CHEMICAL ENGINEERINGTRANSACTIONS VOL. 56, 2017 A publication of The Italian Association of Chemical Engineering Online at www.aidic.it/cet Guest Editors: JiříJaromírKlemeš, Peng Yen Liew, Wai Shin Ho, JengShiun Lim Copyright © 2017, AIDIC ServiziS.r.l., ISBN978-88-95608-47-1; ISSN 2283-9216 Spatial Inventory of CO2 Emissions and Removals from Land Use and Land Use Changes in Thailand Poonsak Miphokasap Department of Geography, Faculty of Social Science, Kasetsart University. 50 Ngam Wong Wan Rd., Ladyaow, Chatuchak, Bangkok, Thailand. fsocpsm@ku.ac.th The main objective of this study aims to implement the spatial inventory of Carbon Dioxide (CO2) emissions and removals during 2007-2012 resulting from carbon stock changes in biomass for all types of land. The methodological approach and land-use categories were based on the 2006 Intergovernmental Panel on Climate Change (IPCC) guidelines in the Agriculture, Forestry and Other Land Use (AFOLU) sector. To depict the spatial extend and distribution of CO2 emissions and removals, Geographic Information System (GIS) was applied as a key tool for accounting the land-use changes and quantifying amount of CO2 emissions per unit area. The results indicated that changes of above and below ground biomass during five years period lead to the net total CO2 values of 15.32, 0.93, -0.59, and 0.02 MtCO2 in Nan, Udonthani, Suratthani and Bangkok respectively. The proportions of CO2 emissions in Nan were larger than CO2 removals which were a result of conversions of forest land to crop land. In Udonthani, there are small amount of CO2 emission due to the conversions of forest land to the various types of cropland. In Bangkok, there are the lowest CO2 emissions during the study time period. This is because of the small proportions of above and below ground biomass. To accomplish the inventory in Bangkok, it needs to calculate an amount of CO2 emission from other sectors. As a result of net CO2 among emission and removal, it could be concluded that Suratthani was identified as a CO2 sink. This is mainly due to the large coverage of forest land and perennial crop e.g. rubber and oil palm. In contrast, Nan was recognized as a CO2 source with the large amount of CO2 emission. The success of this study will lead to benefit for monitoring CO2 status and for implementing more sustainable action, strategies, good practices and practical policy for CO2 reductions at the local level for supporting the low carbon city. 1. Introduction Global warming and climate change is recognized as the critical issue which recently influences to the physical environment, social and economic developments (Intergovernmental Panel on Climate Change: IPCC, 2007). In additions, climate change phenomenon affects to the variation of meteorological factors e.g. rainfall, wind velocity, and temperature (Xu et al., 2015).The Greenhouse Gas (GHG) concentrations in the atmosphere have been proportionally emitted from the human activities as shown in the changes of land-use patterns. Among GHGs, Carbon Dioxide (CO2) with the lowest global warming potential contributes most to the current warming. The main sources of CO2 are fossil fuel combustion, deforestation, and land-use changes related to above and below ground biomass, dead organic matter and soil carbon stock changes. Land use changes and land management influence a variety of ecosystem processes that affect greenhouse gas fluxes such as photosynthesis, respiration, decomposition, nitrification and combustion.The inventory of CO2 emissions and removals can occur across all different types of land categories (IPCC, 2006). To implement the CO2 spatial inventory, it needs to conclude the total change in areas of land categories which were sub-divided into land remaining in the same category and land converted from on category to another. In the developing countries, patterns of land-use have been dramatically changed relying on the seasonal, physical and socioeconomic conditions (IPCC, 2007).One of the most significant elements of CO2 reduction leading to low carbon society is the reliable and precise of GHG emission and removal baseline. To determine such baseline, amount of CO2 distributions in the spatial extent with the specific locations of source and sink have to be delineated. DOI: 10.3303/CET1756003 Please cite this article as: Miphokasap P., 2017, Spatial inventory of co2 emissions and removals from land use and land use changes in thailand, Chemical Engineering Transactions, 56, 13-18 DOI:10.3303/CET1756003 13 IPCC released the guidelines for national greenhouse gas inventories, e.g. revised 1996 and 2006 guidelines. There are six land-use categories in the 2006 IPCC guidelines including forest land, cropland, grassland, wetland, settlement, and other land. This guideline improved the