Indirect emissions of forest bioenergy: detailed modeling of stump-root systems

Transcription

1 GCB Bioenergy (2014) 6, , doi: /gcbb TECHNICAL ADVANCE Indirect emissions of forest bioenergy: detailed modeling of stump-root systems JARI LISKI*, SANNA KAASALAINEN, PASI RAUMONEN, ANU AKUJ ARVI*, ANSSI KROOKS, ANNA REPO* and MIKKO KAASALAINEN *Finnish Environment Institute, Natural Environment Centre, Ecosystem Change, PO Box 140, Helsinki, FI, 00251, Finland, Department of Remote Sensing and Photogrammetry, Finnish Geodetic Institute, Geodeetinrinne 2, P.O. Box 15, Masala, FI, 02431, Finland, Department of Mathematics, Tampere University of Technology, PO Box 553, Tampere, FI, 33101, Finland Abstract Indirect carbon dioxide emissions from producing bioenergy from tree stumps and roots depend critically on the decomposition rate of these harvest residues if they were left in forest to decay. We developed a method to improve the current estimates of these emissions. First, the 3D structure of uprooted stump-root systems was modeled based on terrestrial laser-scanning data. Second, information obtained on the size distribution of the stumps and the roots was used to simulate their decomposition and to estimate the indirect emissions. The method was able to describe the structure of stump-root systems at a clear-cut boreal Norway spruce test site. Compared with earlier results based on the diameter of stumps alone, the new estimates of the decomposition rate were slightly higher and, consequently, those of the indirect emissions slightly lower. The method is useful to collect information on the indirect emissions of stump-root bioenergy quickly in different forests. Keywords: bioenergy, carbon, decomposition, forest, harvest residue, indirect emissions, modeling, roots, stump, terrestrial laser scanning Received 20 March 2013 and accepted 1 April 2013 Introduction Interest in using tree stumps and roots for energy production is increasing in Northern Europe (e.g., Eriksson & Gustavsson, 2008) and North America (Hannam, 2012). For example in Finland, the annual land area where the stump-root systems are collected after forest felling has expanded from nearly none in 2000 to ha in 2010 (Finnish Forest Research Institute, 2011). Over the same period, the annual harvest volume of the Finnish stumps and roots has grown to more than 1Mm 3 (Finnish Forest Research Institute, 2011). The stump-root systems are favorable raw material of bioenergy for technical and economic reasons (e.g., Walmsley & Godbold, 2010). In addition, high targets of renewable energy in Europe create pressure to intensify the use of residual forest biomass, such as stumps and roots (UNECE/FAO, 2008; The European Parliament & The Council, 2009; Beurskens & Hekkenberg, 2011). Bioenergy production from stumps and roots is controversial from a climate perspective (Hope, 2007; Zabowski et al., 2008; Jarvis et al., 2009; Melin et al., 2010; Lindholm et al., 2011; Repo et al., 2011; Repo et al., 2012). Correspondence: Jari Liski, tel , fax , jari.liski@ymparisto.fi This biomass is renewable but using it for energy production may cause relatively high indirect carbon dioxide emissions. These emissions result from losses of forest carbon namely the dead stumps and roots that would eventually become a part of soil organic carbon. In particular, stumps decompose slowly and retain carbon for a long time if left in forest (Melin et al., 2009; Palviainen et al., 2010; Shorova et al., 2012). When they are removed and combusted, the carbon is released to the atmosphere at once. The use of less recalcitrant residual biomass, such as branches or small-sized thinning wood, has a smaller and more temporary effect on the forest carbon balance (Repo et al., 2011; Repo et al., 2012). This is because these components decompose more rapidly and release carbon relatively quickly even in forest. The indirect emissions from using stumps and roots for energy production depend critically on their decomposition rate in forest. An important factor affecting the decomposition rate of any woody litter is the size of the fragments decomposing (Harmon et al., 1986; Tuomi et al., 2011a). Thus, to estimate reliably the indirect emissions from using stump-root systems for energy production, it is necessary to know their size distribution. The size distribution of the stump-root systems is difficult to measure manually because the work is slow and hard and these systems take often very irregular 2013 John Wiley & Sons Ltd 777

