How do actions to decarbonise the energy and mobility sectors affect consumption-based carbon footprints? A case of historic and predicted actions in a suburb in Finland

Household consumption accounts for 72% of the global greenhouse gas (GHG) emissions. To obtain consumption-based emissions in accordance with the 1.5-degree target, the carbon footprint of individuals should be reduced to 2.5 t CO2e a−1 by 2030, which means at least a 70% reduction in high-emitting countries. The decrease in consumption-based GHG emissions can be achieved through both technological and behavioural changes. Currently, climate measures are largely dependent on technological efficiency, although behavioural changes are also needed. In this paper, we study how technological actions to decarbonise the energy and mobility sectors affect consumption-based carbon footprints in the years 2010–2030 in a suburb in Finland. Based on the results, emissions from the mobility sector decreased by 50%, and those from the energy sector decreased by 68% in 2010–2030, when only technology development and society-level actions are considered. These emission reductions affected the carbon footprint of an average Finnish suburb by decreasing it by 37%. This study demonstrates that technological actions to decarbonise the energy and mobility sectors do not guarantee adequate emission reduction by 2030 to achieve the 1.5-degree target; therefore, a change in lifestyle and consumption habits is also needed. Further research should take behavioural changes into account when assessing the development of a consumption-based carbon footprint.


Introduction
Overconsumption and its links to planetary boundaries (i.e. climate change and biodiversity) cause significant uncertainty to humankind now and in the future (Rockström et al 2009;Steffen et al 2015). To limit global warming to 1.5°C, global greenhouse gas (GHG) emissions per capita must be reduced to 2.5 t CO 2 e a −1 by 2030 and to 0.7 t CO 2 e a −1 by 2050 (Akenji et al 2021). According to Rogeli et al (2018), emissions should be reduced to zero by the middle of this century. However, the responsibility of mitigating climate change should be fairly shared between countries because global emissions are highly concentrated in 10% of households, with the highest per capita emissions contributing to 34%-45% of global GHG emissions and the bottom 50% contributing to 13%-15% (IPCC 2022). Citizens from North America and the European Union are among the top 10% emitters (Chancel and Piketty 2015).
Traditionally, climate change mitigation actions have focused on production-based or territorial GHG emission accounting. These methods allocate GHG emissions according to the place of origin and exclude international imports and exports, whereas consumption-based emission accounting considers GHG emissions caused by consuming goods and services and does not depend on where emissions are produced (Ottelin et al 2019; Davis and Caldeira 2010). In some developed countries, although production-based GHG emissions have decreased, consumption-based GHG emissions have increased (Peter and Hertwich 2008). For example, in Finland and Sweden, a decrease in GHG emissions in the energy sector has been achieved as a result of technological development, but the growth of household consumption takes away the benefits of low-emission technology (Salo and Lettenmeier 2022). Therefore, an adequate reduction in GHG emissions cannot be achieved only through technological innovations. Behavioural changes are also needed.
The climate change mitigation potential of consumers has recently been studied. Household carbon footprints can be decreased with already available low-emission consumption options (Ivanova et al 2020), and behavioural change in households has been recognised as an inexpensive and rapid way to achieve climate change mitigation goals (Stankuniene et al 2020). Van de Ven et al (2018) found that carbon footprint per capita could be decreased by 6%-16% through behavioural change in the European Union. In Finland, Salo and Nissinen (2017) estimated that the carbon footprint of an average Finn could decrease by approximately 37% through consumption choices. Even though households have great potential to mitigate climate change, there are many barriers to changing household behaviour to be more sustainable, for example, the socioeconomic characteristics of the household, access to credit, knowledge and understanding of climate change and the political environment (Stankuniene et al 2020). In addition, government and education institutions focus recommendations on lower-impact actions instead of high-impact climate actions, creating a misunderstanding between official recommendations and individual climate actions (Wynes and Nicholas 2017). Evidence shows that people underestimate GHG emissions from highest-intensity emission actions, such as air travel and meat consumption (Wynes et al 2020). Therefore, mitigation responsibility cannot be given only to households. The climate mitigation potential of societal actions and technological development should also be assessed. Currently, national and international climate targets rely mostly on technological innovations instead of behavioural change (O'Brien 2018).
Our study provides information on how technological and societal decarbonisation actions affect consumption-based GHG emissions in suburban areas. In general, consumption-based emission accounting studies mainly focus on international, national or city scales (Ottelin et al 2019). Knowledge is still lacking about consumption-based emissions accounting on a sub-city scale, for example, a suburban scale. Therefore, this work focuses on consumption-based accounting in a suburban area. According to Gill and Moeller (2018), GHG emissions per capita decrease when municipality size increases. The lowest emissions are found in municipalities with 100,000-500,000 residents. However, emissions have started to increase again in the category, including municipalities with more than 500,000 residents.
This research was conducted in the city of Lahti in Finland, specifically in the neighbourhood of Mukkula. Lahti is located in southern Finland with 120,000 residents. Finland is a high-income country where 66% of consumption-based GHG emissions are linked to households (Nissinen and Savolainen 2019), and the consumption-based carbon footprint of Finns is approximately 9.7 t CO 2 e a −1 (Akenji et al 2021), which is one of the highest carbon footprints in the world. In 2021, consumption-based GHG emissions were calculated for Lahti for the first time in terms of mobility, food, energy and construction sectors and consumer goods and services. Previously, only territorial-based GHG emissions were calculated, but these emissions covered only a part of the carbon footprint of citizens. The current consumption-based carbon footprint in Lahti is 8.4 t CO 2 e a −1 per capita (Sitowise 2022), which, according to the study, is 1.3 tCO 2 e lower than the average carbon footprint of a Finn.
The consumption-based carbon footprint of Finnish households increased by 12% during the years 2000-2016 (Nissinen and Savolainen 2019). Various studies have indicated that income level affects carbon footprint (Christis et al 2019;Feng et al 2021;Steen-Olsen et al 2016;Ivanova et al 2017). In Finland, the annual carbon footprint is 7.2 t CO 2 e in the lowest income decile and 19 t CO 2 e in the highest income decile (Nissinen and Savolainen 2019). Kalaniemi et al (2020) calculated the consumption-based carbon footprints of people participating in a universal basic income (UBI) experiment. As the calculations were based on minimum reference budgets, they were considered carbon footprints of the lowest income decile. According to their calculations, the average carbon footprint at the UBI level was 4.8 t CO 2 e a −1 . In the study, the UBI average was 10,100 €/capita. We assume that the carbon footprint of Mukkula residents is lower than the average Finnish carbon footprint due to socio-economic characteristics; thus, the carbon footprint of Mukkula is assumed to be closer to the sustainable level.
This study aims to examine how consumption-based GHG emissions can be decreased through societal action in a Finnish suburban area. We focus on the energy and mobility sectors and study how much technological development and societal decarbonisation actions only in these sectors can decrease the total carbon footprint of residents in a suburban area. In this paper, behavioural changes are not included in calculations. Nissinen et al (2015) examined the effect of policy instruments in decreasing the climate impacts of housing, mobility and food sectors in Finland and found that the GHG emissions of household consumption decreased with policy instruments by 6% of Finland's average GHG emissions for the years 2008-2012 and that the most effective reduction potential was in housing and transport policies. Therefore, only the energy and mobility sectors are considered in the current study. To avoid generalisation, this study is limited to a local scale and determined for a specific suburban area. The study is based on the actual consumption estimates of residents and the projections of societal and technological changes in the energy and mobility sectors in a specific Finnish suburban area for the years 2010, 2021 and 2030. Our findings are important in the suburban context because GHG emissions from urban areas are substantial and expected to increase (Lwasa et al 2022).