estimate methodology and provided the three different approaches used to represent the areas of land-use categories depending on the details of data and acquisition (IPCC, 2006). An approach 1 identifies the total change in areas for each individual land-use category within a country, but does not provide information on the nature and area of conversions between land-uses. An approach 2 introduces the tracking of land-use conversions between six land categories, but is not spatially explicit. Approach 3 extends approach 2 by allowing land-use conversions between two time periods to be tracked on a spatially explicit basis. In general, only approach 3 will allow for the CO2 spatial distributions and geo-reference of CO2 sources and sinks. To establish the CO2 spatial inventory based approach 3, Geographic Information System (GIS) was needed to apply both for detecting the extents of land-use change and for quantifying the CO2 emissions and removals per unit area. GIS based GHG inventory allow us to identify what an activity is (urbanization, deforestation, etc.), where the emissions source or sink are, and how much GHG emissions. GIS based tools have been also developed for accounting GHG emissions and reductions with the various purposes, for example, GHG from reforestation and deforestation (Richards and Brack, 2004), mapping individual anthropogenic CO2 point sources in fossil-fired power plants geo-referencing carbon sources and sinks based land-use change (Ponce- Hernandez, 2004). In Thailand, Initial National Communication (INC) and the Second National Communication (SNC) which represents the national greenhouse gas inventory were submitted to the United Nations Framework Convention on Climate Change (UNFCCC) in November 2000 and in March 2011, respectively (ONEP, 2010). Annual GHGs Inventory reports in Land sector were implemented using the statistical data based approach 1based the national data level. Thus, the inventory results cannot support the facilitation of low-carbon policy making at the local or municipal level. To mitigate greenhouse effects, it is essential to issue the policy and produce an action plan particularly in regional, local and city scale based temporal-spatial data (Asdrubaliet al., 2013).The main objective of this study aims to implement the spatial inventory of CO2 emissions and removals during 2007-2012 resulting from carbon stock changes in biomass for all types of land. The methodological approach and land- use categories were based on the 2006 IPCC guidelines in the Agriculture, Forestry and Other Land Use (AFOLU) sector. GIS based spatial inventory is beneficial for monitoring CO2 status, and implementing more sustainable action, strategies, good practices and practical policy for CO2 reductions at the local level for supporting the low carbon city. 2. Study Area In this study, the four provinces with the different dominance in land-use types were selected for evaluating the proposed GIS-based CO2 inventory. Nan province with the complicated mountain and the large proportion of forest land and cropland is located at the north of Thailand (centered at 18 51’ 3.96” N 100 50’ 6.36” E). Udonthani province is situated at the north eastern region and most areas are covered with an annual crop plantation such as rice, corn and cassava (centered at 17 25’ 30.54” N 102 51’ 47.52” E). Suratthani province located in the south region has the large spatial extents of perennial tree crops e.g. oil palm, rubber (centered at 9 3’ 3.79” N 99 5’ 21.4” E). At last in Bangkok, all of areas are mainly utilized for settlement, business and industry (centered at 13 46’ 17.22” N 100 37’ 19.92” E). 3. Research Methodology A GIS-based technical framework and five crucial steps of accounting and mapping the CO2 emissions and removals were illustrated in Figure 1. The details of each step were summarized as below. 3.1 Re-arranging land use categories The spatial layers (1: 25,000) of land-use categories which were interpreted from satellite image, aerial photo and ground survey operated by the Land Development Department (LDD) were used as the activity data source. However, five main land-use categories of LDD differ from those of the 2006 IPCC guidelines. Thus, it needs to re-arrange the land-use categories to be consistent. The six land categories based the 2006 IPCC guidelines were sub-divided into land remaining in the same category and land converted from on category to another. Each land category was further subdivided into sub-categories For instance, classes of cropland categories consist of the long-term cultivated, paddy rice and perennial tree crop. 14 3.2 Classifications of land-use change The CO2 inventory based 2006 IPCC guidelines focused both on the remaining and conversion of land-use category. Before measuring the CO2 