case study: Environmental performance (i.e., direct contributions to environmental or climate benefits), Political feasibility (i.e., attitude of stakeholders towards cost efficiency and policy possibility/stringency for noncompliance), and Feasibility of implementation (i.e., feasibility of climate mitigation measures implementation and financial feasibility).
Step 2: Weighting mitigation measure. AHP-pairwise comparison method was employed by interviewing the following experts and decision makers in Ayutthaya Municipality (n = 4): (i) Representative of Ayutthaya Provincial Office for Natural Resources and Environment, Provincial Energy Office, Provincial Electricity Authority, and Ayutthaya Municipality. Theoretically, as presented in Eq. (12), the weighting of each GHG mitigation measure and preference for climate policy implementation (from step 1) was conducted following pairwise comparison matrix with 9-point scale (i.e., less important variables were valued from 1 to 1/9) 15 . All pairwise comparison numerical values were normalized and summed to 1. The consistency ratio (CR) was computed to avoid any incidental judgment in the pairwise comparison matrix (Eqs. 12 and 13). The calculated weighting coefficients are acceptable if the final CR is less than 0.1.
where A = [a ij ] is a representation of the expert's preference for each GHG mitigation measure (i, j = 1, 2, …, n) where λ max is the greatest eigenvalue of the pairwise comparison matrix; n is the factor number where CI is the consistency index; RI is the random consistency index (i.e., RI for 9 factors is 1.25).

Results and discussion
Total city-wide emissions. In 2018, the total city-wide GHG emissions of Ayutthaya Municipality were approximately 99,137.04 tCO 2 eq. Energy sector was by far the biggest contributor to the total emissions (49%; 48,216.54 tCO 2 eq), followed by the waste sector (36%; 35,659 tCO 2 eq), transportation (11%; 11,191.75 tCO 2 eq), and AFOLU (4%, 4,069.75 tCO 2 eq) (Fig. 2a). Overall, estimated GHG intensity was 1.93 tCO 2 eq per capita. Similarly, a study conducted by the Thailand Greenhouse Gas Management Organization 16 reported that the annual per capita GHG emissions of some cities (i.e., at the municipal level) in Thailand ranged from 1.28 to 3.74. It should be noted that municipal-level GHG emissions were slightly lower than in denser cities and metro areas. Under such situations, per capita emissions seemed to decline with decline in urban density. For instance, residents in the capital city of Thailand, Bangkok, were responsible for emitting 7.01 tCO 2 eq per capita in 2016 16 . In developed cities like London and New York, per capita emissions were 6.18 (in 2006) and 5.8 (in 2014) tCO 2 eq, respectively 17 . Shan et al. 18 conducted an inventory of CO 2 emissions of Chinese cities in 2010 and revealed that Hohhot and Nanping generated the highest (29.67 tCO 2 eq per capita) and lowest CO 2 emissions per capita (2.38 tCO 2 eq per capita), respectively. On the basis of scope, in this study, indirection emissions from consumption of (9) GPP = GDP current yr GDP first yr www.nature.com/scientificreports/ purchased electricity (Scope 2) in the Ayutthaya Municipality represented the largest source of emissions (38%), followed by Scope 3 and Scope 1 (Fig. 2b).