Materials and methods
The analysis of this study is based on two key approaches. Person-specific information from people living in the case suburb of Mukkula was collected using a survey. Information related to GHG emissions was collected from the literature, databases and local operators to assess consumption-based carbon footprints based on the survey data. This study used a quantitative approach. GHG emission-related data for mobility and energy systems were collected and analysed for the years 2010, 2021 and 2030. This research focused especially on the energy and mobility sectors because societal actions and policy instruments to decrease consumption-based GHG emissions are more efficient in the mobility and energy sectors than in the food and other consumption sectors (Nissinen et al 2012).

Study area
The neighbourhood of Mukkula is located approximately 5 km from the Lahti city centre and has a population of 7,676 (in 2020), comprising 47% men and 53% women. The Mukkula area was selected for this study because of its socio-economic status and housing stock. The average income level in Mukkula is lower than the average Finnish income level and more than 70% the housing stock is apartment buildings. The majority of the population is middle aged or retired, and the average age is 46 years (Finnish average 43 years). In the area, the majority of residential buildings are 1960s apartment buildings, and the household size is 1.8 on average. The average income level in Mukkula (35,027 €/year/household) is lower than the average Finnish income level (41,651 €/year/household) (Statistics Finland 2022a).
In recent years, the city of Lahti has made significant environmental sustainability actions, which have affected carbon footprints of citizens. In 2021, Lahti was nominated in the European Green Capital Award for being a pioneer in environmental actions and developing innovative solutions to environmental challenges. In 2019, Lahti reduced its GHG emissions from energy production by abandoning coal and replacing it with a low emission-intensive bioheat powerplant (European Union 2020). In the mobility sector, Lahti provides electric city bikes, and two-thirds of the city's public transport (i.e. bus) runs on electric or renewable biodiesel (Green Lahti 2022). In addition, Lahti has engaged residents towards a more sustainable lifestyle, for example, by conducting an experiment on personal carbon trading on mobility (Kuokkanen et al 2020;Uusitalo et al 2022).

Survey
The survey of the Mukkula residents' initial data for consumption-based carbon footprints was conducted online between August and September 2021 in Finnish. The functionality of the survey was tested on 13 test users before it was sent to the respondents. A link to the digital survey was sent to a random sample of 2,000 Mukkula residents via post letters. To ensure the representativeness of the sample, the random sample was provided by the Finnish Digital and Population Data Services Agency, and it was limited to residents who lived in the postal code area of 15240, who were aged 18-80 and whose mother language was Finnish. The total number of respondents to the survey was 264, and the response rate was 13%. Consent of the participants was obtained for the survey. The data were collected using the Webropol 3.0 survey and reporting tool and analysed using the SPSS 26 statistical programme. The survey questions were related to the background, housing and mobility information of the respondents. Most of the questions were related to estimations of different consumption volumes. A copy of the questionnaire translated into English is presented in appendix A.
The reliability of the survey was analysed by comparing the socio-economic characteristics of the sample (a number of acceptable survey participants) and those of the population of Mukkula (table 1). Information on the Mukkula inhabitants was acquired from Statistics Finland's online database Paavo, which provides open data through postal code areas (Statistics Finland 2022a). Despite the relatively low response rate, the sample was sufficiently representative of the population of Mukkula (table 1). However, there were a few differences between the sample and the population of Mukkula. For example, those who were over 60 years old, highly educated and in the highest income category were overrepresented in the sample (table 1).
In addition to socio-economic characteristics, the representativeness of the sample was analysed by the confidence interval. Confidence interval for GHG emissions from mobility and energy sectors per respondent was calculated using the 95% confidence level. The average GHG emissions from mobility and energy sectors is 3641 kg CO 2 e a −1 per respondent with a 95% confidence interval [3249, 4033].

Calculation of consumption-based carbon footprint
Carbon footprint is defined as the sum of GHG emissions and removals in a product system based on a life cycle assessment (LCA) using climate change as a single impact category. Different GHGs (e.g. CO 2 , CH 4 and N 2 O) are presented as mass in CO 2 equivalents using the global warming potential (GWP) factors of GHGs (ISO 14067). According to the ISO 14067 (2018), ISO 14040 (2006) andISO 14044 (2006) standards, carbon footprint calculations should follow the four phases of LCA: 1) goal and scope definition, 2) inventory analysis, 3) impact assessment and 4) interpretation.
In this chapter, the goal, scope definition and impact assessment are presented for the carbon footprint calculations of housing and mobility. The inventory analysis of housing and the mobility carbon footprint calculations are presented in Chapters 2.3 and 2.4. An impact assessment includes the key equations needed for carbon footprint calculations, which are presented in Chapters 3.2, 3.3 and 3.4. This interpretation is presented in chapter 3.

