beta

You're viewing our new website - find out more

Publication - Report

Climate change: evidence review in Agriculture, Forestry, Land Use, Waste

Published: 19 Jan 2017
Part of:
Environment and climate change
ISBN:
9781786527530

Evidence review of potential climate change mitigation measures in Agriculture, Forestry, Land Use and Waste.

156 page PDF

1.7MB

156 page PDF

1.7MB

Contents
Climate change: evidence review in Agriculture, Forestry, Land Use, Waste
Appendix A2. Review of models and tools for quantitative assessment of the wider impacts of ALULUCF GHG mitigation options

156 page PDF

1.7MB

Appendix A2. Review of models and tools for quantitative assessment of the wider impacts of ALULUCF GHG mitigation options

A2.1 Models and tools for air quality ( WI1- WI4)

The models described for assessing air quality focuses on a combination of air dispersion models ( EMEP4UK, SCAIL) which can output deposition and concentrations values to a grid or receptor, and an integrated model which can explore abatement scenarios and provide benchmarks for protection of ecosystems and air quality and human health ( UKIAM, GAINS). DNDC is a process based model which predicts crop yield, carbon sequestration, nitrate leaching loss, and emissions of carbon and nitrogen gases in agroecosystems. Most of the MOs outlined in this report can be assessed by way of altering input emissions to the models. The models can be used to explore national and local scale effects, although the models are restricted down to a resolution at the 1km scale ( EMEP4UK, UKIAM). However, individual local scale modelling can be carried out by models such as SCAIL to assess source to receptor impacts at the farm level.

A2.1.1 EMEP4UK

Table 30 Model description: EMEP4UK

Model/tool name EMEP4UK
References
Impacts assessed Air pollutants
Sectors covered Agriculture, industry, transport, stationary combustion (all emission sectors)
Geographical scope Country/Regional
Modelling approach The EMEP4UK model is a 3D eulerian atmospheric chemistry transport model ( ACTM) driven by the numerical weather prediction model weather and research forecast ( WRF).
The model is used to simulate photo oxidants and both inorganic and organic aerosols. The EMEP4UK model calculates hourly to annual average tropospheric atmospheric composition and deposition of various pollutants; including speciated components of PM 10, PM 2.5, secondary organic aerosols ( SOA), elemental carbon ( EC), secondary inorganic aerosols ( SIA), sulphur dioxide, NH 3, NO x, and ozone. Dry and wet deposition of pollutants are routinely calculated by the model.
EMEP4UK initially was developed as a regional application of the EMEP MSc-W model which is used to support the Convention on Long Range Transboundary Air Pollution ( CLRTAP). However, now the EMEP4xyz can be apply virtually anywhere in the world from Global run to nested regions at high resolutions.
Vieno et al. 2010

Simpson et al. 2012
Main model outputs UK pollutant maps (up to 1km x 1km grid)
e.g. https://eip.ceh.ac.uk/apps/atmospheric http://www.emep4uk.ceh.ac.uk/2014

Main data needs Country/Global emission inventory
Driven by real meteorology, therefore an EMEP compatible meteorological dataset is required. The EMEP4UK rv4.8 currently uses the WRF model version 3.7.1 (Weather Research and Forecasting) as meteorological driver.

Main limitations Level of expertise to run and making scenarios into emission maps.
Validation/ robustness EMEP4UK has been compared with other models.
Also validated with measurement networks.
The EMEP MSC-W model is extensively validated and verified and the model performances are reported annually in the EMEP status report.
http://emep.int/publ/emep2016_publications.html.
Carslaw et al. 2011a
Carslaw et al. 2011b
Dore et al. 2015
Vieno et al. 2010
Vieno et al. 2014
Vieno et al. 2016a
Scottish/ UK case study examples EMEP4UK has been used to model:
Ozone during a summer heat wave
Multiple years UK atmospheric composition PM air episodes
PM2.5 mitigation

Vieno et al. 2010
Vieno et al. 2014
Vieno et al. 2016a
Vieno et al. 2016b
Examples of integrated use The EMEP-MSc-W model has been integrated with the GAINS model Simpson et al. 2012

A2.1.2 UKIAM

Table 31 Model description: UKIAM

Model/tool name UKIAM ( UK Integrated Assessment Model)
References
Impacts assessed An integrated assessment modelling tool to support policy in relation to air pollutants and GHGs.
Sectors covered Agriculture, industry, transport - all sectors

Pollutants covered: sulphur dioxide, NO x, PM, NH 3. UKIAM has also been extended to include GHG emissions.
ApSimon et al. 2009
Geographical scope UK and regional
Modelling approach UKIAM projects UK emissions for sulphur dioxide, NO x, NH 3, PM 10 and PM 2.5 for future scenarios providing data on pollutant deposition, criteria for ecosystem protection, urban air quality and human health and data on potential emission abatement measures.

UKIAM uses pre-calculated source-receptor matrices derived from atmospheric modelling to estimate the response of baseline concentrations and deposition to changes in different sources both within and outside the UK. Abatement measures have been defined and incorporated into a Multi-Pollutant Measures Database giving percentage reductions in emissions achieved for each pollutant for a selected source, together with unit costs.

UKIAM remains an independent model paralleling GAINS but model at 1 to 5 km resolution over the UK using the FRAME model.
Oxley et al. 2003, Oxley et al. 2013


Oxley et al. 2013, AMEC 2009



Dore et al. 2007, Fournier et al. 2004
Main model outputs Cost data analysis tables, deposition maps.
Main data needs Emissions inventories and scenarios
Main limitations Time consideration need to be given for the multiple model runs.
Validation/
robustness
Model output from FRAME have been validated against measurements and compared with other models.
In general, it is less easy to validate modelled data on source attribution of pollutant concentrations and deposition against measurements.
Dore et al. 2015
Scottish/ UK case study examples PM2.5 emission abatement strategies and sensitivity to human health (in London).
UK assessment of traffic emissions and future scenarios and the UK's air quality strategy.
Oxley et al. 2015
Oxley et al. 2011
Examples of integrated use Already an integrated model

A2.1.3 DNDC

Table 32 Model description: DNDC

Model/tool name DNDC (Denitrification-Decomposition model)
References
Impacts assessed Predicts crop yield, carbon sequestration, nitrate leaching loss, and emissions of carbon and nitrogen gases in agroecosystems.
Sectors covered Agriculture
Geographical scope Site or regional
Modelling approach DNDC is a process-oriented computer simulation model of carbon and nitrogen biogeochemistry in agroecosystems. The entire model is driven by four primary ecological drivers, namely climate, soil, vegetation, and management practices.
The model consists of two components:
  1. Soil climate, crop growth and decomposition sub-models. Predicts soil temperature, moisture, pH, redox potential (Eh) and substrate concentration profiles. These are driven by ecological drivers (e.g., climate, soil, vegetation and anthropogenic activity).
  2. Nitrification, denitrification and fermentation sub-models. Predicts emissions of CO 2, CH 4, NH 3, nitric oxide, N 2O and dinitrogen from the plant-soil systems.
DNDC has been modified for application into the UK to produce UK- DNDC, and which was updated. It uses UK-specific input data. At the regional scale, UK- DNDC utilises its own databases.