Energy. Energy sector was by far the greatest contributor to the total GHG emissions (49%) in Ayutthaya Municipality. Within this, the residential sector was the primary contributor to GHG emissions (37,672.89 tCO 2 eq; 78%), followed by commercial and governmental organizations (18.49%), and public lighting (i.e., highway street lighting, 3.38%) ( Table 2). Among commercial and governmental organizations, hospitals, schools, and hotels located in the municipality released large amounts of GHGs (812.60-4218.84 tCO 2 eq). Compared to other sub-districts, Pratuchai, which is the most populated subdistrict and where the WHS is located, was the largest source of GHG emissions (7797.10 tCO 2 eq). Interestingly, however, the WHS emitted only 206.38 tCO 2 eq (while temples and cultural heritage site emitted 151.63 and 54.74 tCO 2 eq, respectively), accounting for approximately 0.2% of the total city-wide emissions. The results of energy-related GHG emissions obtained using the GPC were similar to those obtained via the LCA approach. Emission results generated using the IPCC2013 GWP 100a method also revealed that the residential sector accounted for a significant share of the total energyrelated CO 2 emissions (74%; 31,866.18 tCO 2 eq). Commercial and governmental organizations emitted almost a quarter (22%; 10,695.76 tCO 2 eq) of this total. Meanwhile, public lighting and the WHS were attributable for only 4% and < 1% of total emissions from energy consumption. Relatedly, the WHS was found to account for the lowest GHG emissions using both the GPC and LCA methods. It is possible that most of the ancient buildings at the WHS of Ayutthaya municipality are located outdoors. Many of these historic heritage sites rely mainly on natural light and ventilation for thermal and lighting comfort during the day. In terms of electricity consumption, only lamps are commonly used for lighting, especially at night. However, improving sustainability and energy efficiency in built historic heritage have become high-interest topics among scholars. For instance, European countries with colder climates are most interested in reducing energy consumption in historic buildings 19 . The European Directives showed the potential of the building sector in achieving energy efficiency and reducing carbon emissions. Specifically, study conducted by Serraino and Lucchi 20 investigated energy retrofit intervention in a castle in Italy and reported that the following criteria must be considered in the implementation of a highly-energy efficient system in historic buildings: achieving building thermal comfort and minimizing CO 2 emissions and running costs. Some studies have focused on natural lighting design and solar radiation control in heritage buildings 21,22 . A study of Martínez-Molina et al. 19 also observed that improving indoor climate, energy efficiency, and thermal comfort in historic buildings have become a major issue. Further, Marchi et al. 11 assessed environmental policies for GHG emission reduction at  www.nature.com/scientificreports/ UNESCO heritage sites of Italy using integrated energy-saving measures and revealed that installation of solar panels on roofs of existing buildings was the most effective environmental policy for decarbonization. GHG emissions equivalent to 17,000 tCO 2 eq per year could be avoided using this measure (57% reduction in GHG emissions in the short term; 10 years).
Transportation. In 2018, the volume of GHGs emitted by on-road transport sector of the Ayutthaya Municipality amounted to 11,191.75 tCO 2 eq (11%). The combustion of diesel fuel generated the largest share of emissions (39%) compared to gasohol 91 (37%) and gasohol 95 (24%) for road transport, respectively. It has been suggested that the main factors affecting commuting-related CO 2 emissions of individuals their and households in local context of Thailand should be more investigated. This finding is in line with Marchi et al. 11 who reported that diesel is most widely used in the transport sector. By reducing transportation emissions, a study of Li et al. 23 suggested that many behavioral targeting policies should be implemented in mobility management in cities to reduce private motorized transport that could contribute to a reduction in CO 2 emissions.
AFOLU. Methane emissions from AFOLU within the city boundary were estimated by employing Eqs. 3 and 6. In this research, methane emissions from rice fields were estimated based on emission factor for a particular harvested area of rice, cultivation period of rice (day), and also annual harvested area of rice. In terms of land use, only 2% of the total area of Ayutthaya Municipality (i.e., Khao Rain, Klong Suan Phlu, and Huntra subdistricts) served as agricultural area. All the rice growing areas in the municipality is under rain-fed cultivation.
The period of rice cultivation in this study area was about 180 days. The results found that the AFOLU sector of Ayutthaya Municipality emitted approximately 4069.75 tCO 2 eq (accounting for 4% of city-wide emissions). This is within the range of estimated GHG emissions from rice cultivation in other municipalities of Thailand (1.13-6106.75 tCO 2 eq) 16 . In terms of alternatives for mitigating emissions related to rice cultivation water management, Alternate Wetting and Drying (AWD) is a possible option to mitigate methane emission from irrigated rice paddies. A research conducted by Oo et al. 24 confirmed that yield-scaled GHG emissions from AWD cultivation were significantly lower than from continuous flooding cultivation.