2 778 J. LISKI et al. and complicated shape (e.g., Laitakari, 1929; Danjon & Reubens, 2008; Kalliokoski et al., 2008). Terrestrial laser scanning (TLS) provides an effective and low-cost in situ approach to measure tree structure. Information on the performance and data accuracy of TLS is constantly increasing as is the number of new applications in environmental studies. TLS has proven to be a useful tool to estimate tree biomass and volume (Koch, 2010). Today, TLS can also be used to collect accurate information on the 3D structure of trees (C^ote et al., 2009), but this requires complementing of the TLS method with efficient computational tools designed for the purpose (Raumonen et al., 2013). So far, most of these computational methods have focused on tree trunks in the framework of traditional forest resource inventory (e.g., Pfeifer et al., 2004; Liang et al., 2012). Applications of the TLS method to study the structure of belowground parts of trees are still quite rare (Danjon & Reubens, 2008). In this study, we describe, first, a TLS-based method to characterize the 3D structure of stump-root systems lifted from soil for energy production. Then, we show, as an example, how this information can be used to simulate the decomposition of these stump-root systems and to estimate the indirect emissions caused using these systems for energy production. Material and methods Stump-root systems studied A forest company lifted six stump-root systems for this study as a part of their ordinary biomass collection practice at a clear-cut Norway spruce site in southern Finland. The only difference compared with the normal practice was that the systems studied were not cut to pieces. Thus, they represented exactly those parts of the stump-root systems that would be decomposing in soil at the site if they were not harvested to produce bioenergy. Laser scanning the stump-root systems The stump-root systems were measured using a stationary TLS, Leica HD6100. The scanner operation is based on phase modulation with three different carrier wavelengths. The unambiguity range of the distance measurement is approximately 79 m and accuracy 2 5 mm with a field of view of The angular resolution can be selected between and The scanner output is a nm laser beam with a 3 mm circular diameter and 0.22 mrad divergence. The stump-root systems studied were scanned from different directions to cover the entire structure (Fig. 1). The scans were registered to a common coordinate system based on spherical reference targets positioned at the study site. Thus, it was possible to form a single three-dimensional point cloud from the surface of each stump-root system. Fig. 1 Laser scanning of stump-root systems. The scans were taken from different directions to cover the systems as a whole. Three spherical reference targets were used to register the scans to a common coordinate system. TLS data processing and modeling the structure of stump-root systems To model the structure of the stump-root systems, we assumed that the point cloud produced by the laser scanner formed a comprehensive sample of the system surface. These point clouds contain usually also points from outside this surface. These points of noise form typically small separate clusters or have a lower point density than measurements from the actual roots or stump do. These clusters may fill the space between the actual neighboring roots randomly, and it may seem that the roots are connected to each other. To correct this error, we filtered out points which either had only one or two neighbors within 1 cm radius or formed separate small clusters. After this, we built a model of each stump-root system using a method similar to that developed for describing the structure of tree biomass aboveground (Raumonen et al., 2013). In this method, the point cloud is first divided into cover sets corresponding to small patches of the stump-root surface. The actual surface of the stump-root systems is then modeled using geometric characterizations and information on the neighboring cover sets. The main difference from the method applied for modeling the aboveground structure of trees was the selection of the starting point. The modeling of the aboveground parts starts from recognizing the base of the trunk, whereas we started modeling the stump-root systems from identifying the cutting surface of the stump. Then, we expanded away from the cutting surface by adding neighboring cover sets. This stepby-step procedure separated the stump from the roots. Stumps are not usually geometrically cylindrical and thus we did not model them with geometric cylinders. However, the volume enclosed by the surface of a stump can be approximated with a polyhedral grid that has topological cylinder support. The stump is divided into angular sectors (36 in this case) according to the axis that goes through the center of the cutting surface and is normal to it (Fig. 2). Then, each sector that has enough height is approximated with grid points that are means of points forming vertical layers (about 5 cm in height). This way the stump is modeled as an assembly of polyhedrons with

