Aarhus University Seal / Aarhus Universitets segl

No. 155: The Danish SINKs project. Final report on the Danish monitoring project for Land Use, Land Use Change and Forestry under the Kyoto Protocol

Steen Gyldenkærne (scientific coordinator), Pia Frederiksen (project coordinator) Eds. 2015. The Danish SINKs project. Final report on the Danish monitoring project for Land Use, Land Use Change and Forestry under the Kyoto Protocol. Aarhus University, DCE – Danish Centre for Environment and Energy, 111 pp. Scientific Report from DCE – Danish Centre for Environment and Energy No. 155



This report describes the final outcome of the Danish SINKs project. The SINKs project was initiated in 2007 to develop methods for fulfilling the needed documentation for the Danish ratification of the Kyoto protocol according to article 3.3 and the subsidiary voluntary election of Forest Management, Cropland Management and Grassland Management under article 3.4. Article 3.3 covers the mandatory estimation of afforestation and deforestation since 1990 and the related carbon stock changes in the areas involved. Forest Management under Article 3.4 covers land which was forested before 1990. Cropland and Grassland Management covers the area under agricultural management. Because of the complex land use in Denmark and difficulties to differentiate cropland and grassland both Cropland and Grassland Management was elected under Article 3.4.

In total 13 different projects were initiated covering forest and agricultural issues. The outcome from the different projects is documented in the scientific literature, in reports and working papers and in the Danish National Inventory Report (NIR) for greenhouse gas emissions (Nielsen et al. 2010, 2011, 2012, 2013, 2014).

In this report the outcome of the project is split in three main chapters, each covering one main topic. Chapter 5 covers the development of the method for producing a Danish land use change matrix (LUM), Chapter 6 is about forest issues and Chapter 7 is about agricultural issues. This structure assists the understanding of how the different sub-projects contribute to the overall objectives.

The LUM was initially planned to be based on a previous Global Monitoring for Environment and Security (GMES) remote sensing project for the Danish forests combined with auxiliary vector map information. During the project it was realised that the uncertainty related to land use change detection based on remote sensing was high and not suitable for the purpose under the complex land-use/land cover conditions present in Denmark. Because of a continuous development in Danish geo-referenced thematic data it was decided to construct a land use/land cover map for 2011 primarily based on digital thematic maps and only where no other information was available to use remotely sensed information, as for the forested area. Two other land use maps were constructed for 2005 and 1990 respectively based on the best available information and a rule based decision tree. The LUM for the years between these three points in time was based on linear interpolation.

The LUM for 2011 to 2012 and future LUM updates are based on annual updates of the different thematic maps. Some of the vector maps, especially from the Danish Geodata Agency are only updated every fourth year. This creates a time lag for some of the data layers but is found to be the best solution for construction of a suitable LUM for the purpose.

The land use/land cover analysis showed that the forest area in 1990 was 543 973 ha. From 1990 to 2012 the afforested area has been estimated to 94 358 ha and the deforested area to 5 784 ha. In this period the area with settlements and infrastructure increased from 485 543 ha to 511 411 ha. New settlements are primarily located on agricultural land. Analyses of the data from the agricultural subsidy program has showed that there is a large annual transition between the land use categories Cropland and Grassland and vice versa indicating that the Danish election of both Cropland Management and Grassland Management was a robust decision.

Forest issues included the analysis of biomass expansion factors for forest species, and soil carbon stocks in forest soils. National biomass expansion factors (BEF) was estimated and implemented in the Danish GHG inventory for beech and norway spruce. These are slightly different from the default values in the Intergovernmental Panel on Climate Change (IPCC) 2003 guidelines (IPCC, 2003).

Prior to the SINKs project there was very little information on the carbon stock in the Danish forest soils. The project has estimated carbon stock changes in 110 paired forest soil samples from 1987 to 2009 and in 22 plots converted to forest during 1990 to 2007. Furthermore, carbon stock measurements have been conducted in 278 plots in the national forest inventory (NFI) for a further baseline on the carbon stock changes. The results have shown that the forest soils were not likely a source for carbon dioxide (CO2) in Denmark during the period 1990 to 2012.

