IDENTIFYING PATHWAYS TO ACCOMPLISH BRAZIL'S NDC THROUGH AN INTEGRATED ASSESSMENT MODELING APPROACH

Background: The increasing awareness about climate change and the adverse effects of global mean temperature increasing beyond +1.5 ∘ C above pre-industrial levels resulted in a historic international climate agreement in December 2015 in Paris. Countries around the globe published their intended Nationally Determined Contributions (hereafter iNDCs)–converted into Nationally Determined Contributions (NDC), committing to take actions post-2020 to tackle global warming, mainly to mitigate greenhouse gases (GHG) emissions. The Brazilian NDC established absolute emissions targets of 1.3 GtCO 2 e by 2025 and of 1.2 GtCO 2 e by 2030 (GWP-100, AR5), corresponding to reductions of 37% and 43%, respectively, compared to 2005. In this work, we studied the role that each economic sector can play to meet the Brazilian NDC through an integrated assessment modeling (IAM) approach. Results: The analysis showed that the AFOLU (Agriculture, Forestry and Land-Use) sector would contribute with mitigation amounts of 25.5 MtCO 2 e in 2025 and 145.8 MtCO 2 e in 2030, considering implementation of no-regret abatement measures (LC0) and LC10 scenario, which implies an economic effort to internalize a carbon price of US$ 10/tCO 2 e in the economy, respectively. Potential emissions reductions in the energy system would contribute to the mitigation of 60.4 MtCO 2 e in 2025 and 211.1 MtCO 2 e in 2030. Additionally, we identified critical measures with higher mitigation potential, for instance, commercially planted forests, integrated crop-livestock-forestry systems, no-tillage systems, biological nitrogen fixation application, intensification of livestock production through cattle confinement, deforestation reduction, expansion of native vegetation, and degraded pastures recovery in the AFOLU sector. Regarding the energy system, the integrated modeling demonstrated high mitigation potential in measures related to energy efficiency in the industry, waste management, and transport sectors; as well as a modal shift from individual to collective passenger transport and highways to railways and waterways to load transportation, and energy utilization of urban solid waste and effluent of treatment plants for the production of biomethane and electricity. Conclusions: Projected emissions for 2025 demonstrate that the NDC target for this year could be achieved with the LC0 scenario (carbon value equal to zero) implementation, while the NDC target for 2030 could be achieved by implementing the LC10 scenario in 2030.

95% of global GHG emissions. These countries are committed to limit the global average temperature increase to below 2°C from pre-industrial levels, through the analysis of action plans called Intended Nationally Determined Contributions (iNDC). These INDCs were altered to Nationally Determined Contributions (NDC) after being ratified by each country. In addition, the Paris Agreement also established that the parties should indicate more ambitious long-term contributions, encouraging efforts to limit the temperature increase by up to 1.5°C (UNFCCC, 2016). However, studies indicate that, in the absence of mitigation efforts, the current Brazilian energy mix will continue on a trend of increasing carbon intensity, with natural gas and coal gaining importance in the power sector, and the sugar-alcohol sector undergoing a severe crisis that has caused the closure of several ethanol distilleries. For instance, the depletion of the hydropower potential outside the Amazon region, and the vulnerability of existing hydro capacity to climate change (LUCENA et al., 2010;LUCENA et al., 2009), it means that other sources would take on increasing roles in meeting baseload demand, with results showing coal to be the least cost solution (NOGUEIRA et al., 2014;LIMA et al., 2015;HERRERAS MARTINEZ et al., 2015;PORTUGAL-PEREIRA et al., 2016).
In the case of agriculture, forestry and other land use (AFOLU) sector, a long stream of studies has affirmed that reducing emissions from deforestation is a cost-effective way to mitigate climate change in Brazil (NEPSTAD, et al, 2009;KINDERMANN, et al, 2008;FEARNSIDE, 2005;KRUG, 2018). Indeed, since 2005 the emissions from deforestation have reduced substantially due to the scaling up of law enforcement, the development of new monitoring systems (RAJÃO; GEORGIADOU, 2014;RAJÃO;VURDUBAKIS, 2013), the expansion of protected areas (SOARES-FILHO et al., 2010) and the creation of deforestation-free supply chains schemes for soy and beef (GIBBS et al, 2015). At the same time, it has become increasingly clear the importance of intensifying cattle ranching and promoting low carbon agricultural techniques in order to keep emissions from growing a scenario of intense expansion of the sector (STRASSBURG et al., 2014).
Integrated assessment models (IAMs) map the interactions between socio-economic systems and energy and environmental processes and are used to develop emission scenarios, estimating the costs and benefits of mitigation policies and the economic impacts of climate change. IAMs experiences combine models from different areas of knowledge (CLARKE et al., 2016;PRADHAN et al., 2019). A detailed representation of the energy system is necessary -considering conventional and alternative energy uses.
The same applies to land use -considering agriculture, livestock, and forests -and the economic system, considering sectoral elasticities and productivities, as well as for ecosystems, even if in a simplified approach.
In this context, this study aimed at highlighting the role that each economic sector can play to meet the Brazilian NDC targets for 2025 and 2030, using the IAM approach to identify the most cost-effective mitigation options that should be given priority in order to fulfill Brazil's NDC commitment.