Mobility
The consumption-based carbon footprint of mobility of the respondents (kgCO 2 eq person −1 a −1 ) depends on personal mobility behaviour (modal shares and distances) and modal-specific GHG emissions (passenger car, bus, train, ferry, airplane, motorcycle/moped, cycle and walking). Carbon footprints include direct emissions from the use phase of each vehicle and emissions from vehicle manufacturing and fuel or energy production.
In the survey, respondent-specific mobility data were explored by asking how many kilometres the respondents travelled with different mobility modes during an average week. Based on information for an average week, the annual emissions were calculated based on the fact that one year has 52 weeks. The following equation was used to calculate the carbon footprint of mobility: where CF mobility is the consumption-based carbon footprint of mobility [kgCO 2 eq a −1 ], ef c is the life cycle emission factor of a passenger car (gasoline, diesel, electric or biofuels) [kgCO 2 eq km −1 ], b c is the number of kilometres driven by a private car per year [km a −1 ], i p is the average number of passenger in a car, ef v is the GHG emissions of vehicles per passenger kilometres (bus, train, bike, motorbike or moped) [kgCO 2 eq pkm −1 ], b v is the number of kilometres travelled by each vehicle per year [km a −1 ], and ef a is the GHG emissions per trip (flight and ferry) [kgCO 2 eq trip −1 ]. Flights were categorised into domestic, intra-European Union (EU) and (1) N = 264, the number of respondents of the survey.
(2) N = 260, the number of respondents of the survey.
(4) N = 6379, total number of residents of Mukkula aged 18 or over (Statistics Finland 2020 intercontinental flights. Ferry trips were divided into trips between Helsinki and Tallinn and between Helsinki and Stockholm, which are the most likely ferry trips for people living in Lahti. In the formula, n a is the number of one-way trips per year.

Housing
The consumption-based carbon footprint of housing for each respondent (kgCO 2 eq person −1 a −1 ) includes direct GHG emissions from heat and electricity production. It can be calculated using questions related to the type of building (apartment house, detached house, terraced house and semi-detached house), the area of the apartment in square metres, the construction year of the house, the main heating system in the household and the type of electricity contract of the household. It can only be calculated when the energy and electricity consumption of the household are determined. The following equation was used to calculate the carbon footprint of housing: where CF housing is the consumption-based carbon footprint of housing [kgCO 2 eq], a h is the area of a household [m 2 ], c h is the energy class of the building [kWh m 2-1 ], ef h is the emission factor for heat production [kg CO 2 eq kWh −1 ], i h is the number of inhabitants in the household, e e is the electricity consumption of the household [kWh], and ef e is the emission factor for electricity production [kg CO 2 eq kWh −1 ].
Heat consumption is calculated based on the construction year of the house and the area of the apartment in square metres. The energy class of the building (c h ) depends on the construction year. If the construction year is 2010 or later, the factor 130 kWh m 2-1 is used; when the construction year is between 1990 and 2009, the factor 160 kWh m 2 -1 is used; and when the construction year is earlier than 1990, the factor 240 kWh m 2-1 is used (Hietaniemi et al 2021). The carbon footprint of heat consumption depends on the energy consumption, the heat production method used (electricity, oil, district heating, wood/pellet, heat pump) and the number of inhabitants in the household.
The electricity consumption of the household depends on the type of house and the number of inhabitants in the household. The electricity consumption [kWh a −1 ] formula for apartment buildings is 1,400 + X * 500 (X = the number of inhabitants in the household − 1), that for detached and semidetached houses is 4,600 + X * 900 and that for terraced houses is 2,600 + X * 700 (Hietaniemi et al 2021).

Historic, current and future predictions of mobility emissions: 2010, 2021 and 2030
For the assessment, we included the life cycle GHG emissions of various mobility modes, such as vehicle manufacturing, fuel or energy production and direct emissions. In some studies, infrastructure construction and the end-of-life emissions of vehicles were included in the assessments. However, as these were assumed to have a minor effect on total emissions, they were excluded from the current study (Chester and Horvath 2009). All emission factors for the different mobility modes and for the years 2010, 2021 and 2030 are presented in appendix B.

General assumptions for multiple mobility modes
Fossil diesel and petrol production from oil is assumed to produce 19 gCO 2 MJ −1 (Directive 2018/2001). There is a variation between biofuels' carbon footprints, but the average is assumed to be approximately 19 gCO 2 MJ −1 from the production phase, especially when biofuels are made from waste and sideflow feedstocks (Directive 2018/2001; Seppälä et al 2019). Renewable hydrotreated biodiesel sold in Finland is estimated to lead to a 90% reduction in total GHG emissions compared to fossil diesel use (Neste 2021). Biofuel distribution obligations with road fossil fuels are 4% in 2010 and 18% in 2021 (Sipilä et al 2020). The proportion of biofuel will increase to 30% by 2029 (Jääskeläinen 2021). Therefore, we assume a 30% biofuel share in 2030.