Manure- DNDC represents the manure life cycle on farms and predict GHG and NH 3 emissions from livestock manure systems.
http://www.dndc.sr.unh.edu/model/GuideDNDC95.pdf
Gilhespy et al. 2014

Brown et al. 2002 Cardenas et al. 2013

Li et al. 2012
Main model outputs Simulated results including daily and annual crop biomass, carbon and nitrogen pools/fluxes, water budget and daily fluxes of NH 3, CH 4, N 2O, nitric oxide, and dinitrogen. These are recorded in a series of files (csv).
Main data needs 3 main datasets are required:
  1. Crop management parameters inputs are required (e.g. crop type, rotation, tillage, fertilization, irrigation etc.).
  2. Climate data for the years to be simulated should be provided (temperature, precipitation are required, additional data e.g. wind speed, solar radiation and relative humility can be provided).
  3. Soil parameters include texture, bulk density, pH etc.
Background concentrations of NH 3 and CO 2 can also be set.

Main limitations Large data input requirements
Validation/ robustness DNDC has now been used to simulate various cropping, grazing and forest systems in many countries. The agreement between the model simulations and measured values vary, with some studies reporting poor agreement. Giltrap et al. 2010
Scottish/ UK case study examples N 2O emissions from soils at county level for the UK. Four MOs were assessed and the results showed there were differences in the emission factors according to location. Cardenas et al. 2013
Examples of integrated use DNDC has been developed into various other sub-models: Wetland- DNDC, Forest- DNDC, CAPRI- DNDC. The INTEGRATOR model uses CAPRI- DNDC. Gilhespy et al. 2014
De Vries et al. 2011

A2.1.4 GAINS

Table 33 Model description: GAINS

Model/tool name GAINS (The Greenhouse gas -Air pollution INteractions and Synergies)
References
Impacts assessed Estimates the environmental effects of air pollution under consideration of GHG emissions. The model simulates the flow of pollutants from their sources to their multiple effects, and estimates costs and impacts of policy interventions. Assesses economic sectors and options for emission control, costs of implementation in terms of reducing ecosystem and human health impacts. GAINS agriculture: An NH 3 module for GAINS has been developed for NH 3 emissions from animal manure at 4 stages - housing, storage, application and grazing. Emission factors and a set of abatement measures are defined for each stage. Amann et al. 2011a Klimont & Winiwarter 2014
Sectors covered Agriculture, Industry, Transport
Pollutants covered: sulphur dioxide, NO x, volatile organic acid, PM, NH 3, CO 2, CH 4, N 2O.

Geographical scope Individual countries, regions and global
Modelling approach Cost-benefit source-receptor model taking into account atmospheric chemistry, quantification of ecosystem and human health responses
Main model outputs Cost data analysis tables, deposition maps.
Main data needs Cost data (investment costs, operating costs (fixed & variable)), future scenarios & baseline projections of economic activities.
Main limitations Dependent on complete emission inventories
Validation/ robustness No information
Scottish/ UK case study examples EU member states including UK. Outputs included: Health impact indicators, critical load exceedance for nitrogen and acidification. Amann et al. 2011b
Examples of integrated use Various assessments of EU and UNECE policies, e.g. National Emission Ceilings Directive, Gothenburg Protocol

A2.1.5 MODDAS- THETIS

Table 34 Model description: MODDAS- THETIS

Model/tool name MODDAS- THETIS
References
Impacts assessed Estimates the pollutant recapture by trees for NH 3 and PM
Sectors covered Agriculture ( NH 3) and combustion sources ( PM)
Geographical scope Site based assessments (single source)
Modelling approach MODDAS- THETIS is a flexible two-dimensional (along wind and vertical) model that can be used to examine the pollutant abatement potential of tree shelter-belt structures in the landscape. MODDAS is a Lagrangian stochastic model for gaseous dispersion and THETIS is turbulence model designed for transfer within the planetary boundary layer as well as within a plant canopy.

The model scenario setup is based around a woodland schema where different blocks of canopy are designed of varying height and width and density (Leaf Area Index - LAI). Source strength and the source length can also be configured.
Loubet et al. 2006 Foudhil 2005




Bealey et al. 2014
Main model outputs Data table of pollutant recapture %
Concentrations and deposition plots - before, within and after the canopy.

Main data needs Source emissions
Main limitations Can only be used for single sources
Validation/
robustness
Both models have been validated in conditions similar to those modelled here, specifically MODDAS in an NH 3 release experiment over a developed maize canopy and a grassland, and THETIS over several canopy arrangements. Loubet et al. 2006,
Foudhil 2005
Dupont and Brunet, 2006
Scottish/ UK case study examples No real-life scenarios applied as yet.
Modelling of a housing scenario showed that a 30-50 m deep tree shelter belt could capture up to 15-20% of the NH 3 emitted.

Bealey et al. 2014
Examples of integrated use Not yet.

A2.1.6 SCAIL

Table 35 Model description: SCAIL

Model/tool name SCAIL (Simple Calculation of Atmospheric Impact Limits)
References
Impacts assessed Estimates concentrations and deposition from local sources
Sectors covered Agriculture ( NH 3, nitrogen and acid deposition, PM) and combustion sources ( NO x, sulphur dioxide, nitrogen and acid deposition, PM)
Geographical scope Site based assessments (multi-sources)
Modelling approach SCAIL is a suite of screening tools for assessing the impact from agricultural and combustion sources on semi-natural areas like SSSIs and SACs. SCAIL provides an estimate of the amount of acidity, nitrogen or sulphur deposited to an ecosystem.
Meteorology in the model is provided by 40 meteorological stations around the UK SCAIL uses the air dispersion model Aermod.
Hill et al. 2014a www.scail.ceh.ac.uk
Main model outputs Data table of source contribution to pollutant concentration and deposition. Provides critical load exceedance statistic for ecosystems
Main data needs Background concentration and deposition maps

Meteorological data (wind speed, wind direction)

Emission, livestock numbers, storage/spreading volumes etc

Main limitations Only for use in local site-based assessments
Validation/
robustness
SCAIL has been validated against measurements taken around farms and anaerobic digesters.

Provides a best estimate for pollutant impacts.
Hill et al. 2014b, Bell et al. 2016
Scottish/ UK case study examples Used across UK and in Scotland by SEPA for permitting purposes
Examples of integrated use Not yet.