RESULTS
As detailed in the methods section, the three modeling tools were integrated to ensure that the energy system results were consistent with the macroeconomic outputs while also agrees with land-use evolution in Brazil (cost and productivity, and final energy demand from the agricultural sector). In other words, an iterative procedure was performed integrating the analysis until the technical coefficient of the Computable General Equilibrium (CGE) model agreed with the MESSAGE, while the output in terms of bioenergy agreed with the OTIMIZAGRO model. For this reason, our findings are entirely consistent and very detailed, allowing us to indicate where different mitigation options could be implemented to help comply with Brazil´s NDC.
The integrated modeling of GHG emission scenarios was performed until 2050 ( Figure   1). However, the results related to the accomplishment of Brazilian NDC emission targets (2025 -2030 period) are emphasized ( Figure 1A). demonstrate that the NDC target for this year (1,300 MtCO2e) involves a 4% emission reduction effort regarding the REF scenario; thus, the commitment could be achieved with the LC0 scenario (carbon value equal to zero) implementation ( Figure 2B). It is worth noting that although technically and economically feasible, the LC0 scenario is not cost-free and comprises non-economic barriers to its implementation.  Figure 1B). When analyzing the AFOLU sector, it is possible to see that it contributes with mitigation amounts of 25.5 MtCO2e in 2025 and 145.8 MtCO2e in 2030, considering LC0 and LC10 scenarios, respectively (Table 1). However, it is noteworthy that emissions related to waste burning and agricultural soils are higher in LC0 and LC10 scenarios compared to the REF scenario. It happens due to an increase in biofuels production (required as a mitigation measure implemented on the transport sector), as well as mitigation actions related to degraded pastures restoration and planted forests expansion. In Table 2 are demonstrated the AFOLU sectoral mitigation options to be implemented in 2025 and 2030 that allow achieving the sector mitigation potential described above. In 2025, under LC0 scenario implementation, the main measures would be increasing commercial planted forests, integrated crop-livestock-forestry systems, no-tillage systems, and biological nitrogen fixation (BNF) application.
Aiming to reach the LC10 scenario in 2030, mitigation actions related to cattle ranching intensifying through confinement, as well as increasing deforestation reduction, expansion of native vegetation, and degraded pastures recovery, would be implemented together with the measures related to the LC0 scenario.
Deforestation reduction involves reduction rates in Amazon, Caatinga, Pampas, and Pantanal biomes. In addition, it also considers the recovery of 9.3 million hectares of native vegetation by 2030.
In summary, it involves promoting actions for the reduction of deforestation in conjunction with the expansion of native vegetation areas and planted forests, as well as increasing the stock of carbon in the soil with the expansion of integrated systems and the recovery of degraded pastures. hydroelectric potential from 2030 on. Finally, the use of wood fuel biomass for energy is also significant.
The results obtained to the REF scenario indicate that Brazil will follow a conservative path related to the energy mix, with fossil sources ranging between 50% and 60%, reaching its maximum value in 2040, when the oil and gas supply would also reach its production plateau. Regarding the LC0 and LC10 scenarios, primary energy consumption is reduced, mainly from 2040 to 2050, with particular emphasis on lower consumption of oil in refineries instead of sugarcane processing for ethanol production. Coal consumption is also reduced due to the lower expansion of coal-fired power generation, especially in the LC10 scenario (2030-2050).    The mitigation options related to GHG emissions reductions described in the energy system are demonstrated in Table 4. Regarding the industry and energy sectors, the mitigation options identified by integrated modeling are efficiency on heat and steam recovery, flare reduction, and installation of steam recovery units in oil and gas E&P platforms, as well as substitution of coal by sugarcane bagasse on thermal plants. In contrast, the mitigation options identified to transport were a modal shift from individual to collective passenger transportation as well as from highways to railways and waterways to load transportation. In the waste management sector, the options identified were mainly the energy utilization of urban solid waste and effluent of treatment plants for production of biomethane and electricity.

DISCUSSION
This study investigated the role that each economic sector can play to meet the GHG emission targets for 2025 and 2030, and thus comply with Brazil's NDC commitments.
An integrated analysis was performed using soft-links between three leading Brazilian developed tools: a CGE model, named EFES, which provides and guarantees the macroeconomic consistency of the analysis; an energy system optimization model, named MESSAGE, which provides different trajectories for the Brazil' 's energy system, in a highly detailed techno-economical fashion (including GHG emissions from fuel combustion, industrial processes, fugitive emissions, and waste treatment); and a landuse optimization model, named OTIMIZAGRO, which can optimize at micro spatial resolution the AFOLU sector in Brazil.
As a result of the iterative procedure, this study can add different mitigation options into a greenhouse gas emission strategy to reach the Brazilian NDC. This strategy is compatible with the same trajectory of the economy, the energy system (supply and demand), and the land-use sector; this approach avoids incompatibility of measures between the sectors. For instance, it is expected to measure the mitigation potential and the marginal cost of GHG emission reduction by mapping technical-economic parameters of low carbon activities. However, the sectoral analysis does not allow the detection of non-additive mitigation potentials that may derive, for example, from the competition for energy inputs and technologies aimed at the reduction of GHG emissions. It is the case of natural gas, which is a contested energy input to mitigate emissions in thermoelectric generation and production of heat and steam for industrial processes, replacing more energy-intensive fuels (coal and fuel oil). Sectoral models are not able to measure the effects of the dispute on the availability and prices of natural gas and, as a consequence, tend to over and underestimate, respectively, the potential and cost of mitigation associated with the substitution of coal and fuel oil by natural gas.
The results showed that the LC0 scenario is compatible with meeting the NDC target for scenarios, representing a higher national effort for compliance with the NDC.