Local buses
Public transportation within the city of Lahti is operated by buses. All buses in 2010 operate using fossil diesel, and all of them are assumed to be EURO5 buses. Among all buses in 2021, 17% run on electricity, 41% on biodiesel, and 42% on fossil diesel. All diesel buses in 2021 are EURO6 buses. A target for 2030 is that all buses must operate using either electricity or biodiesel. For this study, a 50-50 assumption is made between electricity and diesel. The direct emission per passenger in street driving is 58 gCO 2 km −1 for EURO 5 diesel buses and 53 gCO 2 km −1 for EURO 6 diesel buses. Fuel consumption per passenger is 0.80 MJ km −1 for both bus types, and electricity consumption is 0.078 kWh km −1 for electric buses (Technical Research Centre of Finland 2017). According to Chester and Horvath (2009), bus manufacturing emissions are 8.0 gCO 2 pkm −1 . Based on these assumed values, the local bus-related emissions per passenger kilometre are as follows: 78.9 gCO 2 pkm −1 (2010), 36.2 gCO 2 pkm −1 (2021) and 23.4 gCO 2 pkm −1 (2030).
2.4.3. Passenger cars including petrol, diesel, gas, ethanol, electric and hybrid cars All passenger cars in 2010 are assumed to use petrol or diesel as their fuel, as the use of other fuels was minimal in Finland, and electric and hybrid cars were not yet available during that time. The distribution of passenger cars in 2021 is based on the survey. Finland's target for electric passenger cars is 700,000 in 2030. Of these cars, at least half are battery electric cars, and the rest are hybrids (Jääskeläinen 2021.) The total number of passenger cars is predicted to be 2.85 million (LVM 2021b). Therefore, the share of electric passenger cars will be 24.6%, and the share of battery electric passenger cars will be 12.3% in 2030. The share of electric cars is assumed to be the same in Mukkula as in Finland. Finland's target for gas cars and vans is 130,000 in 2030 (Jääskeläinen 2021). In 2021, approximately 15,600 cars or vans use gas as fuel in Finland, and 93% of them are passenger cars (Traficom 2022a). In this study, the share is assumed to be the same in 2030, that is, 120,900 gas cars. Therefore, the target is that 4.2% of cars will use gas in 2030. Table 2 shows the distribution of passenger cars for the years 2010, 2021 and 2030.
Direct GHG emissions and fuel consumption of passenger cars with internal combustion engine are extrapolated from historic emission data provided by the Technical Research Centre of Finland (2017). According to the survey, the passenger cars of the respondents are nine years old on average, which is less than the average of 12 years in Finland (Statista 2021). A number of diesel and petrol cars used in Finland are also assumed. Passenger cars operating with high biofuel blends (biodiesel, ethanol and biogas) are assumed to have only biogenic direct emissions. Battery electric cars are assumed to consume 0.17 kWh km −1 of electricity (Uusitalo 2020). Hybrid passenger cars in Finland operate with 50% electricity and 50% fuel on average, but this is highly dependent on users' behaviour (Autoalan Tiedotuskeskus 2020). According to Uusitalo et al (2020), the manufacturing emissions of internal combustion engine-operated passenger cars are 17.1 gCO 2 pkm −1 , and electric car manufacturing emissions are 28.6 gCO 2 pkm −1 . The average of the manufacturing emission factors of battery electric and internal combustion engine-operated passenger cars is used for hybrid passenger cars due to a lack of specific data. For the same reason, how passenger car manufacturing emissions will change in 2030 has not been analysed. The average GHG emissions for passenger cars are 214.6 gCO 2 km −1 (2010), 145.9 gCO 2 km −1 (2021) and 82.3 gCO 2 km −1 (2030).

Motorcycles/Mopeds
Motorcycles and mopeds use liquid petrol or diesel as fuel. The GHG emissions for this category of mobility are calculated with the assumption of a 50-50 ratio between mopeds and motorcycles. Direct emissions and fuel consumption are 68 gCO 2 km −1 and 0.87 MJ km −1 for mopeds and 109 gCO 2 km −1 and 1.5 MJ km −1 for motorcycles (EURO 4), respectively (Technical Research Centre of Finland 2017). There have been no improvements in GHG emissions or in fuel consumption between EURO 2 and EURO 4 motorcycles, and it is assumed that there will not be significant reductions in the future (Technical Research Centre of Finland 2017). Therefore, the same data are applied for 2010, 2021 and 2030, and any differences are caused by the different biofuel shares in the fuel blend. Unfortunately, no information is available on moped or motorcycle manufacturing emissions; therefore, they are not included in this study. The GHG emissions for this category are 105.7 gCO 2 km −1 (2010), 95.9 gCO 2 km −1 (2021) and 82.7 gCO 2 km −1 (2030).

Train
The main energy source of passenger trains in Finland is electricity. Hydropower has been the electricity used by the national railway company in Finland since 2008 (VR Group 2022). The GHG emissions from hydropower are assumed to be 2 gCO 2 eq kWh −1 (Vattenfall). Approximately 5% of passenger trains use diesel because not all routes have electricity. However, the train company has compensated for these emissions since 2019, and it is assumed that the emissions will also be compensated for in 2030 (V R Group 2022). For 2010, diesel usage in trains is approximately 1.4 gCO 2 eq pkm −1 (Uusitalo 2020). The energy consumption for trains using electricity is assumed to be 0.3 MJ pkm −1 , which is the average value of four different trains in 2016 in Finland (Technical Research Centre of Finland 2017). GHG emissions from the manufacturing of trains is 3 gCO 2 pkm −1  (Chester and Horvath 2009). Based on these assumed values, the GHG emissions from train transport are 4.6 gCO 2 eq pkm −1 for 2010 and 3.2 gCO 2 eq pkm −1 for 2021 and 2030.

Ferry
This study includes two possible ferry trips: 72 km (Helsinki-Tallinn) and 383 km (Helsinki-Stockholm). The emission factor in 2016 for ferry travelling at a speed of 27 knots is approximately 282 gCO 2 eq pkm −1 with 3.6 MJ pkm −1 fuel consumption (Technical Research Centre of Finland 2017). The emissions are assumed to be the same in 2010 and 2021 because the renewal of ferries is slow. The European Parliament has voted that maritime transport will be included in the EU Emissions Trading System from 2022 onwards and that CO 2 emissions will be reduced by 40% by 2030 (Erbach 2020). Unfortunately, no information is available on ferry manufacturing emissions; therefore, they are not included in this study. The total GHG emissions for Helsinki- Tallinn  According to Graver et al (2020), regional flights emit 160 gCO 2 per passenger-km, medium-haul flights emit 110 gCO 2 per passenger-km, and long-haul flights emit 90 gCO 2 per passenger-km. These emission factors are used as the 2021 factors for aviation.
According to Finland's Government Programme, the target for the ratio of renewables in aviation fuels will be 30% by 2030 (LVM 2021a). Therefore, it is assumed that 30% of the fuel is sustainable aviation fuel (SAF). Direct emissions from the SAF are assumed to be only biogenic. Using HEFA technology for producing SAF produces 9.7% emissions on average compared to fossil jet fuel when considering conversion, transport and use (Neuling and Kaltschmitt 2018). The historical annual 1.9% reduction in passenger-km emissions is assumed to continue (Larsson et al 2019).