A2.2 Models and tools for water quality ( WI5- WI7)

The models selected are able to assess the effects of MOs at at least farm scale, with scope for upscaling to catchment, regional or national scales. All of the models listed have been successfully used in UK studies to assess the wider effects of GHG MOs. The Farmscoper model and the ADAS Wales Framework are related suites of models able to assess the impacts of mitigation measures on a range of pollutants and pathways. The LUCI model is a GIS based ecosystem services model which assess the effects of land use and management and is able to include the effects of afforestation and peatland restoration as well as measures to reduce GHG emissions from agricultural activities. DNDC, mentioned in the previous section (A2.1.3), is also capable of modelling certain water quality impacts.

A2.2.1 ADAS Wales

Table 36 Model description: ADAS Wales

Model/tool name ADAS Wales  
  Diffuse Pollution Emission Modelling Framework References
Impacts assessed Nitrate, phosphorus, sediment, pesticides, veterinary medicines, N 2O, CH 4 and CO 2 Emmett et al. 2014
Sectors covered Agriculture
Geographical scope Catchment scale, parameterized for Wales.
Modelling approach The framework is similar to the Defra Farmscoper model (see Section A2.2.3), and combines a suite of models to calculate emissions of nitrate, phosphorus, sediment, pesticides, veterinary medicines, N 2O, CH 4 and CO 2. The framework is stratified by Robust Farm Type and reported emissions for each of the Water Framework Directive river catchments in Wales. The modelling framework uses a combination of process based and inventory models.
Emissions of pesticides and veterinary medicines are calculated using the regulatory MACRO and PRZM models. Phosphorus and sediment losses are calculated using the PSYCHIC model.
Nitrate losses are calculated using the N-CYCLE, NITCAT and MANNER models.

CH 4 and N 2O emissions are calculated using the tier one and two IPCC methodology with modifications to represent the effects of observed levels of soil compaction and poaching on N 2O emissions. Indirect N 2O emissions from leached nitrate were calculated using the appropriate nitrogen leaching model.
The framework contains a meta-model of export coefficients derived from process based models describing the effects of 40 individual mitigation methods for pollutant emissions to air and water. The modelling framework provides a consistent assessment of multiple pollutants to air and water from agriculture in Wales, which explicitly links the impact of MOs intended to improve water quality with their secondary impacts on emissions of GHGs.
Anthony and Gooday 2010, Emmett et al. 2014




Jarvis 1994, Carsel et al. 1984
Davison et al. 2008
Scholefield et al. 1991, Lord 1992, Chambers et al. 1999 Baggott et al. 2006, IPCC 2006
Main model outputs Emissions of nitrate, phosphorus, sediment, pesticides, veterinary medicines, N 2O, CH 4 and CO 2
Main data needs Spatial database of agricultural activity, separated by farm system type. Data on agricultural practices (stocking levels, crop rotations, fertiliser application rates, manure management) and uptake of the mitigation measure e.g. June Agricultural Census, British Survey of Fertiliser Practice and Farm Practice Survey.
Water Framework Directive Catchment boundaries.
Monthly average rainfall, temperatures and number of rain days on a 5 by 5 km 2 grid.
Soil particle size distribution (percentage sand, silt and clay), organic matter content bulk density and HOST class of the dominant soil series within each 1 km 2 squares.
Digital Elevation Model.
Land cover data.
Discharge consents database for non-agricultural pollution inputs.
Anthony et al. 2012
Main limitations Has currently only been developed for Wales, although parameterisation may be similar between Wales and Scotland.
Validation/
robustness
Gives a consistent framework for using several existing well established models.
Scottish/ UK case study examples Used to evaluate the effect of Welsh Government Agri-Environment schemes. Anthony et al. 2012
Examples of integrated use This tool is itself an integrated suite of models. It has been integrated with the LUCI model as part of the Welsh Government Glastir Monitoring and Evaluation Programme.
The Farmscoper model incorporates a similar suite of models to the ADAS Wales model but also includes additional models to assess emissions of NH 3. Farmscoper integrates emission data with unit costings for measure implementation in an algorithm which optimizes measures to maximize benefits for the range of wider impacts.
Emmett et al. 2014


Anthony and Gooday 2010

A2.2.2 LUCI

Table 37 Model description: LUCI

Model/tool name LUCI (Land Utilisation and Capability Indicator) formerly Polyscape
    References
Impacts assessed Water quality (nitrogen, phosphorous and sediment run-off) flood risk, carbon sequestration, habitat connectivity.
Sectors covered Mountains, moors and heaths; Semi-natural grasslands; Enclosed farmland; Woodland; Freshwater, wetlands and floodplains; Urban
Geographical scope Site to catchment or landscape scale.
Modelling approach LUCI is GIS-based spatially explicit ecosystem service model. It is a process-based tool which maps ecosystem services using a range of algorithms that maintain biophysical principles and spatial connections using lookup tables, combined with topographic routing of water, sediment and nutrients over the landscape. It is spatially explicit at the resolution of the topographic data layer used: model applications to date have used a 5m by 5m resolution. Jackson et al. 2013
Main model outputs Agricultural productivity, carbon stock and condition, flood mitigation and concentration. Accumulation of nitrogen, phosphorous over the landscape. In stream discharge, nitrogen and phosphorous concentration and load
Main data needs Required spatial data layers: Digital Elevation Model topography layer, Land use (several supported, for UK LCM2007), Soil type (several supported, for UK NATMAP) If available: Long term annual average precipitation and predicted evapotranspiration, Detailed river network Lookup tables (values provided for supported datasets): Soil and biomass carbon, land use export coefficients for nitrogen and phosphorous, cost distance for species dispersal, soil fertility, drainage and waterlogging
Main limitations Does not report uncertainty. Does not include valuation.
Validation/ robustness Quantitative. Provides spatially explicit ecosystem service trade off maps.
Scottish/ UK case study examples Welsh Government Glastir Monitoring and Evaluation Programme ( GMEP).
Loweswater catchment modelling for Defra.
Natural England Bassenthwaite catchment project.
Emmett et al. 2014
Norton et al. 2014
Examples of integrated use In the GMEP project LUCI has been integrated with the Multimove habitat and species model and with ADAS Wales Diffuse Pollution Emission Modelling Framework.

A2.2.3 Farmscoper

Table 38 Model description: Farmscoper

Model/tool name Farmscoper  
    References
Impacts assessed NH 3, nitrate, phosphorus, pesticides, N 2O, CH 4 and CO 2 Anthony and Gooday 2010
Sectors covered Agriculture
Geographical scope Farm scale, England and Wales and had been scaled to catchment level Zhang et al. 2012
Modelling approach The framework is similar to the ADAS Wales Diffuse Pollution Emission Modelling Framework, and combines a suite of models to calculate emissions of NH 3, nitrate, phosphorus, sediment, N 2O, CH 4 and CO 2. Farms can be models based on Robust Farm Type within soil and climate zones. Emissions of pesticides are calculated using the MACRO and model.

Phosphorus are calculated using the PSYCHIC model.

Nitrate losses are calculated using the NEAP_N, N-CYCLE, NITCAT, MANNER and EDEN models.

NH 3 emissions are calculated using the NARSES and MANNER models.

