Land use change impacts of increased bioenergy demand in Brazil (FAPESP SPRINT 2016/50495-4)

Basic data for this project

Type of project: Own resources project
Duration: 01/03/2017 - 01/03/2018

Description

Introduction In its INDC (Intended Nationally Determined Contribution), presented at UNFCCC COP-21, Brazil has set ambitious goals for reducing emissions of greenhouse gases. The country has pledged to end illegal deforestation in Amazonia and to regrow 12 million hectares of new forest areas. And Brazil has also committed to increase the share of renewables in its energy matrix. One of the main challenges for public policy in Brazil is how the country can achieve all of these goals balancing production and conservation in the land sector. A substantial category in renewable energy is that of biofuels. Two main types of biofuels exist: bioethanol (replacement for petroleum gasoline) and biodiesel (replacement for petroleum diesel). Brazil is the second largest producer of bioethanol. Its annual production increased from 11.5 billion liters in 1990 to 28.0 billion liters in 2013, mostly from sugarcane (MAPA, 2015). Biodiesel is produced in Brazil mainly from soy (80%). Current production is estimated to be around 3 billion liters per year (MAPA, 2015; OECD/FAO, 2015). The production of both bioethanol and biodiesel is expected to increase given incentives of the government, e.g. the INDC, and the biofuel industry (OECD/FAO, 2015). Previous studies (Lapola et al., 2010; Verstegen et al., 2016) have shown a risk for land use change (LUC), either direct (dLUC) or indirect (iLUC), resulting from land expansion for biofuel production. This would jeopardize Brazil’s aspired production-conservation balance. This project will work to increase the scientific knowledge on this area, which is still subject to intense national and international debate. Objectives The goal of this research is to use land use models to capture the direct and indirect effects of expansion of bioenergy production in Brazil. We will address two main scientific questions: (a) What could be the land use effects of increased bioethanol production from sugarcane in Brazil? (b) What could be the land use effects of increased biodiesel production from soy in Brazil? Relevance Brazil has made a strong investment in domestic bioethanol and biodiesel production as a means to reduce greenhouse gas emissions and exposure to volatility of oil prices. Policy-makers in Brazil have high expectations of export opportunities of ethanol and biodiesel, based on expected increases of biofuel consumption in the USA and the European Union. These expectations have been limited by scientific studies that have argued that biofuels expansion in Brazil has significant environmental costs (Lapola et al., 2010; Arima et al., 2011). The studies have had an impact on public policy decisions such as the US EPA’s ruling that requires assessment of iLUC effects on biofuel usage (EPA, 2010). We consider that there is a need to revisit these studies, in the light of recent advances in land use modelling and measurement of land trajectories. Previous studies focused on first generation ethanol only, and not on second generation, as aimed for in the INDC. Using second generation instead of first generation can cause large differences in land use change impacts, compared to the above-mentioned studies, because the second generation production of ethanol from ‘residues’, e.g. filter cake, does not require any additional land. We also need new studies that consider the recent gains in agricultural productivity in Brazil. The percentage of soy grown in double cropping systems has grown rapidly in the past few years (Conab, 2016), which reduces the net land area required to cultivate the feedstock. Past studies have ignored or underestimated the percentage of soy grown in double cropping systems (Lapola et al., 2010; Özdemir et al., 2009), and therefore overestimate the land use change impacts of an increased biodiesel demand. Several conservation policies were implemented or amended in last decade, such as the soy moratorium in 2006 (Rudorff et al., 2011). These gains in productivity and conservation policies need to be incorporated in a new generation of models. Methods To better understand the impacts of biofuel expansion on land use, scientists use land use change models. Both applicants have extensive experience with the development of such models (Verstegen et al., 2016) and with statistical analyses of the factors driving land use change in Brazil (Aguiar et al., 2007; Espindola et al., 2012). In this project, we will use two land use change models that complement each other: 1. The GLOBIOM-Brasil model, a global recursive dynamic partial equilibrium model that includes the main sectors competing for land (agriculture, forestry and bioenergy). It operates on 50 km x 50 km grid cells, and will provide large-scale regional impacts on deforestation, emissions, agriculture, livestock and wood production, exports and imports. GLOBIOM-Brazil is an adaptation of the GLOBIOM model developed by IIASA (Havlík et al., 2011) and has been used as one of the scientific bases for Brazil’s INDC on the area of land use change. 2. The PLUC model, a demand-driven, spatially-explicit land use change model. It can be used in combination with an equilibrium model (see e.g. Verstegen et al., 2016) or stand-alone (Verstegen, Karssenberg, van der Hilst, & Faaij, 2012) to assess the dLUC and iLUC effects of demand increases for specified feedstocks. The current implementation for Brazil operates on 5 km x 5 km grid cells and has a framework for ensemble runs (Monte Carlo simulation). Data One important contribution of the study is to enable the inclusion in both models of the new land use data sets being produced by the e-sensing project, which is FAPESP’s thematic project to which this proposal is linked. The e-sensing project is developing innovative ways of extracting land use information from big Earth observation data sets (Maus et al., 2016). The new data sets being produced by the e-sensing project will provide new information required by land use models, such as the extent of double cropping and the location of pasturelands in the Cerrado and Amazonia.

Keywords: Modelling; Land use change; Bioenergy; Brazil