Cycling and walking
Cycling and walking do not cause direct GHG emissions, nor do they consume energy. Nevertheless, we include GHG emissions related to bicycle manufacturing, which are approximately 5.0 gCO 2 eq km −1 (European Cyclist Federation 2021).

Historic, current and future predictions of energy emissions: 2010, 2021 and 2030
To assess the development of GHG emissions in the energy sector, we examine GHG emissions from electricity and heating production. Only direct emissions are considered; that is, emissions from energy infrastructure are not considered. The GHG emissions from the construction of the respondents' apartments are included in the assessment. The emission data for energy calculations are presented in appendix C.

Electricity production
Annual electricity production-related GHG emissions are obtained from Statistics Finland (2020). There are two ways to allocate emissions for electricity from combined heat and power production: the benefit allocation method and the energy method. The benefit allocation method causes slightly higher emissions than the energy allocation method. It was selected for this study because it allocates emissions more evenly between power and heat. The emission factor for electricity production is 270 gCO 2 kWh −1 in 2010. The newest value available is 89 gCO 2 kWh −1 for 2020.
In Finland, electricity can be marketed as renewable energy only if it has been issued a guarantee of origin by the transmission operator system Fingrid (Energiavirasto 2022a). As this share is sold to customers separately, the average emission factor of the Finnish electricity grid mix is lower than emissions from fossil-based energy sources. Therefore, in this study, the emission factor for the residual mix is used for households that do not buy renewable energy.
The residual mix is based on the production mix of electricity produced in Finland from which the renewable sources are deducted (Energiavirasto 2022a). For 2021, the residual mix is 234.9 gCO 2 kWh −1 (Energiavirasto 2022b). The residual mix is only available from 2015 onwards. Therefore, for the years 2015-2019, the average ratio between the average mix and the residual mix is calculated. The ratio is used to estimate the residual mix for 2010, which is 551 gCO 2 kWh −1 . Based on annual values, an exponential extrapolation is conducted for 2030. Based on this approach, the predicted value is 104 gCO 2 kWh −1 .

District heat production
Lahti Energia provides district heat for the Mukkula region. There has been a significant drop in district heat production-related GHG emissions, especially in 2019, when coal use was replaced with biomass. The emission factor with benefit allocation for district heat is 58.5 gCO 2 kWh −1 in 2020, and it can be assumed to be the same for 2021. In 2014, the value is 190 gCO 2 kWh −1 , and it can be assumed to be the same for 2010 (Lahti Energia 2021).

Construction
According to , the average share of embodied GHG emissions of buildings (i.e. emissions from manufacturing and processing building materials) is approximately 20%-25% of the life cycle GHG emissions of buildings. The emission factor for the embodied GHG emissions of residential buildings is 6.7 kgCO 2 eq m 2 a −1 . Röck et al assessed that the embodied GHG emissions of residential buildings had not declined but rather had increased. In this study, we assume the embodied GHG emissions to remain the same during the years 2010, 2021 and 2030.

Results and discussion
To calculate GHG emissions from mobility and housing sectors, mobility and housing information of Mukkula residents (modal shares and distances per week, heating and electricity system in household etc.) was collected through the survey. The housing and mobility information of the survey respondents is presented more detailed in appendix D and E. Based on survey information and collected emission data, the development of mobility GHG emissions and that of GHG emissions from housing for the years 2010, 2021 and 2030 are presented in sections 3.1 and 3.2, respectively. In section 3.3, the total carbon footprint of Mukkula residents is presented for the years 2010, 2021 and 2030. In addition, the development of mobility and energy GHG emissions and how the decrease in GHG emissions in these sectors affects the consumption-based carbon footprint are discussed. In section 3.4, the sensitivity analysis for four carbon footprint scenarios is conducted. Finally, section 3.5 examines the limitations of the study. Figure 4 shows the respondents' average modal share of mobility modes, including passenger car, bus, train, walking, bicycle, motorcycle/moped, aviation and ferry. As shown in the figure, the average modal shares of aviation and ferry trips are based on the number of trips reported by the respondents and the distances presented in section 2.3. Other distances are based on the respondents' answers. The most common mobility mode is a passenger car based on the distance travelled, accounting for 43% (figure 1). Aviation is also one of the most commonly used mobility modes, accounting for 28% of the distance travelled. Public transport (bus, train) covers 11% of the distance travelled. More detailed mobility information is presented in appendix D.