CH 4 and N 2O emissions are calculated using the tier one and two IPCC methodology with modifications to represent the effects of observed levels of soil compaction and poaching on N 2O emissions. Indirect N 2O emissions from leached nitrate were calculated using the appropriate nitrogen leaching model.
The framework contains a meta-model of export coefficients derived from process based models describing the effects of 97 individual mitigation methods for pollutant emissions to air and water. FARMSCOPER can estimate the cost and effectiveness of mitigation methods individually, so that mitigation methods of interest can easily be identified. It also allows for the evaluation of multiple mitigation methods, as these will not simply be the sum of the impacts of the individual methods, due to interaction and competition between methods
Emmett et al. 2014
Anthony and Gooday 2010
Jarvis 1994
Davison et al. 2008 Lord & Anthony 2000, Scholefield et al. 1991, Lord 1992, Gooday et al. 2008 Webb & Misslebrook 2004, Chambers et al. 1999 Baggott et al. 2006, IPCC 2006



Anthony and Gooday 2010
Main model outputs Emissions of NH 3, nitrate, phosphorus, pesticides, N 2O, CH 4 and CO 2. Optimisation of combined MOs. Anthony and Gooday 2010
Main data needs Spatial database of agricultural activity, separated by farm system type. Data on agricultural practices (stocking levels, crop rotations, fertiliser application rates, manure management) and uptake of the mitigation measure e.g June Agricultural Census, British Survey of Fertiliser Practice and Farm Practice Survey.
Monthly average rainfall, temperatures and number of rain days on a 5 by 5 km 2 grid.
Soil particle size distribution (percentage sand, silt and clay), organic matter content, bulk density and HOST class of the dominant soil series
Digital Elevation Model.
Land cover data.
Anthony and Gooday 2010
Main limitations Has currently only been developed for England and Wales, although parameterisation may be similar for Scotland.
Validation/ robustness Gives a consistent framework for using several existing well established models.
Scottish/ UK case study examples Farmscoper was developed and used to model the benefits of MOs at a farm scale in Defra project WQ0106.
It has been upscaled for use at a catchment scale in the Hampshire Avon Demonstration Test Catchment, and have a modified version of Farmscoper forms the basis of the ADAS Wales Diffuse Pollution Emission Modelling Framework
Anthony and Gooday 2010
Zhang et al. 2012 Emmett et al. 2014
Examples of integrated use This tool is itself an integrated suite of models Anthony and Gooday 2010

A2.2.4 NIRAMS

Table 39 Model description: NIRAMS

Model/tool name NIRAMS (Nitrogen Risk Assessment Model for Scotland)  
    References
Impacts assessed Nitrogen leaching Dunn et al. 2004a, 2004b
Sectors covered Agriculture
Geographical scope Scotland
Modelling approach Calculates N balances, weekly nitrogen leaching and catchment scale nitrogen transport
Main model outputs Streamwater nitrogen concentrations draining from agricultural land; outputs are reliable above 30 km 2 resolution
Main data needs Land use, soil, topographical, meteorological data
Main limitations Predicting long term changes; uncertainty of grassland N balances
Validation/ robustness Successfully reproduced weekly nitrogen flows in eight test catchments
Scottish/ UK case study examples Nitrogen leaching and water nitrate concentration in Scotland
Examples of integrated use

A2.3 Models and tools for soil quality ( WI8)

CARBINE is a forest carbon model which has been developed by Forest Research (the Forestry Commission's research agency) to assess the effects of forest-related activity on soil carbon stocks under UK conditions. It has been developed primarily to assess changes in mineral soils, but is being developed to improve modelling of effects on organic soils. The Windfarm Carbon Calculator has been developed specifically to assess the effects of wind turbine developments on the carbon stocks landscapes in Scotland, particularly those with high carbon soils. Although primary developed to assess the effect of large windfarm developments it could also be applied to smaller on-farm turbines schemes.

A2.3.1 CARBINE

Table 40 Model description: CARBINE

Model/tool name CARBINE  
    References
Impacts assessed C stocks of stands and forests in living and dead biomass and soil, and associated harvested wood products
Sectors covered Forestry
Geographical scope UK at stand, forest and national level
Modelling approach The model consists of four sub-models or 'compartments' which estimate carbon stocks in the forest, soil, and wood products and, additionally, the impact on the GHG balance of direct and indirect fossil fuel substitution attributable to the forestry system.
The model is able to represent all of the introduced and native plantation and naturally-occurring species relevant to the UK. The forest carbon sub-model is further compartmentalised to represent fractions due to tree stems, branches, foliage, and roots. The soil carbon sub-model runs independently of the forest sub-model. Initial soil carbon is estimated based on land use/cover and soil texture (sand, loam, clay and peat). The timecourse of any soil carbon stock change is assumed to follow an exponential form with the magnitude of the stock change and rate constant dependent on the soil type and on the particular land-use transformation.
Robertson et al. 2003
Main model outputs C stocks of stands and forests in living and dead biomass and soil, and associated harvested wood products.
Impact on the GHG balance of direct and indirect fossil fuel substitution

Main data needs Areas and age-class distributions of each tree species.
Estimates of stand structure and growth obtained from yield tables applied at the stand level. Pre-afforestation land use/cover and soil texture (sand, loam, clay and peat).

Edwards and Christie 1981
Main limitations The impact of different forest management regimes can only be assessed for the range of tree species, yield classes and management regimes represented in published yield tables.
The standard thinning regime assumed for most species is based on recommended practice. However, actual forest management departs significantly from these recommendations. Unmanaged or 'semi-natural' forest is poorly modelled as it is assumed to follow the same growth patterns as unthinned productive forest up to the maximum potential carbon stock.
Uncertainty of modelling change in soil carbon stocks for forests on organic soils is high, although an improved version of CARBINE is being developed which incorporates elements of the ECOSSE soil model to address this.

Validation/ robustness Although widely used by Forest Research the model has not been subject to peer view. However, the results have been validated against available field data. The soil sub-model is based on the established Roth-C model. Coleman et al. 1997
Scottish/ UK case study examples CARBINE is used by Forest Research to model carbon stocks of UK forests, and is used to generate estimates of change in forest carbon stocks for the UK LULUCF inventory. Thompson and Matthews 1989, Mason and Kerr 2004, Broadmeadow and Matthews 2003, Brown et al. 2016a
Examples of integrated use An improved version of CARBINE is being developed which incorporates elements of the ECOSSE soil model to improve modelling of change in soil carbon stocks for afforested organic soils.