3 INDIRECT EMISSIONS OF STUMP-ROOT BIOENERGY 779 (a) (b) (c) Fig. 2 Modeling of a stump-root system using grid points on the stump surface. The blue points form a sector on the surface (a). Each sector is divided into smaller layers (red points) having a grid point (green circle) as a mean (b). The blue circles represent the grid points on the data-covered surface whereas the red circles show the grid points interpolated based on the blue circles (c). four facets (one quadrilateral and three triangles). The volumes of these polyhedrons are easy to calculate, and for each layer of polyhedrons (36 per layer) we get the minimum and maximum diameter. If there were gaps in the point cover of the stump such that some grid points could not be determined, we interpolated the grid points for each layer as follows. We used the distances r b and r e of the boundary grid points (points between which there were missing grid points) from the center of the layer and linearly interpolated distances r i for the missing n grid points or r i =r b + i/n 9 (r e -r b ) held. The interpolated grid points are with these distances from the center so that they are spaced with equal angles (Fig. 2). After separating the stump from the roots, we segmented the rest of the point cloud so that each segment corresponded to a part of a root or a whole root. Each segment was thus a connected and nonbifurcated part of the point cloud. Then, we estimated each segment (root) by a collection of cylinders approximating the local orientation and diameter of the segment. The resulting cylinder model can be used to approximate the size distribution of the parts forming the stump-root system and its topological branching structure. The biomass of the stump-root systems was calculated based on the volume of the cylinders by assuming a dry mass density equal to kg dm 3 for each cylinder (Palviainen et al., 2010). For comparison, we estimated the biomass also based on the diameter of the cutting surface using allometric equations (Repola, 2009). Decomposition and indirect emissions of the stumproot systems The decomposition of the stump-root systems was simulated using the user interface of the Yasso07 litter decomposition and soil carbon model (Tuomi et al., 2011b). In this model, the decomposition of woody litter is affected by the diameter of the woody fragments (Tuomi et al., 2011a). Other factors controlling the decomposition are the chemical composition of the woody litter and climate. To simulate the decomposition of the stump-root systems, their biomass was divided into 1 cm diameter classes according to the models fitted to the laser-scanning data. The biomass was assumed to comprise 1% water-soluble compounds, 1% ethanol-soluble compounds, 68% acid-hydrolysable compounds, and 30% nonsoluble and nonhydrolysable residue (Repo et al., 2012). Annual mean temperature at the study site was 3.7 degrees C, maximum difference in the monthly mean temperature between months 12.1 degrees C and annual precipitation 585 mm. The indirect carbon dioxide emissions from harvesting the stump-root systems and using them for bioenergy production were calculated for a scenario in which this practice was started and continued at a constant level for a 100-year period (Repo et al., 2011; Repo et al., 2012). The heating value of the biomass was taken to be equal to 11.6 MJ kg 1 and the moisture content 35% (v/v) (see Repo et al., 2012). Results Evaluation of the models of the stump-root systems The models of the stumps and the roots covered the majority of the laser-scanning data collected from the stump-root systems (Fig. 3). However, some data were found outside the modeled structure of the stumps, especially stumps b, d, and e. The models did not also capture the structure of small roots in every detail. The estimates for the total biomass of the stump-root systems based on the modeled structure (Fig. 3) were quite similar to these estimates calculated using the allometric equations (Fig. 4). The structure-based estimates of the stump biomass were higher than the allometrybased ones, but the structure-based estimates of the root biomass were lower. These differences were probably caused by different definitions of stumps and roots in the methods. The similarity of the total biomass