Mobility
The use of public transport is clearly higher among the respondents than the average in the region (11% vs. 8% of distance travelled), despite the ongoing pandemic that greatly decreased the use of public transport in Lahti in 2020 (Finnish Transport Infrastructure Agency 2018; Kareinen et al 2022). On average, Mukkula residents travel 28 km d −1 (aviation not included), which is close to the average mobility in Lahti during the COVID-19 pandemic (33 km d −1 ) but less than that before it (40 km d −1 ) (Uusitalo et al 2022; Kareinen et al 2021). The differences may be due partly to the COVID-19 pandemic and partly to the study area being a suburb in a city, whereas the Päijät-Häme region also includes sparsely populated areas with long distances. GHG emissions from mobility have a clear declining trend from 2010 to 2030, when modal shares and distances are assumed to be the same as in 2021. According to the results in figure 2, GHG emissions from mobility decrease by 25% from 2010 to 2021 and continue to decrease by 34% from 2021 to 2030, leading to a total 50% decrease from 2010 to 2030. The average GHG emissions per capita, including the different mobility modes, in 2010, 2021 and 2030 are illustrated in figure 2. The GHG emissions from mobility are presented in more detail in appendix B.
Most of the GHG emissions from mobility are produced from aviation (55%, 2021) and passenger car use (38%, 2021). Aviation and passenger car use also have higher emission factors than motorcycle/moped use. Almost 50% of the climate impacts from aviation are indirect. Emissions from aviation decreases by 41% and those from passenger car use by 62% between 2010 and 2030. The share of emissions from other mobility modes is under 7% in all the examined years.
These results slightly contradict previous research. According to Akenji et al (2021), the average mobility GHG emission of a Finn is 3,650 kgCO 2 e a −1 , which is significantly higher than that calculated in the present study. Akenji et al (2021) found that 55% of mobility emissions are linked to passenger car usage and 35% to aviation. In our study, the average emissions from aviation are slightly lower than those in Akenji et al's study, and passenger car-related emissions are significantly lower. This difference can be explained by the fact that the amount of mobility for Mukkula residents is generally lower, especially due to the COVID-19 pandemic.
To meet the 1.5-degree target, Akenji et al (2021) estimated two carbon budget scenarios for the Finnish lifestyle with consumption-focused and system-focused perspectives for 2030. They found that the mobilityrelated carbon footprint should be reduced by around 70% in the system-focused scenario and by 80% in the consumption-focused scenario. To meet these reduction targets, they suggested that transport demand should be halved, or the average carbon intensity of mobility should be at the level of electric vehicles. That is, all current  fossil-based mobility modes should be changed to electric ones by 2030. Based on our results, mobility emissions can be reduced from the 2021 level by 34% by 2030. To meet the 1.5-degree target, the share of electric cars as passenger cars should be increased significantly in 2030, and changes in mobility behaviour are needed.
This study did not consider changes in mobility habits. Lapp et al (2018) reported that distance travelled per capita would increase by 6% in 2017-2030 in Finland as the number of trips decrease and the length of trips increase. Specifically, the distance travelled increases in aviation and rail transport, although the change is small. This leads to an increase in the mobility emissions of the 2030 scenario. However, this estimation does not include the effect of the COVID-19 pandemic.
As the demand for biofuels increases, the demand for biofeedstocks also increases. Competition for raw materials is increasing because the production of energy and materials has an increasing trend in the use of biofeedstocks (Norman and Talalasova 2021). The availability of sustainable feedstocks can become a bottleneck that limits real market uptake (Prussi et al 2019). In addition, the use of biofuels does not remove the multiplying climate effects in aviation. Javed et al (2020) summarised the forecasting of biofuel production and demand as a 'very complex issue since it is dynamically influenced by policy changes, biofuel price, vehicle population, farming practices, local water resources, land demand, cost of production, international oil prices, food prices, and competition with other grain food supply that can lead to unintended food security issues'. For example, the proportion of biofuels in petrol and diesel has been proposed to increase to 34% or higher from 30% by 2030 in Finland to consider the increasing use of biogas (Jääskeläinen 2021). However, the distribution obligation of biofuels decreases by 7.5% percentage points in 2022, which is also presented for the year 2023, due to a rapid increase in energy prices (TEM 2022). Therefore, there are uncertainties in the amount of biofuels used in the 2030 scenario.
The target of Finland is to have 700,000 electric cars account for all cars in 2030 (LVM 2021b) and for the share to increase further after 2030. Climate actions and low renewable electricity production costs may further increase the share of renewables in the electricity grid, thus decreasing emissions from electric mobility (Viktoria 2021). In the aviation sector, electrification may be conducted in short-haul flights, and renewable hydrogen may provide zero emission flights after 2035 (Airbus 2022; Innovation Origins 2021).

Energy
In this study, GHG emissions from the energy sector per capita are calculated to have decreased by 56% from 2010 to 2021 and to still decrease by 28% from 2021 to 2030. This leads to a 68% decrease in emissions from 2010 to 2030. Most of the reductions between 2010 and 2021 are due to the significant change in emissions from district heating. The reduction between 2021 and 2030 is due to the predicted higher share of renewable energy in electricity production. The average GHG emissions of housing per capita in 2010, 2021 and 2030 are presented in figure 3. The GHG emissions from the energy sector are discussed in more detail in appendix C.
Based on our results, 28% of GHG emissions from housing can still be decreased by 2030 ( figure 3). According to Akenji et al's (2021) scenario estimations for 2030, housing-related carbon footprint should be decreased 62% in the consumption-focused scenario and 76% in the system-focused scenario. In the consumption-focused scenario, emission reductions are projected to be achieved by reducing living spaces and total energy consumption demand. To meet the targets in the system-focused scenario, the efficiency of buildings should be improved, and the efficient use of space should be increased (e.g. decreasing additional office buildings). The type of fuel in heating and electricity systems affects the carbon footprint of housing. Electricity accounts for 38% of housing-related GHG emissions, and heating accounts for 46% in Finland. In Mukkula, the share of GHG emissions from heating is 64%, and electricity accounts for 17% of total GHG emissions from housing. The reason that the share of GHG emissions from electricity is significantly lower in Mukkula than in Finland is that almost one-third of the survey respondents (27%) had renewable grid electricity, which is assumed to be emission-free in this study.
In 2030, the annual carbon footprint of housing in Finland should be 600 kgCO 2 e to meet the 1.5-degree target (Akenji et al 2021). Based on our calculations, the carbon footprint of housing in Mukkula per resident could be 1,320 kgCO 2 e in 2030, when only technological development and societal actions are considered. Housing-related emissions in Mukkula should still be decreased by almost 55% to meet the 1.5-degree target in 2030. Therefore, technological efficiency in electricity production does not guarantee adequate emission reduction. In the energy sector, the focus should be on phasing out fossil-based fuels (e.g. coal, oil and peat) before 2030. Even though we did not assess the emission development of construction in this study, the GHG emissions of construction can be expected to decrease because the building and construction sector has set the goal of decreasing embodied emissions by 40% by 2030 (World Green Building Council 2022). Furthermore, as the climate change mitigation of established cities includes measures such as replacing, repurposing or retrofitting the building stock (Lwasa et al 2022), the construction of new buildings will not be essential in the Mukkula area.
In this study, we did not assess the role of residential oil heating in 2030 because it was not an essential part of heating systems in Mukkula, with only 2% of the respondents having oil heating systems in their households. In Finland, the residential oil heating system is commonly used in detached or semi-detached houses, whereas in the Mukkula area, residential buildings are mainly multifamily dwelling stocks heated by district heating. Replacing oil heating with pellet heating or a ground-source heat pump may be the most economical solution for a consumer if a subsidy for heating renovation is available (Hast et al 2016). The Finnish government has declared to phase out residential oil heating by the early 2030s, and the government offers subsidies for owners of detached houses for changing oil heating to another form of heating (Ministry of the Environment 2022). Vihola and Heljo (2012) estimated that only a small share of detached or semi-detached houses in Finland would be heated by oil in 2028 if the present rate of abandonment remained the same.