A2.3.2 SPACSYS

Table 41 Model description: SPACSYS

Model/tool name SPACSYS  
    References
Impacts assessed Predicts crop yield, carbon sequestration, nitrate leaching loss, and emissions of carbon and nitrogen gases in agroecosystems. Wu et al. 2007
Sectors covered Agriculture
Geographical scope Site
Modelling approach The model describes crop yield, nitrate and carbon cycling, and it includes a soil water component that includes representation of water flow to field drains as well as downwards through the soil layers. The model is process based for the crop and the soil components. The model can also be run in a 3D-root mode at the single plant level to assess the effects of root growth on the uptake of nitrogen. The root growth, direction and elongation rates are modelled. Wu et al. 2007
Main model outputs Simulated results including daily and annual crop biomass, carbon and nitrogen pools/fluxes, water budget and daily fluxes of NH 3, CH 4, N 2O, nitric oxide, and dinitrogen. Wu et al. 2015
Main data needs Crop management, Soil parameters include texture, bulk density, pH etc., and daily climate data (max & min temperature, precipitation, wind speed and either vapour pressure or relative humidity, and either global and net radiation or cloudiness and sunshine hours).
Main limitations Required inputs, and the sparse level of validation, particularly for cropping systems in the UK.
Validation/ robustness The model has been validated for N 2O emissions for grassland and arable systems in the UK and Italy. At the China sites the model was validated for soil carbon and N 2O emissions. The 3D root model has been validated against white clover, winter wheat. Wu et al. 2015,
Abalos et al. 2016
Perego et al. 2016
Zhang et al. 2016a & 2016b
Bingham & Wu 2011
Scottish/ UK case study examples The model has been validated against N 2O emissions from a manuring trial conducted in Edinburgh. The model has also been validated against grassland for the south-west of England. Wu et al. 2015
Abalos et al. 2016
Examples of integrated use No information

A2.3.3 Windfarm carbon calculator

Table 42 Model description: Windfarm carbon calculator

Model/tool name Windfarm carbon calculator  
    References
Impacts assessed The Windfarm carbon calculator is the Scottish Government's tool to support the process of determining wind farm developments in Scotland. The tool assesses, in a comprehensive and consistent way, the carbon impact of wind farm developments.
Sectors covered Windfarms
Geographical scope Scotland on a site by site basis
Modelling approach The latest version of the carbon calculator is a web-based application linked to central database, which stores all of the data entered.
Emissions due to construction and operation of the windfarm area estimated from life cycle analysis.
Peat which is removed is assumed to be instantaneously oxidized. The carbon dynamics of disturbed peat on site are modelled using IPCC Tier 1 methodology as default, although more complex modelling can be accommodated where available.
Change in the carbon stocks of forests can be modelled using either a simple methodology based on yield classes or more detailed modelling based on the 3 PG tree growth model.
Scottish Government
Nayak et al. 2008, Nayak et al. 2010, Smith et al. 2011



Xenakis et al. 2008
Main model outputs Loss of carbon due to production, transportation, erection, operation and decommissioning of wind farm and back up generation provision; change in carbon dynamics of peatlands; changes in carbon stocks due to forestry clearance; impacts of forestry management on windfarm carbon emission savings
Main data needs Emission factor for displaced power source, site capacity factor, and rated capacity of turbines.
Life cycle analysis data for carbon losses due to production, transportation, erection, operation and decommissioning of wind farm.
Areas of peat affected removed and affected by drainage. Peat depth. Data on the extent and type of structures on site and extent of restoration of drained peat.
Area and average carbon stock of forest felled.
Average temperature.

Main limitations IPCC emission factors are used for emissions from drained peatlands. These may not be appropriate for UK peatlands, particularly blanket bogs.
Model is site specific.

Validation/ robustness Model has been peer reviewed Nayak et al. 2010
Scottish/ UK case study examples Model and previous versions of it is the standard tool for assessing the carbon balance of Scottish Windfarms.
Examples of integrated use No information

A2.4 Models and tools for flood management and water use ( WI9)

The models selected were chosen because they are best able include the effects of some or all of the land cover and soil factors which are likely to be affected by the MOs and have consequences for flood risk and water use.

A2.4.1 IHMS

Table 43 Model description: IHMS

Model/tool name IHMS  
    References
Impacts assessed Changes in water resources (surface and groundwater) availability due to land use and climate changes Ragab and Bromley 2010, Ragab et al. 2010
Sectors covered Water resources, Hydrology and Agriculture
Geographical scope Catchment scale
Modelling approach Distributed - Physically based hydrological process-daily based.
Main model outputs All water balance components, evaporation, infiltration, stream flow, groundwater recharge, runoff, plant water uptake, groundwater levels, soil moisture, wetness index, Ragab and Bromley 2010
Main data needs Rainfall, climate, soils, land cover, elevation, vegetation/ land cover parameters, stream parameters Ragab and Bromley 2010
Main limitations Not for national scale (i.e. UK as a whole), best for catchment scale.
Validation/
robustness
Has been validated for several catchment without problems. D'Agostino et al. 2010; Montenegro and Ragab 2010, 2012
Scottish/ UK case study examples Currently is successfully used for Eden Catchment, Scotland and 5 other catchment across the UK; Pang, Don, Frome, Fowey and Ebbw. DRY project- NERC grant (2014-2018): http://www1.uwe.ac.uk/et/research/dry/dryprojectsummary.aspx
Examples of integrated use Linked to MODFLOW (groundwater flow model)and SWI (Seawater intrusion model) models Ragab et al. 2010

A2.4.2 SALTMED

Table 44 Model description: SALTMED

Model/tool name SALTMED  
    References
Impacts assessed Changes in water balance components, crop growth, yield and nitrogen cycle due to changes in land use, water availability, field, Nitrogen fertilizers, and climate changes (e.g. CO 2, temperature, drought etc.) Ragab 2015a
Sectors covered Agriculture
Geographical scope Field scale
Modelling approach Field scale model, physically-biologically based process-daily based Ragab 2015a
Main model outputs All water balance components, evaporation, infiltration, irrigation, drainage, biomass, dry matter, yield, plant water uptake, soil moisture, soil salinity, soil nitrogen, etc. Ragab 2015a
Main data needs Rainfall, climate, soils, land cover, vegetation (crops/trees) parameters, land management parameters, nitrogen-fertilizers (organic, inorganic) input, .yjv
Main limitations Field scale only
Validation/
robustness
Has been validated for several fields worldwide without problems Ragab et al. 2015b, Pulvento et al. 2015 (There are at least 20 papers on validation of SALTMED)
Scottish/ UK case study examples Currently is in use at Harper Adams University, UK See more at Water4Crops EU funded project web site at: http://www.water4crops.org/
Examples of integrated use Will be integrated into a catchment scale model as part of the DRY project DRY project- NERC grant (2014-2018): http://www1.uwe.ac.uk/et/research/dry/dryprojectsummary.aspx

A2.5 Models and tools for land use and land cover ( WI10)

Changes in land use and land cover are likely to be driven by a number of other factors as well as climate change mitigation measures, and the effect of these measures may be small compared to other demands on land such as the need to provide timber, food, housing and recreational opportunities and may also change in response to climate change and market and policy forces. It is therefore difficult to separate out the effects of MOs on land use and land cover from wider effects. The LULUCF inventory contains information on land use and land cover for each UK administration, and is able to protect change in land use and management and consequent change in GHG emissions and soil carbon stocks. The LULUCF inventory is able to produce projections of the effect of land use and management on GHG emissions and carbon stocks to 2050 using scenarios which can be developed based on policy aspirations or projected market trends.

