Tier 3 modelling of carbon stock change in cropland mineral soil

Norway submits the national greenhouse gas inventory report under the UNFCCC each year, which forms the basis for accounting under the bilateral agreement with EU and Norway’s commitments under the Paris Agreement. Improved methodologies will improve the quality and accuracy of the numbers submitted. NIBIO is part of the National Inventory System, together with The Norwegian Environment Agency and Statistics Norway (SSB). NIBIO’s responsibility is the compiling of data and reporting of emissions from the Land Use, Land-Use Change and Forestry (LULUCF) sector, work which is led by the Department of Forest and Climate.

Carbon stock changes (CSC) is estimated for the different carbon pools, and for the different land-use categories. In this project we target the mineral soil pool in croplands. The aim is both to develop a methodology for a Tier 3 approach (spatial-temporal modelling of emissions) and provide data to verify the implementation and use of such an approach in the national greenhouse gas inventory.

Cropland definition

Cropland is defined in the greenhouse gas inventory as lands that are annually cropped and regularly cultivated and ploughed. Both annual and perennial crops are grown. It also encompasses grass leys that are in rotations with annual crops, which may include temporarily grazed fields that are regularly cultivated. This category includes arable land that was previously annually cropped and regularly ploughed but has since been abandoned. These areas remain in the cropland category until they have a regrowth of trees that make them unsuitable for ploughing. All cropland is considered managed.

The project target cropland remaining cropland, which is lands that has been cropland for at least 20 years. Emissions and removals for land use changes to and from cropland is estimated with other methodologies.

Phase 1: Model selection, calibration, and evaluation

The project has tested different models and has concluded on the ICBM model: The Introductory Carbon Balance Model for Exploration of Soil Carbon Balances. This is a model and approach also used e.g., in the Swedish GHG inventory.

Phase 2: Identify model inputs and develop the methodology for national scale

Currently the National inventory report (NIR), uses a Tier2 methodology involving the use of ICBM at steady state, for CSC in cropland mineral soil.

In order to develop Tier3 version of ICBM at national scale we implemented the following steps:

In ICBM the effect of temperature and moisture on Soil Organic Matter (SOM) decomposition is controlled by a rate modifier parameter, the external response factor, which is a function of soil type, climatic conditions, and the vegetation cover. To take into consideration the diversity of conditions presents in Norway as well as the spatiotemporal variation of climatic conditions, we set up a hydrological model that enable to compute the external response factor daily at commune level using weather data interpolated on a 1 by 1 km grid, for different combination of soil type and crop type.

We introduced a small model that compute the carbon input to ICBM as a function of annual crop yield. Previously in the Tier2 version of ICBM, the same set of fixed carbon inputs, defined by crop management practices, were used invariantly every year.

Because there are around 420 municipalities in Norway, the code was set up to enable parallel computing (the computation is distributed municipality by municipality on different CPU cores).

Thanks to the modifications (i.e., hydrological model + C input model + stratification at municipality level), the ICBM is now able to account for:

1. C input variation between year and between municipalities.
2. variation of the SOM decomposition rate within a year and between municipalities (e.g., it would integrate the effect of local drought, frost, and cold summer).

However, to work at this level the upscaled model requires new model inputs such as detailed data on weather, soil texture, vegetation cover per crop type, crop yield, proportion of cultivated crop type, and manure. To progress we therefore need to work on the following points:

• Weather data: the precise dataset interpolated on a 1 by 1 km grid we used, only includes data between 1990 and 2016. This need to be updated for the most recent year. Average daily data derived from the period 2006-2016 would be used for the 2016-2020 period until we get the update.
• Soil texture distribution is only defined for thirty-one agrozone (AZ). We will investigate the possibility to define it at commune level, using the soil database (initially used to define soil texture at AZ level).
• Vegetation cover and C input are now function of crop yield. However, at the moment the model only takes into consideration the yield measured on one location (the highly monitored Skuterud catchment of the JOVA program). We will try to extract the yield data from the totality of the seven catchments monitored by the JOVA program). Each catchment will be used to produce reference yield for representative group of communes. We will try to investigate alternate source of Yield data to complement this approach (e.g., SSB, Jordplan).

Phase 3 Verification – CSC data for cropland mineral soil

There are strong requirements for verification before a Tier 3 methodology can be implemented in the national GHG accounting. In general models used for Tier 3 methodologies require enough observations to capture the diversity in climate, topography, and land-use management that is found within the member states borders.

Soil sampling at the Skuterud catchment

The Skuterud catchment is part of the JOVA monitoring program and as such hold a precise account of the management practice history on all fields for the last ~30 years. In 2001, about 250 georeferenced samples were taken from the catchment on a one hundred by 100 m grid and analyzed for C and N content and bulk density. This sampling campaign therefore enable to estimate the carbon stock over the catchment. In 2020, to assess the carbon stock change as a function of the field management history (i.e., ploughing type and frequency, use of fertilizer and pesticide, chalking, crop type, planting and harvesting date, etc.), we precisely resampled the positions selected in 2001 with the same methodology. So far, the carbon stock change data are being statistically analyzed with regards to the management history. The ambitions are to use the collected data set as part of a verification of ICBM in Norway.

Other data for soil carbon stock change in cropland

Through the project we will seek cooperation with other groups within and outside NIBIO for gathering and evaluating soil carbon data which can be used in verification of ICBM Tier 3 modelling of soil carbon stock changes in cropland mineral soil.

Phase 4: Documentation

The last phase of the project will focus on documentation. As a part of the implementation of the Tier 3 method in the GHG inventory, detailed documentation of the method is required. Documentation of the model and method as well as documentation of the QA/QC procedures and verification process.