Total carbon footprint in the years 2010, 2021 and 2030
Based on the calculations of mobility and energy GHG emissions per person in the years 2010, 2021 and 2030, we estimated how increasing efficiency and technological development in the mobility and energy sectors could affect the total carbon footprint per capita ( figure 4). In addition to mobility and housing, we took into account the GHG emissions of food and other forms of consumption (e.g. hobbies, leisure, clothes). GHG emissions from food and other forms of consumption are based on Akenji et al's (2021) study of the average Finnish carbon footprint calculation. As the GHG emissions of food and other forms of consumption mostly depend on the consumption and lifestyle choices of people and are not easily influenced by society, we assumed that GHG emissions from food and other forms of consumption remained the same in all years.
As shown in figure 4, societal emission reduction measures only in the mobility and energy sectors do not achieve a sustainable level of carbon footprint per capita in 2030. In 2010, most of the emissions are from housing and mobility. When Lahti abandoned coal in its district heating system in 2019, housing emissions dropped significantly. More than half of the carbon footprint per capita consists of emissions from food and other forms of consumption in 2030 ( figure 4). When only decarbonisation actions are executed in the mobility and energy sectors, the total carbon footprint per capita may decrease by 14% in 2021-2030 and by 37% in 2010-2030. According to Akenji et al (2021), the carbon footprint of Finns should be reduced by more than 70% by 2030. This means that the annual carbon footprint of Finns should be less than 3 tonnes of CO 2 e in 2030. Based on our results, the average carbon footprint of Mukkula residents will be 6.8 t CO 2 e in 2030. Therefore, it should be decreased by more than half to achieve the 1.5-degree target.
In terms of carbon footprint reduction measures, many individual actions can supplement societal decarbonisation actions. To meet the 1.5-degree target, individual actions and behavioural changes should be considered in emission reduction measures. Especially in the food and other consumption sectors, individual actions have more potential to reduce emissions than societal actions. Wynes and Nicholas (2017) and Tolppanen et al (2021) found that everyday lifestyle actions, such as living car-free, avoiding air travelling and eating a plant-based diet, have the highest potential to reduce GHG emissions. Salo and Nissinen (2017) calculated that an average Finn could decrease the personal carbon footprint by 4,300 kgCO 2 e through consumption choices. Claudelin et al (2020) examined the potential climate impact of anti-consumption and found that in Finland, people could reduce their annual personal carbon footprint by 2,085 kg CO 2 e through anti-consumption choices. Considering the above, we can assume that Mukkula residents can reduce their personal carbon footprint by 2,000-4,000 kgCO 2 e. Our estimations of societal decarbonisation actions can decrease the carbon footprint to 6,800 kgCO 2 e in 2030. Therefore, if both individual climate actions and societal actions are considered, the average carbon footprint of Mukkula residents can be at the level of the 1.5-degree target in 2030. This estimate assumes that household consumption will not increase significantly in the future.
There can also be different capabilities to reduce personal consumption-based GHG emissions, for example people with a high-income level may have more potential to reduce consumption-based GHG emissions because high-income level usually correlates to high carbon footprint level ( Uusitalo et al (2021) have studied the effect of personal carbon trading (PCT) on mobility emissions, and they have found that people with a higher income level need to reduce emissions from mobility more than people with a lower income level due to allocations of the carbon allowances in PCT method.
Based on our results and according to Matthews and Wynes (2022), increased technological efficiency, societal decarbonisation actions and mid-century net zero goals are insufficient to limit global warming to the 1.5-degree target. Matthews and Wynes (2022) pointed out that system-level and individual lifestyle changes should be recognised and supported by political and corporate sectors to commit seriously to the 1.5-degree target.
In our calculations, we did not consider how geopolitical changes could affect societal decarbonisation actions in 2021-2030. For example, the military invasion of parts of Ukraine by Russia in 2022 changed clean energy policies in Europe (Steffen and Patt 2022). The European Commission (2022b) presented a plan (REPowerEU) to reduce dependence on Russian fossil fuels and speed up the green energy transition in Europe. The measures of the REPowerEU plan mainly focus on energy savings, diversification of energy supplies and acceleration of the rollout of renewable energy. Steffen and Patt (2022) asserted that Russia's war in Ukraine increased the level of public support for clean energy policies and led to an opportunity for an energy transition in Europe. The possible rapid energy transition in Europe may affect the climate targets of 2030 because the EU has presented new targets and additional investments that could be executed before 2027. For example, the European Commission proposes increasing the 2030 target for renewables from 40% to 45%, which doubles the solar photovoltaic capacity and the rate of deployment of heat pumps (European Commission 2022b). If the rapid energy transition is executed and the EU's dependence on Russian fossil fuels is terminated during the next decade, we assume that it will strengthen the possibility of reaching the 1.5-degree target through societal and technological actions.

