Data-intensive tools for effective carbon mitigation in forestry
Jorge L. Zapico (LNU), Department of Computer Science and Media Technology jorgeluis.zapico@lnu.se Rafael M. Martins (LNU), Department of Computer Science and Media Technology rafael.martins@lnu.se Johan Bergh (LNU), Department of Forestry and Wood Technology johan.bergh@lnu.se
Örjan Vorrei (Sydved) Orjan.Vorrei@sydved.se Keywords: Visualization, Sustainable Forestry, ICT4S.
1. INTRODUCTION
Sustainable forestry can be a key solution contributing to climate change mitigation (IPCC: Rogelj et al, 2018). But the carbon balance (carbon sequestered minus carbon emissions) can vary depending on many factors, including how the forest is managed by the owners and how and when it is harvested (Lundmark, Bergh et al. 2014; Canadell & Raupach, 2008). Besides that forests store large amounts of carbon in biomass and soils wood products can replace fossil fuels and greenhouse gas-intensive materials (Sathre
& O’Connor 2010, Sathre et al. 2010). Calculating a carbon balance model and simulating the impact of different decisions is a data-intensive task which can use the increasing digitalization of forestry data.
Optimizing the carbon balance of forestry through better data-intensive management (See Zou et al 2019 for data-intensive approaches in forestry) could help optimizing the climate mitigation impact of forestry in Sweden, and helping forest owners and industry in their work towards sustainable forestry. This project explored the possibilities of using data and visualization in helping forest owners to understand the carbon balance of their forest, and help them in making informed management decisions to improve the carbon capture potential of their forestry.
2. EXISTING DATA AND MODELS
One part of the project was to explore the existing data and carbon models which could provide affordances for data-driven approaches for carbon mitigation.
For calculating the carbon balance (a balance how much carbon is stored in the forest different carbon stores, how much is emitted, how much i taken when harvesting) there are a number of existing models.
Most relevant for this project was SLU’s software Heureka was used as it is developed for a Swedish
1context, but a number of other models exist internationally . Heureka can calculate and make prognosis of
21https://www.heurekaslu.se/wiki/About_Heureka
2 Such as CO2FIX
https://www.wur.nl/en/product/Carbon-balance-model-CO2FIX-downloaded-by-over-5000-people-worldwide.htm and CBM-CFS3
https://www.nrcan.gc.ca/climate-change/impacts-adaptations/climate-change-impacts-forests/carbon-accounting/car bon-budget-model/13107
the carbon balance in a forest based on different parameters such as location, size, soil, tree type composition, etc. This data can for the most part be extracted from existing forest management plans which the forest owners and forest companies have access to. Heureka and other modeling tools are on the other hand mostly tailored for research activities and not for single forest owners to use.
Forest management plans include GIS information on forest properties, dividing the forest in sectors with similar conditions which are managed as a unit and providing information such as age, tree composition and suggested actions in the near future for the forest owners. These forest management plans exist in digital format, even if access to the raw data may not be straightforward and can requires proprietary software.
With the increase in digitalization of the forest industry there is also a growing number of other datasets that could be used for calculating a forest carbon balance in more detail, this includes for example LiDAR
(Airborne Laser Scanning data) which can be used for calculating the actual volume of biomass in a
3
forest and growth rate. There is also a variety of relevant datasets and databases from for example SLU ,
4the data format and access of these databases is for the most part not straightforward for using in mashup applications.
3. PROTOTYPE
A mockup prototype of how a tool using data-intensive visualizations and simulations for supporting carbon mitigation in forestry can look like. In the prototype we used a fictive forest property placed in Småland. The carbon balance for the forest was modeled using Heureka and presents a prognosis for the duration of an average clear-cut forest process, from clear-cut, planting, management and harvest after 80 years. To be closer to reality we broke down the data in smaller sectors and moved the dates of harvest to different years to represent better a real forest with stands at different points in their growth.
The prototype showcases the possibilities of presenting the carbon balance of the forest in a simple visualization to provide an insight about the carbon stores:
1. Carbon stored in the forest (top section of the chart):
a. Biomass above ground (stemwood, branches...) b. Carbon in the soil
c. Carbon in dead wood
2. Carbon extracted from the forest (lower section of the chart) a. Stemwood
b. GROT (branches, roots and other biomass)
3 Se till ex:
https://www.lantmateriet.se/sv/Kartor-och-geografisk-information/geodataprodukter/produktlista/laserdata-skog/
4https://www.slu.se/en/Collaborative-Centres-and-Projects/the-swedish-national-forest-inventory/
The visualization starts with a view of the whole forest property and allows to select different sectors to see the specific carbon balance there. At the moment the only possibility to explore changes is selecting between a standard management program and a no-management at all.
4. POSSIBILITIES AND RISKS
This project brought together an interdisciplinary team which has created a better understanding of the possibilities and risks of data-intensive approaches for climate mitigation in forestry. From the explored existing data and models and the created prototype we can see a number of opportunities and risks.