emissions and removals, types of land-use change were derived by the overlay of re-arranging land use layers based IPCC category in 2007 and 2012. The comparisons of six land- use categories in 2007 and 2012 resulted in the 34 different types. Subsequently, 34 IPCC default codes were determined into the vector attributes to classify the types of land-use change. Figure 1: Research Methodology Framework 3.3 Conversions of land-use changes based vector to raster format To make the possibility to detect the total areas of land conversion type, polygon features of land-use change were converted to the raster format which represents the data with the equally square grid. The pixel size of each grid was 100 x 100 in meter which corresponds to the minimum unit area of IPCC emission factor (tonnes C per ha). 3.4 Calculation of CO2 emissions and removals The emissions and removals of CO2 for the AFOLU Sector, based on change in ecosystem C stocks, are estimated for each land-use categories including both land remaining in the same category and land converted to another land-use (IPCC, 2006).For all subdivisions of land area category, carbon stock changes were separately calculated and summarized in six land-use categories as shown in Eq(1). ∆CAFOLU= ∆CFL+∆CCL + ∆CGL+ ∆CWL+∆CSL +∆COL (1) where ΔCAFOLU is carbon stock change in AFOLU sector, FL is Forest Land, CL is Cropland, GL is Grassland, WL is Wetlands, SL is Settlements and OL is Other Land Annual carbon stock changes in each land-use type can be calculated using the Gain-Loss method. The Gain- Loss Method requires the biomass carbon loss to be subtracted from the biomass carbon gain as shown in Eq(2). Gains can be attributed to biomass growth at above and belowground components. Losses are categorized into wood harvesting, and losses from natural disturbances such as fire and flooding. ∆CB = ∆CG - ∆CL (2) where ∆CB is annual change in carbon stocks in biomass for each land sub-category (tonnes C yr-1). ∆CG is annual increase in carbon stocks due to biomass growth for each land sub-category. ∆CL is annual decrease in carbon stocks due to biomass loss for each land sub-category. The inventory of CO2 emissions and removals in the four provinces with the different of dominant land use types were conducted over the inventory period. In this study, the parameter values such as emission and stock change factors that are based on the country-specific data were used to calculate the carbon change in the forest land. However, the country specific emission factors in other land-use categories were not yet readiness. Thus, the tier 1 default emission factors which were provided by the 2006 IPCC guidelines were utilized. The default values were appropriately determined by considering both an ecological zone and climate domain of LU in 2012 based IPCC LU in 2007 based IPCC Emission Factor (per ha) LU and LU Change based Raster (Six Layers) Re-Arranging LU Classes Consistency with IPCC Categories LU in 2007 by LDD LU in 2012 by LDD LU Change CO2 Emissions and Removals Map Vector to Raster Conversion (1 pixel =100 m x 100 m) 34 Type of LU Change Sum CO2 for each Source Categories by LU and Stratum GIS-Overlay Analysis 15 Thailand. Subsequently, each grid cell was possible to assign a specific value of CO2 which were corresponding to the types of land-use changes, emission and stock change factors. The CO2 emissions or removals have been represented using a mesh of 1 ha square cells. 4. Results and Discussion 4.1 Magnitudes of the land use remaining and land conversions The main proportions of land-use categories in Nan province were the forest land and cropland with 64.74 % and 32.49 %, respectively. As a result of land-use change detections, it can be noticed that Nan province loss the forest lands approximately 74,400 ha during five years of investigation. The forest lands were converted to other land use types, particularly to croplands with the total areas of 89,741 ha. However, there is the increasing of forest lands with the total areas of 15,962 ha. In Udonthani province, most of areas in 2012 were utilized for cropland and forest land with the total area of 697,565 ha (67.68 %) and 193,559 ha (18.78 %), respectively. In additions, the total forest areas of 12,332 ha (1.12 %) were changes to cropland e.g. rubber, sugarcane. In Suratthani province, there is the largest coverage of perennial crop parcel in the southern region of Thailand. Most of croplands cultivated the rubber, orchard and oil palm with the total area of 789,985 ha (60.4%) in 2012. It can observe that Suratthani loss the forest lands with the total areas of 13,895 ha during five years. In Bangkok, there are the total croplands and settlements of 28,518 ha (18.21 %) and 100,432 ha (64.14%) in 2012, respectively. It can be concluded that