Waste. The GPC report 10 highlighted that city should report GHG emissions from disposal of waste generated within the city boundary, whether treated inside or outside the boundary of city. In this study, the total solid waste sent to landfill in 2018 was 18,130 tonnes, comprising 95.4% of non-food organic waste, 3.3.4% of food waste, and 1.43% of solid waste that is paper and textiles, respectively. The default values of methane correction factor (MCF) of 1 and oxidation factor (OX) of 0.1 for managed landfill were used in estimation of methane emission. In 2018, Ayutthaya Municipality generated 18,130.19 tons of municipal solid waste, all of which was sent to a landfill located outside the city boundary. Within Scope 3 of the GPC, waste sector was the second largest source of GHG emissions, accounting for 35,658.61 tCO 2 eq (36% of the total emission). This result is consistent with other reports in literature; for example, Hoklis and Sharp 25 reported that in 2009, 338.51 GgCO 2 eq of methane was emitted from municipal solid waste in Phnom Penh, Cambodia. It was clearly observed in this study that waste disposal activities in tourism cities may result in direct and indirect GHG emissions. Further, previous studies reported that in 2018, tourist areas of Thailand generated approximately 1-2.5 kg of waste per person per day 26,27 . Therefore, behavior of both resident and tourist plays an important role in mitigating climate change, as it determines the solid waste generation rate of the community.

GHG mitigation scenarios and AHP assessment of climate change mitigation policy.
Under the BAU scenario, total GHG emissions of Ayutthaya Municipality would increase from 99,137.04 tCO 2 eq in 2018 to 161,123.67 tCO 2 eq and 113,316.63 tCO 2 eq in 2030 in case of forecasted annual economic growth rates of 4.13% and 1.12%, respectively. The residential sector was attributable for the largest proportion of GHG emissions (78%). Under BAU condition, estimated GHG emissions from the residential sector were about 37,672.89 tCO 2 eq. This is projected to increase to 43,056.66 tCO 2 eq and 61,221.79 tCO 2 eq in 2030 at annual economic growth rates of 4.13% and 1.12%, respectively. Potential GHG mitigation options were proposed based primarily on Thailand's NDC. If all policy interventions as indicated in Thailand's NDC Roadmap on Climate Change Mitigation (2021-2030) are "fully" implemented in the residential sector, total GHG emissions would be reduced by 9735.47% tCO 2 eq and 6846.86 tCO 2 eq in 2030 assuming 4.13% and 1.12% annual economic growth rates, respectively. These policy interventions include (i) energy-saving measures and strategies and (ii) applying renewable energy in residential buildings. As shown in Fig. 3a,b, the forecast results revealed that applying all energy-saving strategies could provide greater GHG reduction than the implementation of renewable energy policy in the local case study. According to the judgment on pairwise comparisons of AHP, Tables 3 and 4 revealed that high-efficiency cooling systems should be excluded in the AHP. Installation of LED lighting showed the highest score, followed by improving energy efficiency of air conditioners, and energy efficient appliances and cooking stoves. Establishing renewable energy technologies presented the lowest score in the AHP comparison. In the case of factors that affect climate mitigation policy decisions and implementation, feasibility of implementing mitigation measures (i.e., technical possibilities) presented the highest score (0.45), followed by policy feasibility (0.39) (i.e., policy possibility and cost effectiveness of implementing climate strategies). Surprisingly, environmental performance (i.e., benefits of policy on environment and climate change) showed the lowest score (0.16) in terms of stakeholder's preference for tackling climate change in the case study. These results are consistent with Heinrich et al. 14 who revealed that feasibility of implementation was the most important criteria for developing climate change mitigation measures in the power sector. Direct contribution to GHG mitigation as climate benefit was one of the highest AHP criteria, which is inconsistent with this research. www.nature.com/scientificreports/ Overall, as clearly pointed out above, energy consumption in the residential sector was attributable for the largest share of GHG emissions in this research case study. The following recommendations were proposed: Aligning city policies for a low-carbon society: It is strongly suggested that each city in Thailand quantify citywide CO 2 emission inventory and their reduction targets. The country's NDC targets must be linked to citylevel policy. Moreover, policy decision makers, local authorities, and all stakeholders should urgently prioritize climate change action policy and mitigation strategies, which had the lowest AHP score, in their local area. Specific detailed information on city-level CO 2 emissions would be much more meaningful to the decisionmaking process of local authorities than aggregated information at the national level 28 . However, interestingly, a study conducted by Gouldson et al. 29 revealed that the absence of multi-level governance arrangements that potentially enable the implementation of low carbon development strategies at the urban level in Asian cities  www.nature.com/scientificreports/ will have global implications for climate change. Moreover, there are still some challenges associated with creating city-level GHG emission inventories such as the use of different methods and approaches and the difficulty of defining city boundary and cross-boundary activities. Further, activities data at the city level are limited and incomparable. It is therefore necessary that data collection for GHG inventory at the city level be as accurate as possible and all uncertainty associated with inventories be avoided.