CONCLUSIONS
Through the IAM approach, we demonstrate that the NDC target for 2025 could be achieved with the LC0 scenario (carbon value equal to zero) implementation, while the NDC target for 2030 could be achieved by implementing the LC10 scenario in 2030, which implies an economic effort to internalize a carbon price of US$ 10/tCO2e. In addition, it was identified the low carbon activities with higher mitigation potential in all sectors of the economy, which need to be implemented to comply with the NDC targets.
The results obtained indicated a rationale that takes into consideration sector cost-carbon effectiveness and key technologies for compliance with the NDC. However, it is necessary to establish a national consensus where technology action plans can build from, aiming to support the policy-making process. Therefore, this study contributed to the development of a comprehensive Technology Needs Assessment, which is a methodology defined as a group of activities conducted by a country leading to the identification, prioritization, and diffusion of environmentally friendly; the initiative is funded by the Green Climate Funding (www.greenclimate.fund/document/strategicframeworks-support-brazil-through-unep), and it is expected to be concluded at 2021.

METHODS
The integrated modeling of GHG emissions scenarios started with boundary conditions from a macroeconomic consistency model (dynamic stochastic general equilibrium -DSGE) that generated data for the EFES Model (KANCZUK, 2001 and2004;HADDAD, DOMINGUES, 2016). The key variables used for the construction of sectoral scenarios of energy supply and demand, as well as land use and land use changes, were projected at EFES, including Gross Domestic Product (GDP), the gross value of production, valueadded, staff employed, work income.
Due to the correlation between the level of economic activity and GHG emissions, it was considered projections of macroeconomic aggregates that reflect, especially in the short term, the current conditions of the Brazilian and world economy. The GDP projections considered are shown in Table 5. Those data, together with variables of characterization and evolution of the economic sectors, such as production, technological, energy, land use, and GHG emissions profile, enabled a bottom-up disaggregated sectoral modeling. Optimization and simulation models were developed and applied for the following sectors: industrial, energy, transport, buildings (residential, commercial, and service), AFOLU, and waste management.
The sectoral modeling was applied to simulate the sectoral agents' behavior, translating it into energy demand and land use and land-use changes to project GHG emissions.
Therefore, the use of sectoral models allowed a database elaboration to enable the integration of projections into the energy, AFOLU, and economic models (MESSAGE, OTIMIZAGRO, and EFES, respectively). This integration was a soft-link type, which demanded the transposition of results between the models. More information on these models can be found in the Supplementary Material.   Table 6 summarizes the main assumptions considered in the reference and low-carbon scenarios. Increase to 90% of these areas with conservation systems.
Target of the ABC Plan for the area occupied with integrated farming systems by 2020 and maintenance of the adopted ratio between 2021 and 2050.
Target of the ABC Plan to 2020 and a 50% increase in the target of the occupied area between 2021 and 2050.
Application of BNF in 100% of areas planted with soybean and 10% of rice, bean, corn and wheat areas.
Expansion of BNF to 30% of the areas of rice, beans, corn and wheat and the inclusion of sugarcane areas.

Livestock
Projection of cattle ranching aiming to meeting the expected demand for meat, according to Decrease in the ratio of native forest fuel wood to 10% by 2050.

Native forests
Deforestation reduction targets of 80% and 40% in the Amazon and Cerrado biomes, respectively, applied to the deforestation target for 2002 and 2010, and prohibition of cutting native vegetation in the Atlantic Forest; Recovery of environmental liabilities of 12.5 million hectares in 20 years and additional recovery of 4.5 million hectares between 2035 and 2050.
Same as the previous item, with only legal deforestation in the Amazon, and the application of a 40% deforestation reduction target in the Caatinga and Pantanal biomes, and 58% in the Pampas biome; Expansion of native vegetation restoration to 21 million hectares by 2050.

Energy System
Expansion of the energy system at minimum cost; Introduction of available technologies in the baseline; No adoption of additional mitigation policies; Predominance of the sectoral perspective on modeling; Short-term trajectory adhering to the current and planned expansion of the energy system.
Expansion of the energy system, considering different levels of carbon pricing; Introduction of the best available technologies and production practices; Internalization of different levels of carbon pricing in the economy; Freedom to select the evolution of the technological profile and the optimization of the energy system, according to a logic of GHG emission mitigation.