there are the extents of settlement with approximately 4,021 ha during five years as illustrated in Table 1. Table 1: The sample of Land- use Change Matrix in Suratthani province Province Land categories Land-Use Change Matrix (ha) FL CL GL WL SL OL Final area (2012) Suratthani FL 398,053 196 44 1 - 3 398,297 CL 13,942 759,859 9,084 2,016 4,560 524 789,985 GL 131 771 17,441 4 78 43 18,468 WL - - - 45,470 - - 45,470 SL 23 906 299 44 33,317 6 34,595 OL 43 245 39 14 75 20,681 21,097 Initial area (2007) 412,192 761,977 26,907 47,549 38,030 21,257 1,307,912 4.2 The spatial distributions of GHGs concentration With the multiplication between the total areas of land remaining or land conversion and emission factor (tonnes C per ha), it is possible to map the spatial distributions of CO2 emissions and removals within a mesh of 1 ha square cells. The total emission causes the carbon stock loss in biomass which corresponds to the land-use categories including forest land, cropland, grassland, settlement, wetland and other lands. The results indicated that changes of above and below ground biomass during five years period lead to the net total CO2 values of 15.32, 0.93, -5.94, and 0.02 MtCO2 in Nan, Udonthani, Suratthani and Bangkok respectively. The details of CO2 emissions and removals were illustrated in Table 2. The loss of forest land during five years period of 89,741 ha in Nan province was mainly through the conversions to crop land which resulted in the large proportions of CO2 emission with values of 17.52 MtCO2. However, the remaining of forest lands and croplands during five years can uptake the CO2 values of 1.88 and 0.54 MtCO2. Thus, it could be conclude that Nan is the largest CO2 source among the four study areas. This is mainly due to the loss of forestlands resulting in the loss of carbon stock. When considered the subdivided land-use classes within the changed areas, it found that most of forest land was changed to maize and rubber plantations. In Udonthani, the total CO2 emissions mainly cause the forest land converted to cropland with the total area of 12,332 ha resulting in emissions of 1.63 MtCO2. However, the remaining of forest lands and croplands during five years can absorb the CO2 values of 0.92 and 0.85 MtCO2 respectively via the photosynthesis process. Most of the cultivated croplands were an annual trees e.g. sugarcane, oil palm and cassava. With the net total CO2 value of 0.93, it could be said that Udonthani is the rather small CO2 source. Suratthani is considered as the CO2 sink. This is mainly because of the large coverage of forest lands and perennial crop e.g. rubber, orchard and oil palm. The remaining of forest lands and croplands during five years can remove the CO2 from the atmosphere with the values of 2.39 and 7.12 MtCO2 respectively. 16 Table 2: CO2 emissions and removals from Land Use and Land Use change during 2007- 2012 Land Categories CO2 Emissions and Removals (tCO2) Nan Udonthani Suratthani Bangkok Total 15,324,674 937,355 -5,943,558 21,475 3.B.1 - Forest land -2,027,118 -1,181,318 -2,388,746 -957 3.B.1.a - Forest land Remaining Forest land -1,882,660 -924,529 -2,387,120 -832 3.B.1.b - Land Converted to Forest land -144,458 -256,789 -1,626 -125 3.B.1.b.i - Cropland converted to Forest Land -134,700 -230,118 -1,354 -82 3.B.1.b.ii - Grassland converted to Forest Land -8,784 -21,648 -259 -20 3.B.1.b.iii - Wetlands converted to Forest Land -41 -862 -3 - 3.B.1.b.iv - Settlements converted to Forest Land -931 -1,384 - -10 3.B.1.b.v - Other Land converted to Forest Land -2 -2,777 -10 -13 3.B.2 - Cropland 17,026,942 951,370 -3,753,462 -17,271 3.B.2.a - Cropland Remaining Cropland -537,530 -7,120,040 -16,668 3.B.2.b - Land Converted to Cropland 17,564,472 1,796,843 3,366,578 -603 3.B.2.b.i - Forest Land converted to Cropland 17,515,000 1,627,661 3,151,370 - 3.B.2.b.ii - Grassland converted to Cropland 52,350 188,442 282,872 4,229 3.B.2.b.iii - Wetlands converted to Cropland -219 -11,045 -19,213 -391 3.B.2.b.iv - Settlements converted to Cropland -2,659 -1,839 -43,457 -3,469 3.B.2.b.v - Other Land converted to Cropland - -6,376 -4,994 -972 3.B.3 - Grassland 233,651 827,554 156,887 5,354 3.B.3.a - Grassland Remaining Grassland - - - - 3.B.3.b - Land Converted to Grassland 233,651 827,554 156,887 5,354 3.B.3.b.i - Forest Land converted to Grassland 56,260 179,711 35,718 1092 3.B.3.b.ii - Cropland converted to Grassland 179,016 677,156 127,471 20,400 3.B.3.b.iii - Wetlands converted to Grassland -90 -21,029 -212 -559 3.B.3.b.iv - Settlements converted to Grassland -1,535 -4,886 -3,940 -14,728 3.B.3.b.v - Other Land converted to Grassland - -3,398 -2,150 -851 3.B.5 - Settlements 68,356 135,714 33,757 30,538 3.B.5.b - Land Converted to Settlements 68,356 135,714 33,757 30,538 3.B.5.b.i - Forest Land converted to Settlements 