To assess the land use effects of larger scale changes caused by afforestation policy or a change in demand for livestock products, land allocation models, like spatial econometric models, can be used.

A2.5.1 LULUCF Inventory

Table 45 Model description: LULUCF Inventory

Model/tool name LULUCF Inventory  
    References
Impacts assessed GHG emissions and removals and change in carbon stocks in living biomass, soil, dead organic matter and harvested wood products as a result of change in land use and management. Brown et al. 2016a
Sectors covered Grassland, Cropland, Forest, Wetland, Settlement Land, Other Land
Geographical scope UK administrations, Jersey, Guernsey, the Isle of Man and the Falkland Islands. Can be disaggregated to local authority level.
Modelling approach The LULUCF inventory uses methodology laid out by the Intergovernmental Panel on Climate Change ( IPCC).
Much of the UK LULUCF inventory uses a simple "Tier 1" approach in which a default emission factor ( EF) for an activity is multiplied by "activity data" such as the area of land undergoing a particular activity or the quantity of material involved. For more significant activities more complex methodologies are used e.g the CARBINE model is used to generate estimates of change in carbon stocks in Forests, an exponential model is used to assess change in soil carbon stocks, and UK specific emission factors are being developed for peatland drainage and rewetting.
The UK LULUCF inventory is compiled by aggregating inventories for the constituent administrations. Emissions and removals can be disaggregated to a statistical basis to be mapped at Local Authority level.
The LULUCF inventory is able to produce projections of the effect of land use and management on GHG emissions and carbon stocks to 2050 using scenarios which can be developed based on policy aspirations or projected market trends.
IPCC 2006, IPCC 2013
Main model outputs GHG emissions and removals; change in carbon stocks in living biomass, soil, dead organic matter and harvested wood products.
Main data needs Data on land use and management, including the extent of farming practices, peat extract activity, and wildfires.
To produce projections to 2050 scenarios for change in land use and management are needed.

Main limitations In its current form the LULUCF inventory is does not use spatially explicit activity data, and so apportions activity to soil type and climate on a statistical (proportional) basis. However a methodology is being developed which will allow the LULUCF inventory to assimilate spatially explicit land use data and track "land use change vectors" for particular land parcels.
In some cases the IPCC default Tier 1 EF may not fully reflect UK conditions. For example the IPCC Tier 1 EFs for Wetland Drainage and Rewetting ( WDR) are more relevant to fens and raised bogs than to the blanket bogs prevalent in much of Scotland. A DBEIS (formerly DECC) funded research project which is due to report in autumn 2016 is compiling improved EFs and activity data for WDR activities in the UK.

Validation/ robustness The LULUCF inventory is compiled using internationally agreed methodology, and the annual inventories are subject to international review.
Scottish/ UK case study examples Used to produce annual GHG inventories for the LULUCF sector for the UK and its constituent administrations. Brown et al. 2016a, Salisbury et al. 2016
Examples of integrated use Uses the CARBINE forest carbon model to assess change in forest carbon stocks.

A2.5.2 Spatial econometric models

Table 46 Model description: Spatial econometric models

Model/tool name Spatial econometric models  
    References
Impacts assessed Agricultural land use and production, impacts of market and policy changes (e.g. prices, subsidies) and impacts of changes in biophysical constraints Fezzi and Bateman 2011
Sectors covered Agriculture
Geographical scope England and Wales, 5x5 km
Modelling approach Spatially disaggregated, structural econometric model of agricultural land use and production
Main model outputs Land use shares in each grid square, crop and livestock production
Main data needs Historic spatial data on land use, livestock and crop production, prices Data for future scenarios (e.g. prices)
Main limitations Does not exist for Scotland (though being developed by a PhD student in SRUC)
Validation/ robustness No information
Scottish/ UK case study examples Climate change impacts on food production
Fezzi et al. 2015
Examples of integrated use No information

A2.5.3 Agent based land use models

Table 47 Model description: Agent based land use models

Model/tool name Agent based land use models  
    References
Impacts assessed Agricultural land use and production, impacts of market and policy changes (e.g. prices, subsidies) and impacts of changes in biophysical constraints Murray-Rust et al. 2014a, 2014b
Sectors covered Rural land use
Geographical scope Europe/ UK
Modelling approach Empirical agent-based model
Main model outputs Depends on the model, an example: economic (gross margin difference), environmental (land use cover, nitrogen use, diversity) and social (access to green space) outputs Guillem et al. 2015
Main data needs Spatial land use data, climatic and soil data, data on farmers' behaviour
Main limitations Difficult to validate, mostly only calibration happens Brown et al. 2016b
Validation/ robustness See above
Scottish/ UK case study examples Land use and ecosystem services in a Scottish arable catchment Energy crop production in the UK Guillem et al. 2015
Alexander et al. 2013
Examples of integrated use Skylark population model Guillem et al. 2015

A2.6 Models and tools for biodiversity ( WI11)

The selection of models and tools for the assessment of biodiversity impacts focussed on those determining direct effects on terrestrial biodiversity. Collectively these models cover a range of indicators recommended by the European BioBio project (Herzog et al. 2012) including those relating to Habitat Diversity and Species Diversity of key groups (i.e. spiders, vascular plants and bees). Models/Tools for assessing biodiversity impacts are primarily related to habitat type with some tools (i.e. AgBioscape and SRUC's Biodiversity Calculator) also having the potential to model different land management options (e.g. crop rotations in the case of AgBioscape). Consequently the models outlined below are typically effective for detecting impacts of MOs that result in changes to landcover (i.e. Agroforestry, Afforestation, Peatland restoration, Reduced livestock product consumption and Incorporating legumes in grass mixes and crop rotations). Models/tools are less sensitive in detecting impacts of MOs that influence habitat quality or that involve finer changes to land management (e.g. Increased uptake of precision farming techniques, Achieving and maintaining optimal soil pH level and Optimising mineral nitrogen fertilisation).