Sensitivity analysis
As our results are based on predictions of future, there are some uncertainties in emission factors. During the data collection, we recognized four scenarios in which the emission factors and emission predictions are uncertain. There are uncertainties in indirect aviation emissions (S1), the share of electric cars in 2030 (S2), the residual emission factor for electricity production (S3), and the reduction of district heating emissions in 2030 (S4). To illustrate the role of different predictions and estimations, we carried out a sensitivity analysis for carbon footprint calculations (figure 5).
In S1, the indirect emissions of aviation are two times higher than in our baseline scenario. There is a relatively high uncertainty in the magnitude of indirect climate impacts on aviation, thus causing a relatively high uncertainty in aviation-related emissions. Some studies did not consider these indirect impacts but recommended their inclusion. We hope that future research can provide more accurate information about the magnitude of these impacts. We also used rough assumptions from flight distances because exact destinations were not asked about in the survey. As can be seen from figure 5, indirect emissions of aviation can increase the whole carbon footprint significantly if they are higher than predicted emissions in general.
The national target for electric vehicles for a year 2025 (100,000 cars) has already been fulfilled in the begin of 2022 and the share of electric cars in Finland has been doubled between 2021-2022 (Traficom 2022b). Therefore, in S2 we assessed scenario where the share of electric cars in 2030 will be increased +20% compared to the national target for a year 2030 (700,000 cars). However, based on figure 5, this scenario does not reduce the carbon footprint significantly compared to baseline scenario.
In S3, the emissions of electricity production are calculated with average national electricity emission factor (Statistics Finland 2020). In baseline scenario electricity emissions are calculated with the emission factor of residual mix of electricity. Data on the residual mix of Finnish electricity were available only for 2015-2021, and the values for the grid mix were available only for 2010-2019. Thus, estimations for 2010 and 2030 were conducted based on the ratio of emissions factors of the residual mix and grid mix between 2015 and 2019. These assumptions and predictions make the emission factor for residual electricity mix uncertain and therefore, electricity emissions are also calculated for average electricity grid mix. The emission factor of average electricity grid mix is lower than the emission factor of residual grid mix and thus the carbon footprint of S3 is lower than in the baseline scenario.
The city of Lahti aims to be carbon neutral by 2025, which means that energy production of the city can also be assumed to be carbon neutral. In S4, we assumed that the district heating production is considered as carbon neutral by the 2030. Based on figure 5, the carbon neutral district heating provides the most significant emission reductions between years 2021-2030.

Limitations
Finally, some important limitations need to be considered. In the survey, consumption habit-related questions were asked to understand and calculate the estimation of the Mukkula residents' consumption-based carbon footprint. However, it is challenging to include all detailed aspects of the respondents' lifestyles in one survey, and therefore, there are some uncertainties. In addition, the results depend on the respondents' own consumption estimates, which may increase uncertainty. It is possible that people under-or overestimated their mobility in the survey. In the future, it is recommended to collect information on mobility distances and modes using sensors or other methods that are not dependent on users' own estimations.
The survey was about climate issues, which could have influenced the respondents' interest in participating in the survey. The topic might have driven away people who were not interested in climate-related issues or attracted people who were interested in the topic. Therefore, the respondents of the survey might have already made conscious consumption choices and their carbon footprint could be lower than average.
There are also uncertainties in the data collection and future predictions, as can be seen in sensitivity analysis (figure 5).The results of this study are based on current existing data and estimates about emissions in 2030. The current study was unable to determine whether the rate of technological development would change in the next decade. Eight years is a relatively long time in terms of technology. This means that emission predictions for the year 2030 may be lower if technological development is faster than we have estimated.
The indirect effects of electricity and heating energy production are not considered in the study due to the uncertainties related to them. The shares of different electricity production methods vary from year to year. Also, reliable data on indirect emissions for energy production methods is scarce, except for certain methods related to renewable electricity such as solar and wind. Moreover, relatively little data were available on the manufacturing emissions of different mobility modes. Some data were also found from relatively old sources. It is important to include the vehicle manufacturing step in assessments to obtain a better understanding of the overall climate impacts of different mobility modes. We hope that more updated data on vehicle manufacturing, as well as on indirect effects of energy production, will be available in the near future.
The shift towards biomass use in district heat production was one of the key reasons for the reduction of consumption-based carbon footprint. However, if forest biomasses were utilised, there could be risks of additional emissions from forestry that were not considered in this study. In addition, there could be other sustainability effects related to the multiple actions considered in this paper that were not assessed. In the future, it would be important to assess multiple sustainability impacts instead of focusing only on climate impacts in order to obtain a more holistic view of the impacts.
The survey was conducted during the COVID-19 pandemic, which affected the survey results, especially those regarding mobility issues. During the first wave of COVID-19 in the spring of 2020, mobility decreased by approximately 40% in Lahti, and the GWP of mobility decreased by 36% (Kareinen et al 2022). Mobility in Lahti may have increased slightly since the first wave of COVID-19, but we assumed that it did not increase to the same level as before the pandemic in the autumn of 2021 when this study was conducted. Therefore, the emissions from the mobility of Mukkula residents could be lower than those before the pandemic. Although mobility increased toward the autumn of 2021 worldwide (Hoek and Laauwen 2021), it might not have reached the postpandemic level. For example, the number of passengers in Finnish airports in August 2021 was over four times smaller than the pre-pandemic level in August 2019 (Statistics Finland 2022c) (Traficom 2021).

Conclusion
This study examined the effect of societal decarbonisation actions on the consumption-based carbon footprint in a Finnish suburban area. The study showed that technological development and societal actions could decrease the mobility-based GHG emissions of suburban residents by 50% and housing-based GHG emissions by 68% in the years 2010-2030. The total annual carbon footprint per capita could decrease by 37% between 2010 and 2030.
The results of this study imply that emission reduction through societal decarbonisation actions and technological development only in the mobility and energy sectors does not achieve an adequate level of GHG emissions per capita to limit global warming to 1.5 degrees. This finding enhances our understanding of the need for behavioural change and public support to reduce GHG emissions in a high-emitting country such as Finland. Although the current study is based on a small sample of participants, the findings show that both technological and behavioural actions are needed to meet the 1.5-degree target in a suburban context in which emissions continuously increase.
At the city level, main emission reduction measures in energy and mobility sectors can be provided by accelerating the energy transition and reducing the use of private cars. Local energy companies of the city have a major role in reducing emissions from heat and electricity production. Emission reductions of energy sector at the city level affect significantly the personal carbon footprint of inhabitants, as can be seen in the city of Lahti. Another possibility to decrease the emissions at city level is to reduce the use of private cars by supporting the change of modal shares, community planning and proving tolls or car-free zones in a city. In addition to these emission reduction measures of mobility, the electrification of transport can be supported by enhancing the car charging infrastructure.
Considerably more work is needed to determine the role of individual actions in reducing consumptionbased GHG emissions. Another important practical implication is that Finland's current societal actions are not achieving the 1.5-degree target per capita; thus, more ambitious decarbonisation actions from other consumption sectors are also needed.  fuel/energy production 20.7 fuel/energy production 20.7 fuel/energy production 20.7 vehicle manufacturing (unknown) vehicle manufacturing ( (2017)