4 780 J. LISKI et al. (a) (b) (c) Fig. 3 study. Laser-scanning data (left column) and the models fitted to these data (right column) for stump-root systems a to f of this estimates is noteworthy because the allometric equations represent the entire stump-root systems excavated carefully for research purposes whereas the stump-root systems we studied were drawn from soil by an ordinary stump harvest machine. Based on the visual comparison of the modeled stump-root structure to the laser-scanning data (Fig. 3) and the similarity of the biomass estimates to those obtained using the allometric equations (Fig. 4), we concluded that the new method described the majority of

5 INDIRECT EMISSIONS OF STUMP-ROOT BIOENERGY 781 (d) (e) (f) Fig. 3 Continued the structure of the stump-root systems. Consequently, it was appropriate to use the results and investigate the structure of the stump-root systems and the implications for bioenergy production in more detail. Characteristics of the stump-root systems The dry weight of the stump-root systems varied from 9 to 92 kg (Fig. 4) and the diameter of the cutting surface from 21 to 37 cm. The stumps represented 36 65% of the total dry weight depending on the stump-root system. The median diameter of the biomass varied from 11 to 37 cm among the stump-root systems (Fig. 5). The maximum diameter, found in the bottoms of the stumps (Fig. 3), ranged from 36 to 82 cm. Roots thinner than 1 cm represented up to 2% of the total biomass and those thinner than 2 cm up to 10%. The diameter of the

6 782 J. LISKI et al. Allometric model (kg) Stump Roots Sum This study (kg) Fig. 4 Biomass of the stumps, the roots, and the entire stumproot systems according to the method developed in this study (x-axis) and allometric equations (y-axis). The lines represent linear regressions with the intercept equal to zero, stump biomass y = 0.50x (R 2 = 0.60), root biomass y = 1.43x (R 2 = 0.16), and the biomass of the stump-root systems y = 1.03x (R 2 = 0.47). smallest roots detected and modeled varied between 0.3 and 0.5 cm between the stump-root systems. Decomposition of the stump-root systems and indirect emissions of bioenergy The detailed information on the size distribution (Fig. 5) was used to simulate decomposition of the stump-root systems and to estimate indirect carbon dioxide emissions caused by producing bioenergy from these systems. The differences in the size distribution of the biomass caused variability in the simulated decomposition rates. Stump e rich in large-sized biomass decomposed at the lowest rate, whereas stumps c and f comprising smaller sized biomass decomposed at the highest rate (Figs 5 and 6). These small-sized and large-sized stump-root systems had, respectively, 52 or 63% of the original mass remaining after 20 years of decomposition (Fig. 6). After 50 years, these figures were 33 or 43%, and, after 100 years, 23 or 30%. These differences in the decomposition rate affected the estimates of indirect carbon dioxide emissions from producing bioenergy from the stump-root systems. We calculated these emissions for a scenario in which the bioenergy production was started and it was continued at the same level for a 100-year period. These emissions varied among the stump-root systems according to the decomposition rate; after 20 years of bioenergy production, they ranged from 71 to 81 g CO 2 MJ 1, after 50 years from 53 to 64 g CO 2 MJ 1, and after 100 years from 40 to 50 g CO 2 MJ 1. Cumulative frequency (relative unit) Discussion 1 0,75 0,5 0, Diameter (cm) Fig. 5 The biomass of the stump-root systems a to f (see Fig. 3) divided into diameter classes. The graph shows the cumulative frequency distribution of diameter. The method based on TLS and modeling described the majority of the structure of the stump-root systems studied. The comparison to biomass estimates calculated using the allometric equations (Repola, 2009) supported the validity of this method. However, some TLS data were found outside the modeled structure of some stumps and small roots. This indicated that the modeling did not capture these structures completely. We think that a major reason for this shortcoming was that these parts of the stump-root systems were not exposed clearly enough when they were laser scanned. Positioning the stump-root systems more carefully for Mass remaining (relative unit) 1 0,8 0,6 0,4 0, Time (year) Fig. 6 Mass remaining of stump-root systems a to f (see Fig. 2) over a 100-year decomposition period. Decomposition was simulated using the Yasso07 model and the diameter distribution (see Fig. 5) as input to this model. a b c d e f a b c d e f