Creating green tourism culture in the WHS. Ayutthaya, a UNESCO WHS, is one of the most famous landmarks and a major tourist attraction of Thailand. The importance of tourism development and GHG emission is now being emphasized by academic scholars. A report by the UNWTO-UNEP-WMP 1 (2008) revealed that CO 2 emissions generated by tourism was attributable for approximately 3.9 to 6% of the total global emissions in 2015. Similarly, a report by Katircioglu et al. 30 found that tourism directly affected both energy consumption and carbon emissions in the long-term economy of Cyprus. In China, research conducted by Meng et al. 31 assessed carbon emissions of the tourism industry using the Tourism Satellite Account (TSA) and the input-output model and found that the Chinese tourism industry accounted for 2.425-2.489% of the total CO 2 emissions of all industries in China from 2002 to 2010. Overall, transportation activities accounted for about two-thirds of the total direct emissions, followed by accommodation and food services, and shopping. It should be highlighted that specific information on carbon emissions induced by tourism, especially at the city level in Thailand, are missing. This study strongly suggests that further assessments of carbon emissions from tourism be performed in future research. Further, the impacts of tourism activities and GHG emissions associated with solid waste generation should be urgently explored. Moreover, human and social sustainability of all related environmental and climate change mitigation project should be systematically investigated 32 .
Limitations: It should be noted that there are some limitations associated with this research. For instance, lack of access to data or low reliability of existing data for estimating GHG emissions (i.e., N 2 O inventory) in the AFOLU sector, leading to higher uncertainty associated with these emissions. Specifically, Due to the constraints of this study (i.e., lack of usage statistics for fertilizers in AFOLU sector at the local scale), therefore, only methane emissions were quantified. In the waste sector, lack of updated data on composition of solid waste might have a significant effect on the operational GHG emissions results. Therefore, the critical importance of the life cycle approach and uncertainty/sensitivity analysis should be further considered and accounted in estimating GHG emissions at the city scale.

Conclusion
Cities are major contributors to the world's energy-related carbon emissions. However, GHG emission inventories at the city level in Thailand are not well studied. Therefore, the locality of Ayutthaya (a WHS), Ayutthaya Municipality, was selected as a case study. City-wide GHG estimates obtained using the GPC revealed that the most considerable portion of emissions originated from household electricity consumption, especially in Pratuchai, which is the most densely populated subdistrict and the area where the WHS is located. Meanwhile, based on the LCA approach, the WHS contributed only < 1% of the total energy-related CO 2 emissions. Based on expert interviews and multi-criteria analysis, AHP pairwise comparison revealed that energy-saving strategies was the more preferrable climate mitigation policy compared to renewable energy technologies. Feasibility of implementation and policy feasibility were the most important factors influencing climate mitigation policy in Ayutthaya Municipality. Meanwhile, environmental performance (i.e., climate change co-benefits) was the least preferred factor. Further studies on practical policies for low-carbon development and tourism consumer behavior in the WHS should be urgently investigated.

Data availability
The datasets generated and analyzed during the current study are available from the corresponding author on reasonable request.