The carbon balance of a forest is at the moment mostly calculated for research purpose in research-based
tools. The information needed for running the simulation models are not at the moment automated. So
there is a possibility for providing simplified carbon model visualization to the forest owners based on
existing data in automated ways. These simulations may not be needed to be as exact as research oriented
ones, but still based on solid data and assumptions. So the access to precise information of volume such as
with LiDAR data, while interesting, may not be required for such an application. The precision needed
could be explored and the variability calculated with sensitivity analysis.
An important point that this project at the moment has not yet explored is how this information needs to fit the goals and needs of forest owners and (or) the forest industry and which are the drivers, and barriers for forest owners in their work affecting the carbon balance. From the current project we could see that, at the moment, maximizing the carbon balance of the forest is a win-win situation as it mostly means increasing the growth and volume of biomass (meaning more carbon uptake), so the barriers may not be as limiting as expected as when pro-environmental behavior clashes with economic interests.
Such a tool as the one being explored in this project has two main objectives:
- One is providing insight about the current situation, the size of the carbon stores and the
distribution, uptake and emission of carbon through time. The goal is to provide more knowledge and understanding of how the forest work as a carbon storage and how different actions such as harvesting affect the result.
- The second is using this created insight for triggering better management decisions.
A key aspect that needs to is the actionability of the data, meaning what possibilities are provided to the users so they can do something with the information. A main idea that our prototype wanted to showcase is the manipulation of the data by changing different parameters. At the moment this is limited to
management/no management due to using an external model, but there are possibilities of allowing the users to play with and experiment with different options and parameters to explore how different decisions would affect their forest and the carbon balance. This can be useful not only in the decision making, but also for increasing the learning potential about the different connections in the carbon balance of the forest.
An interesting point that the prototype explored but that could be extended is communicating the carbon storage of the wood products. When harvesting timber and wood products, the carbon captured in the trees does not disappear, but it is moved from a carbon store to another, in our prototype this is communicated by having the extraction in a continuum under the forest carbon stores so it adds to the total. The extracted carbon will have different pathways, the volume becoming timber will store the carbon during a building life-cycle, the share becoming paper and other pulp products will store the carbon while in use and recycle, the share becoming biomass for energy will go back to the atmosphere.
But there is also a substitution effect (Sathre and O’Connor, 2010), as the energy from biomass may be substituting fossil fuel, and the timber may be substituting concrete in construction, and in that way not only having embedded carbon but also avoiding other carbon emissions. This part of the visualization has clear possibilities and should be improved upon. The communication of substitution effects would need a clear communication effort to help the users understanding the second order effects in which other types of visualizations could be used, for example comparing alternative scenarios.
In this project we could also see some risks. First, while the relation between maximizing growth and the carbon balance is positive, there are other environmental aspects which are not present such as
biodiversity or eutrophication, so there is a risk of using the optimization of capturing more carbon as a
way of justifying the optimization of growth/profit and overrunning other environmental aspects. This is
always a risk with data and implicit focus on quantitative properties against qualitative ones in data
applications (Zapico, 2014). The creation of the models and actions in the tools can also create bias
towards specific management goals which may not be transparent to the user. To avoid these risks there is a need of both increasing the transparency of the data and assumptions, anchoring in solid research, and good communication to the users. The inclusion of other environmental aspects and goals in the
visualization could also be an interesting path.
4. CONCLUSION AND FURTHER WORK
This SEED project provided the opportunity to explore the topic of using visualization and data-driven approaches for helping the understanding and optimizing of forest’s carbon balance. From the created prototype we could see some clear opportunities, as carbon balance models have been mostly research oriented, and simplified models could help in increasing forest owners understanding and guiding their management. We could also identify some risks that would need to be taken into consideration in further efforts. The closest next step after this exploration are:
1. Based on the identified possibilities explore the most promising areas together with Sydved to match the existing technological affordances with the needs from the forest owners and/or the forestry industry.
2. Based on this, write an external funding proposal with the aim of developing the technology and testing a functional prototype with real data and users. A possibility is the coming Formas call which focuses on the communication of research results to stakeholders .
5BUDGET
The budget for salaries has been used as planned. The budget for travel and activities has not been used.
REFERENCES
Canadell, J. G., & Raupach, M. R. (2008). Managing forests for climate change mitigation. science, 320(5882), 1456-1457. Lundmark, T., Bergh, J., Hofer, P., Lundström, A., Nordin, A., Poudel, B., ... &
Werner, F. (2014). Potential roles of Swedish
forestry in the context of climate change mitigation. Forests, 5(4), 557-578.
Rogelj, J., D. Shindell, K. Jiang, S. Fifita, P. Forster, V. Ginzburg, C. Handa, H. Kheshgi, S. Kobayashi, E. Kriegler, L.
Mundaca, R. Séférian, and M.V.Vilariño (2018) Mitigation Pathways Compatible with 1.5°C in the Context of Sustainable Development. IPCC.
Sathre, R. and O’Connor, J. (2010) Meta-analysis of greenhouse gas displacement factors of wood product substitution. Environmental Science and Policy, 13(2): 104-114.
5 Formas kommunikationsutlysning 2020
https://formas.se/arkiv/alla-utlysningar/utlysningar/2019-10-07-formas-kommunikationsutlysning-2020.html