7 INDIRECT EMISSIONS OF STUMP-ROOT BIOENERGY 783 Indirect CO 2 emissions (g MJ 1 ) Time (year) Fig. 7 Indirect carbon dioxide emissions from a continued annual practice of using stump-root systems a to f of this study (see Fig. 2) for bioenergy production. laser scanning and cleaning them from soil material can solve this problem. The method developed makes it possible to collect a lot of new data on stump-root structure quite quickly compared with traditional manual methods. It takes some minutes to lift a stump-root system from soil using a stump harvest machine, and a couple of hours in total to laser scan the system, analyze the data, and fit the stump and root models to them. Knowledge on the size distribution of stump-root systems is crucial to the decomposition rate estimates of these systems (Tuomi et al., 2011a) and, consequently, the estimates of indirect emissions caused using the stumps and roots for bioenergy production (Repo et al., 2011; Repo et al., 2012). To demonstrate this importance and show an example of using this knowledge, we simulated the decomposition the stump-root systems studied using Yasso07 soil carbon model (Tuomi et al., 2011b) and calculated the indirect emissions based on the results (see Repo et al., 2012). The decomposition was simulated in this study using the detailed information on the size distribution, whereas the diameter of the cutting surface has been used to approximate the average size in similar simulations earlier. The present decomposition estimates were slightly faster. For example, the stump-root systems of this study had 52 to 63% of their original mass remaining after 20 years of decomposition whereas an earlier estimate for stumps in southern Finland was 64% (Repo et al., 2012). Consequently, the present estimates of the indirect emissions were slightly lower than the earlier ones namely g CO 2 MJ 1 after 20 years of constant use compared with an earlier estimate equal to 92 g CO 2 MJ 1 (Repo et al., 2012). a b c d e f The use of stump-root systems for bioenergy production has been started and expanded without proper knowledge on the climate effects of this practice (e.g., Hope, 2007; Lindholm et al., 2011; Hannam, 2012). The majority of the climate effects are caused by the indirect carbon dioxide emissions resulting from changes in the carbon balance of forest namely a decreased carbon stock of stumps and roots in forest, a decreased carbon input to soil and eventually a decreased carbon stock of soil (Palosuo et al., 2001; Repo et al., 2011; Repo et al., 2012). The method presented in this study can be used to make more reliable estimates of these indirect emissions. This information is useful for planning and making decisions on the volumes and methods of lifting stump-root systems for bioenergy production. Acknowledgements This study was funded by the Ministry of Environment in Finland, the Maj and Tor Nessling Foundation (project Climate impacts of forest bioenergy ) and the Academy of Finland (projects /218144, ). We thank UPM- Kymmene for making the stump-root systems available for this study. References Beurskens L, Hekkenberg M (2011) Renewable Energy Projections as Published in the National Renewable Energy Action Plans of the European Member States. ECN-E , Energy Reseach Centre of the Netherlands and European Environment Agency, The Netherlands. C^ote J-F, Widlowski J-L, Fournier RA, Verstraete MM (2009) The structural and radiative consistency of three-dimensional tree reconstructions from terrestrial lidar. Remote Sensing of Environment, 113, Danjon F, Reubens B (2008) Assessing and analyzing 3D architecture of woody root systems, a review of methods and applications in tree and soil stability, resource acquisition and allocation. Plant and Soil, 303, Eriksson LM, Gustavsson L (2008) Biofuels from stumps and small roundwood Costs and CO 2 benefits. Biomass and Bioenergy, 32, Finnish Forest Research Institute (2011) Finnish Statistical Yearbook of Forestry. Finnish Forest