55,205 91,716 4,990 - 3.B.5.b.ii - Cropland converted to Settlements 9,770 23,792 16,607 13,528 3.B.5.b.iii - Grassland converted to Settlements 3,381 20,206 12,160 17,010 3.B.6 - Other Land 22,843 204,035 8,006 3,811 3.B.6.b - Land Converted to Other land 22,843 204,035 8,006 3,811 3.B.6.b.i - Forest Land converted to Other Land 18,570 150,757 1,901 - 3.B.6.b.ii - Cropland converted to Other Land 1,338 29,255 4,491 2,090 3.B.6.b.iii - Grassland converted to Other Land - 24,023 1,614 1,721 3.B.6.b.iv - Wetlands converted to Other Land 2,935 - - - The conversions of forestland to crop lands in Udonthani with the total areas of 12,332 ha resulted in the emission values of 1.62 MtCO2. While the total areas of forestland conversion in Suratthani are a similar amount to Udonthani with the total areas of 13,942 ha, but releasing the larger proportion of emission values of 3.15 MtCO2.This is mainly due to the forest type in Suratthani is the tropical rain forest which stocks the large biomass before the conversion (280 t dm ha-1) comparing to the small one in the dry forest (130 t dm ha-1) which usually found in Udonthani. Bangkok emitted the lowest CO2 among the four study areas because all of areas were used for settlement, business and industry which a small biomass stocks. It lead to released CO2 emission due to biomass change with value of 0.03 MtCO2. However, it could not summarize that Bangkok is the smallest CO2 source. To accomplish the CO2 inventory in Bangkok, it also needs to calculate an amount of emission from energy, industry, and transportation sector. With the proposed method, concentrations and distributions of CO2 sources and sinks were possibly to boundary and locate. Maps of spatial CO2 emission and removal in the four provinces were illustrated in Figure 2. 17 Figure 2: Spatial CO2 emission and removal from Land Use and Land Use Change in (a) Nan (b) Udonthani (c) Suratthani and (d) Bangkok 5. Conclusions In this work, Geographic Information System (GIS) was applied as a key tool for implementing the spatial inventory of Carbon Dioxide (CO2) emissions and removals during 2007 - 2012 resulting from the carbon stock changes in biomass for all types of land. The calculations of CO2 were based on the 2006 Intergovernmental Panel on Climate Change (IPCC) guidelines in the Agriculture, Forestry and Other Land Use (AFOLU) sector. With the performances of GIS, it allows us to map the CO2 spatial distributions and geo-reference of CO2 sources and sinks. It could be summarize from the results that Nan, Udonthani and Bangkok was identified as a CO2 emission source. In contrast, Suratthani was considered as a CO2 sink due to the large coverage of the forest land and perennial tree crop. GIS based spatial inventory is beneficial for monitoring CO2 status, and implementing more sustainable action, strategies, good practices and practical uses to be baseline emission for supporting the implementation of greenhouse gas reduction at the level of district, sub-district or boundary of municipality. However, the uses of the default emission and stock change factors for calculating CO2 emissions and removals which were not consistent with the actual carbon stock in the specific location. Thus, the practical emission factors should be available by considering to the vegetation types and environment conditions. Reference Asdrubali F., Presciutti.A., Scrucca. F., 2013, Development of a greenhouse gas accounting GIS-based tool to support local policy making-application to an Italian municipality, Energy Policy. IPCC guidelines for national greenhouse gas inventories, 2006, National greenhouse gas inventories programme, IGES, Japan. IPCC-Intergovernmental Panel on Climate Changes, 2007, Summary for policy makers, climate change 2007: The physical science basis, contribution of working group I to the fourth assessment report of the IPCC. Cambridge University, United Kingdom and New York, USA. Office of Natural Resources and Environmental Policy and Planning (ONEP), 2010, Thailand’s Second National Communication. Ministry of Science, Technology and Environment, Bangkok, Thailand. Ponce-Hernandez R., 2004, Assessing carbon stocks and modelling win-win scenarios of carbon sequestration through land-use change, FAO, Rome, Italy. Richards G.P., Brack C.L., 2004, A continental biomass stock and stock change estimation approach for Australia, Australian Forestry 67 (4), 284-288. Xu C., Wang R., Ding L., Wen Q., 2015, Study on the climate change of Shiyang river basin in Chinese arid inland area, Chemical Engineering Transactions, 46, 769-774, DOI: 10.3303/CET1546129 Sub - District Boundary Legend District Boundary CO2 Emission and Removal (tCO2) -2.5 to -1 100 to 200 -53 to -10 -10 to -5 -5 to -2.5 -1 to 0 0 to 100 200 to 273 (a) (b) Note: Symbol - = Removal 18