A2.6.1 Interactive Habitat Network User Tool

Table 48 Model description: Interactive Habitat Network User Tool

Model/tool name Interactive Habitat Network User Tool  
    References
Impacts assessed The interactive online tool assesses the impact of land use change (e.g. afforestation/peatland restoration) on structural and functional ecological connectivity for four key habitats (i.e. Broadleaved woodland, Heathland, Neutral grassland and Wetland).
For those wishing to create additional networks (e.g. for a specific species) or utilise the system in GIS additional information is available from SNH Natural Spaces website and/or Phil Baarda, SNH. These include:
Spatial datasets- habitat networks indicated above, acid grassland network, hotspots for habitat creation (i.e. for Broadleaved woodland, Wetland and, Neutral grassland).
Users manual-Outlines modelling procedure using GIS ArcMap including spatial data requirements (Blake & Mattisson 2012).
Tools: ArcMap GIS tool to help automate the creation of new networks.
http://www.snh.gov.uk/land-and-sea/managing-the-land/spatial-ecology/habitat-networks-and-csgn/interactive-habitat-network-tool/
http://gateway.snh.gov.uk/natural-spaces/index.jsp Blake & Mattisson 2012
Sectors covered Agriculture, Forestry, Peatlands
Geographical scope Central Scotland Green Network area, Loch
Lomond and the Trossachs national park and the Scottish Borders

Modelling approach GIS based model utilising least-cost modelling procedures based on Forest Research landscape ecology model BEETLE (Biological and Environmental Evaluation Tools for Landscape Ecology). http://www.snh.gov.uk/docs/B692517.pdf,
Watts et al. 2010
Main model outputs A series of spatial maps illustrating extent of existing habitat networks. Interactive online tool enables altering current land use (e.g. creation of a new woodland of a specific size in a specific location) and determining the impact of this change on the extent of existing networks. Summary information on network metrics (i.e. the number of networks and the size of each network) are provided.
Main data needs Scenarios are inputted by manually drawing the area of proposed land use change and proposed new habitat on the online GIS system. New habitat networks are then generated to determine the impact. New habitat networks can be calculated for species with either high or moderate dispersal.
Main limitations Online tool restricted with respect to the habitats and geographical locations noted above. Networks are not based on actual species but Generic Focal Species for the habitat in question. This generic species is given either moderate or low dispersal powers.
Decisions for land-use change should not solely be based on habitat network modelling and additional factors should be taken into account. For example, creation of native woodland on a SSSI raised bog may increase the extent of a Broadleaved woodland but would result in the loss of a valuable habitat.

Validation/
robustness
Models based on spatial datasets that categorise habitats at a specific point in time. Potential errors with respect to incorrect categorisation of habitats and changes to land cover. Differences in habitat quality are not acknowledged during network creation.
Little scientific evidence investigating the impact of functional/structural connectivity on actual species dispersal.

Scottish/ UK case study examples Habitat network modelling has been used to explore the extent of current habitat networks in Falkirk, Ayrshire and Glasgow to prioritise areas habitat creation to optimise ecological connectivity.
Outputs from this tool are of direct relevance to the BioBio Indicator Habitat Diversity
Chetcuti et al., 2011
Moseley et al. 2008,
Smith et al. 2008
http:/www.forestry.gov.uk/fr/ infd-6w7evk
Herzog et al. 2012
Examples of integrated use Potential integration with other spatial datasets available for Scotland including: suitability mapping for native woodland creation, carbon stock mapping, changes to distribution of 'prime' land under climate change and Ecosystem Service Mapping. Lilly & Baggaley 2013,
Towers et al. 2011,
Brown et al. 2008,
Winn et al. 2015a

A2.6.2 SRUC's Biodiversity Calculator

Table 49 Model description: SRUC's Biodiversity Calculator

Model/tool name SRUC's Biodiversity Calculator  
    References
Impacts assessed The calculator assesses the impact of land use change (e.g. from winter wheat to unimproved pasture) on the number of vascular plant and spider species in a field. Yelloy 1999
Sectors covered Agriculture, Peatlands
Geographical scope Scotland
Modelling approach Biodiversity data were collected from agricultural land covers across Scotland. From these data predictive models were generated from Generalised Linear Interactive Modelling using linear regression to determine the importance of measured environmental variables (e.g. altitude, land use) on response variables (i.e. the number of vascular plant and spider species). Resultant models predict the richness of plant and spider assemblages in a field based on specific input parameters (e.g. current land cover, proposed new land cover, altitude, stocking density). Murphy et al. 1998
Main model outputs Interactive tool provides graphical and textual information on the predicted number of vascular plant and spider species in the current land use and in the proposed new land use, alongside the mean value for a field of the type in question.
Main data needs The interactive tool requires manual inputting via text/drop down menus of the follow information: Field altitude and area, current and proposed land use, years since sown, stocking density, uncultivated headland width, number of cuts, presence of hedgerows and vegetation type.
Main limitations The model is restricted to the following land covers: spring barley, improved pasture, set-aside, winter wheat, oilseed rape, spring barley, heather moorland, gorse grassland, unimproved pasture, root crops.
The model is restricted to spiders and vascular plants.
Model simply reports the number of species and provides no information on which species are present and their rarity/ conservation status.
Interactive tool restricts environmental variables to those that the user can easily determine. Environmental variables included in the initial linear regression models that are not readily measured (e.g. soil organic content) are omitted to facilitate use by target audience (e.g. farmer/agricultural advisor).
The tool calculates the impact of land use change at a field level and the importance of landscape heterogeneity at promoting biodiversity is thus not taken into account.

Validation/
robustness
The original models (i.e. inclusive of environmental variables that are not readily measured) were found to be accurate +/- 4 species for vascular plants and 68% accurate for spiders. Removal of environmental variables that are difficult to determine from the interactive tool will decrease prediction accuracy.
Original modelling determined different optimum models for different field types. The interactive tool draws results from a single model for each response variable and thus robustness of predictions are reduced.
Prior to tool creation a prototype determined functionality of interface and outputs on the proposed interface for non-experts and target users. This determined that the interface was easy to navigate and outputs easy to interpret.
The final online tool was tested by a novice user, by an expert and the author.
Downie et al. 1999,
Wilson et al. 2003
Scottish/ UK case study examples The interactive tool was based on data collected from across Scotland encompassing the main agricultural land uses in Scotland. Predictive models were generated from these data and the accuracy of these models tested.
Impact of changes to management practices (e.g. reduction in grazing intensity, creation of water margins) on spider and vascular plant richness. This information was combined with expert opinion to determine the impact of implementing management practices to promote biodiversity.
Tool outputs are of direct relevance to the BioBio Species Diversity Indicators Vascular Plants and Spiders
Downie et al. 1999,
Wilson et al. 2003,
McCracken 2000
McCracken 2000




Herzog et al. 2012
Examples of integrated use The biodiversity calculator has the potential to generate metrics for use in cost-benefit analyses (e.g. to explore the synergies and trade-offs when implementing different adaptation or agri-environment options).

A2.6.3 Eco-Serve GIS

Model/tool name Eco-Serve GIS  
    References
Impacts assessed GIS based Toolkit that generates spatial maps for nine ecosystem services (i.e. Accessible Nature, Carbon Storage, Local Climate Regulation, Water Purification, Air purification, Noise regulation, Education, Green travel and Pollination). Maps illustrate both requirement for each service (i.e. human demand) and capacity to deliver that service. Multi-functionality of delivery across ecosystem services are also assessed. Winn et al. 2015a
Sectors covered Agriculture, Forestry, Peatlands
Geographical scope England, Scotland and Wales
Modelling approach EcoServ- GIS uses simplified and generalised models of the relationships between landscape variables and ecosystem services. Ecosystem Service Capacity is determined by identifying habitats/ecosystems that provide a particular service and giving these a grade based on their capacity to provide that service. Demand for each service is also graded based on both the number of beneficiaries and the potential benefits derived. Both demand and capacity grades range from low to high (1 to 100) and are relative to the study area in question.
The multi-functionality toolbox creates a multi-functionality score based on the proportion of services that are met.