If forest soils are drained they may emit CO2 and nitrous oxide (N2O). The current Danish forest policy is to create more natural forests and therefore ditches in the national forests are only maintained to a very low degree or even removed, and as a consequence the water level has increased in many areas. It has been estimated that 336 264 ha of the area was drained in 1990 which was reduced to only 281 567 ha in 2011.

Soil sampling in the agricultural soil sampling grid was made in 2009. The network was previously sampled in 1987 and in 1998. In total 464 paired plots were included in the analysis. In general, the lighter sandy soils showed an increase in the carbon stock (0-100 cm) whereas a decrease in the carbon stock was found on the clay soils. This development is explained with a high cattle and grass density, which increase the organic matter input to the sandy soils and the absence of cattle and grass combined with a frequent removal of straw for energy purposes on the clay soils.

In the reporting to the United Nations Framework Convention on Climate Change (UNFCCC) a dynamical model (C-TOOL) is used to estimate the organic matter turnover and the carbon stock change. The output from C-TOOL is similar to what was measured in the agricultural soil sampling grid on clay soils and slightly lower than measurements on sandy soils. This indicates that C-TOOL is suitable for estimating the carbon stock changes. A new parameterization of C-TOOL with new knowledge is currently taking place. The new version is expected to be implemented in near future.

A new map of the agricultural organic soils has been developed. The new map showed that there were 70 300 hectares of organic soils under agricultural influence in 2010 with an organic matter (OM) content of >20 % OM. Of this, 42 000 hectares were cultivated with annual crops and 28000 with perennial grass. The total area is much lower than expected. In 1975 it was estimated that 243000 hectares were organic according to the Danish classification (>10 % OM). The very high disappearance rate is explained by the very shallow and young organic soils in Denmark.

CO2, methane (CH4) and N2O emissions were measured in eight fields on three locations and with different crops. The measured emissions showed a high variability. Based on the measurements it was only possible to distinguish between annual crops in rotation and permanent grass. The measured emissions were up scaled to annual emission factors. The CO2 emission factors are lower than the default IPCC values for temperate regions (IPCC, 2003) and higher than for cold conditions. Denmark is located in a temperature zone between cold and temperate and therefore the measurements were found suitable for Danish conditions and incorporated in the Danish GHG inventory. The measured N2O emissions showed a very high variability, with some values much higher than the default values. The reason may be that one of the plots showed to have very special soil conditions favouring N2O emissions. The measured N2O values are therefore not included in the Danish GHG inventory, where the default values from IPCC (IPCC, 2003) are used. The measured CH4 emissions are not yet included in the inventory.

The total area with hedgerows and small biotopes not classified as forests were estimated for 1990 and 2005 and a model for the carbon stock in the hedgerows has been developed. It has been estimated that the total area of hedgerows has decreased with 2.0 % but the total volume has increased with 6.4 % and the total carbon stock has increased with 14.2 % in that period. These changes are explained with a large replacement of single rowed conifer hedgerows on the sandy soils with 3- or 6-rowed broadleaved hedgerows. These figures are subject to high uncertainty and they will be investigated further in the near future.

In total LULUCF activities from Afforestation have added 184 Gg CO2 eqv to the Danish reduction commitments in the first commitment period. Deforestation has contributed negatively with 439 Gg CO2 eqv, Forest Management added 1 172 Gg CO2 eqv, Cropland Management 8 250 Gg CO2 eqv and Grassland Management contributed negatively with 552 Gg CO2 eqv. This makes up a contribution of 8 614 Gg CO2 eqv to the reduction commitment in total.

The results from the SINKs project have improved the Danish GHG inventory from LULUCF considerably and resulted in an improved understanding of the carbon dynamics in the Danish landscape.