Research Institute, Sastamala. Hannam K (2012) The use of stumps for Biomass in British Columbia a problem analysis. Prov. B.C. Victoria, B.C. Tech. Rep Harmon ME, Franklin JF, Swanson FJ et al. (1986) Ecology of coarse woody debris in temperate ecosystems. Advances in Ecological Research, 15, Hope GD (2007) Changes in soil properties, tree growth, and nutrition over a period of 10 years after stump removal and scarification on moderately coarse soils in interior British Columbia. Forest Ecology and Management, 242, Jarvis PG, Clement RJ, Grace J, Smith KA (2009) The role of forests in the capture and exchange of energy and greenhouse gases. In: Combating Climate Change A Role for UK Forests. An Assessment of the Potential of the UK s Trees and Woodlands to Mitigate and Adapt to Climate Change, (eds Read DJ, Freer-Smith PH, Morison JIL, Hanley N, West CC, Snowdon P), pp TSO, Edinburgh. Kalliokoski T, Nygren P, Siev anen R (2008) Coarse root architecture of three boreal tree species growing in mixed stands. Silva Fennica, 42, Koch B (2010) Status and future of laser scanning, synthetic aperture radar and hyperspectral remote sensing data for forest biomass assessment. ISPRS Journal of Photogrammetry and Remote Sensing, 65, Laitakari E (1929) The root system of pine (Pinus sylvesteris), a morphological investigation. Acta Forestalia Fennica, 33, Liang X, Litkey P, Hyypp a J, Kaartinen H, Vastaranta M, Holopainen M (2012) Automatic stem mapping using single-scan terrestrial laser scanning. IEEE Transactions of Geoscience and Remote Sensing, 50,

8 784 J. LISKI et al. Lindholm E-L, Stendahl J, Berg S, Hansson P-A (2011) Greenhouse gas balance of harvesting stumps and logging residues for energy in Sweden. Scandinavian Journal of Forest Research, 26, Melin Y, Petersson H, Nordfjell T (2009) Decomposition of stump and root systems of Norway spruce in Sweden - A modelling approach. Forest Ecology and Management, 257, Melin Y, Petersson H, Egnell G (2010) Assessing carbon balance trade-offs between bioenergy and carbon sequestration of stumps at varying time scales and harvest intensities. Forest Ecology and Management, 260, Palosuo T, Wihersaari M, Liski J (2001) Net greenhouse gas emissions due to energy use of forest residues - impact on soil carbon balance. European Forest Institute Proceedings, 39, Palviainen M, Finer L, Laiho R, Shorohova E, Kapitsa E, Vanha-Majamaa I (2010) Carbon and nitrogen release from decomposing Scots pine, Norway spruce and silver birch stumps. Forest Ecology and Management, 259, Pfeifer N, Gorte B, Winterhalder D (2004) Automatic reconstruction of single trees from terrestrial laser scanner data. International Archives of Photogrammetry. Remote Sensing and Spatial Information Sciences, 35, Raumonen P, Kaasalainen M, Akerblom M et al. (2013) Fast automatic precision tree models from terrestrial laser scanner data. Remote Sensing, 5, Repo A, Tuomi M, Liski J (2011) Indirect carbon dioxide emissions from producing bioenergy from forest harvest residues. GCB Bioenergy, 3, Repo A, K ank anen R, Tuovinen J-P, Antikainen R, Tuomi M, Vanhala P, Liski J (2012) Forest bioenergy climate impact can be improved by allocating forest residue removal. GCB Bioenergy, 4, Repola J (2009) Biomass equations for Scots pine and Norway spruce in Finland. Silva Fennica, 43, Shorova E, Ignatyeva O, Kapitsa E, Kauhanen H, Kuznetsov A, Vanha-Majamaa I (2012) Stump decomposition rates after clear-felling with and without prescribed burning in southern and northern boreal forests in Finland. Forest Ecology and Management, 263, The European Parliament and The Council (2009) Directive 2009/28/EC of The European Parliament and of the Council of 23 April 2009 on the promotion of the use of energy from renewable sources and amending and subsequently repealing Directives 2001/77/EC and 2003/30/EC. Tuomi M, Laiho R, Repo A, Liski J (2011a) Wood decomposition model for boreal forests. Ecological Modelling, 222, Tuomi M, Rasinm aki J, Repo A, Vanhala P, Liski J (2011b) Soil carbon model Yasso07 graphical user interface. Environmental Modelling & Software, 26, UNECE/FAO (2008) Potential Sustainable Wood Supply in Europe. UNECE/FAO Timber Section, Geneva. Walmsley JD, Godbold DL (2010) Stump harvesting for bioenergy - a review of the environmental impacts. Forestry, 83, Zabowski D, Chambreau D, Rotramel N, Thies WG (2008) Long-term effects of stump removal to control root rot on forest soil bulk density, soil carbon and nitrogen content. Forest Ecology and Management, 255,