Main model outputs The toolkit creates a series of ecosystem service maps (including both requirement and delivery), multi-functionality maps, habitat maps, ecological connectivity maps and Biodiversity Opportunity Areas. The resulting maps are visually interpreted to determine where ecosystem services occur, and indicating where there is relatively high demand for a service, or high capacity to deliver a service.
This tool is designed for simultaneously comparing several ecosystem services. The following metrics are calculated: mean capacity, mean demand, mean GI assets capacity, multi-functionality score, priority multi-functionality score, number of Ecosystem Service Benefiting Areas, and number of Management Zones.

Main data needs OS MasterMap data. Potential to incorporate a range of other datasets (e.g. Digital Terrain Models, Core paths, Native Woodland Scotland Survey). Incorporation of additional spatial datasets will increase the number of ecosystem services that can be mapped.
Software requirements: ArcGIS Desktop (version 10.2.2), an Advanced level license with the Spatial Analyst extension.

Main limitations Ecosystem services mapped are restricted to the nine services outlined above. Mapping output is influenced by the underlying accuracy and resolution of the input spatial datasets. Many of the ecosystem services relate to populated areas and the toolkit is therefore less applicable to remote/non-urban areas (e.g. upland landscapes).
The resultant ecosystem service maps do not attempt to quantify the actual level of service delivery/demand but instead provides a relative measure for the target area. It is therefore not applicable to compare maps from different target areas. Information is largely not incorporated on habitat, or ecosystem quality.
Outputs are based on relatively simple models and capacity and demand only provide a proxy for service provisioning.
Limitations are dependent on the ecosystem service in question.
Winn et al. (2015a) suggest that alterative tools such as InVEST are more suitable. InVEST has not been as extensively tested in the UK. Interpretation of maps for use in decision making requires expert opinion and should consider other information.





















Winn et al. 2015a
Validation/
robustness
The toolkit was developed in Durham, NE England and subsequently tested in the South Downs National
Park and NIA, the Nene Valley NIA (Northamptonshire) and within Somerset.
Reliability of service maps range from Low in the case of Pollination to High in the case of Education and Accessible nature.
Winn et al. 2015a
Scottish/ UK case study examples The EcoServ- GIS toolkit was used to evaluate the multiple benefits derived from green networks in the Cumbernauld Living Landscape project. Identifying the most valuable green networks with respect to the delivery of multiple ecosystem services and helping to define management priorities for each area.
Outputs from this toolkit are relevant to the BioBio Species Diversity Indicator Bees and Habitat Diversity
Winn et al. 2015b




Herzog et al. 2012
Examples of integrated use Potential to incorporate output ecosystem service maps with other spatial datasets such as suitability mapping for native woodland creation and changes to distribution of 'prime' land under climate change.
The digital habitat map produced from the Eco-Serve GIS can be used to produce automated ecological network maps (e.g. thus potential integration with SNH's integrated habitat network modelling tools) and to map biodiversity opportunity areas.
Towers & Sing 2012, Brown et al. 2008

A2.6.4 AgBioscape

Table 50 Model description: AgBioscape

Model/tool name AgBioscape  
    References
Impacts assessed AgBioscape is a GIS based modelling system that simulates interactions between a range of target focal species (e.g. crop pests, natural predators and farmland birds), crop, management and landscape characteristics. Model simulations explore the impact of pre-determined cropping, field and landscape modifications on, for example, pest and natural predator populations across time. This can help to determine optimum modifications (e.g. those resulting in the lowest pest or highest predator densities). Begg 2013, http://www.pure-ipm.eu/
Sectors covered Agriculture
Geographical scope Simulated landscapes.
Modelling approach Discrete time population models (e.g. pest population dynamics) are combined with spatially explicit simulated agricultural landscapes. A matrix population modelling approach is used to spatially simulate the population dynamics of local populations over time. A series of land use/management scenarios are assigned to a simulated agricultural landscape to enable the user to alter landscape metrics (e.g. area and location of hedgerows) and to specify temporal changes in landscape structure (e.g. cropping patterns).The population matrix and simulated agricultural landscapes are overlaid. Transition matrices are used to specify demographical changes in life cycle stages that occur over time as a function of interactions (both within and between species), habitat, landscape and environmental conditions.
Main model outputs Model produces a series of spatially explicit simulated populations (e.g. pest population density) over time based on the specific scenarios inputted (e.g. different crop rotations). Metrics can be obtained from these scenarios (e.g. annual aphid population densities over a 100 year period) to compare scenarios.
Main data needs Modelling requires information on target species ecology (e.g. specificity of pest species, dispersal, life-history information, habitat specific survival rates).
Main limitations Outputs of models are dependent on the availability and reliability of ecological data on target species. Accuracy of ecological inputs will impact model predictability.
Modelling does not take into account impact of habitat quality on target species ecology (e.g. survival rate). Model currently is based on simulated landscapes.
Model is currently not openly available.

Validation/
robustness
AgBioscape model outputs are largely consistent with empirical findings highlighting the influence of landscape composition and crop management on crop-pest systems.
Modelling has, however, only been conducted on simulated landscapes without ground truthing on actual landscapes.
Begg 2013
Scottish/ UK case study examples AgBioscape was used to compare different rotational control strategies for the maize pest Diabrotica virgifera virgifera. Simulated models were used to evaluate strategies for control of cereal aphids by parasitic wasps.
AgBioscape was also used to explore the impact of agri-environment prescriptions on populations of farmland birds, crop pests and natural predators over a 350 year period.
Outputs from this toolkit are relevant to the BioBio Species Diversity Indicator Bees and Habitat Diversity.
Begg 2013
Begg & Dye 2015
Herzog et al. 2012
Examples of integrated use Development of the AgBioscape modelling approach could assist in the development of a decision support tool for land-managers/ policy makers/agricultural advisors. This tool could explore different scenarios with respect to the placement and nature of agri-environment schemes/compulsory greening measures and to spatially determine optimum configurations for pest regulation and/or biodiversity.

A2.7 Models and tools for animal health and animal welfare ( WI12)

Animal health and/or animal welfare are likely to be affected by many MOs. The animal health modelling literature is substantial, usually specific for certain diseases, livestock species and management/treatment. No models or tools were found for assessing the general health or welfare impacts.

A2.8 Models and tools for crop health ( WI13)

Precision farming ( MO2) and Optimal soil pH ( MO3): In principle, the measures that affect the productivity of the crop and therefore may have an impact on the crop health can be assessed with dynamic deterministic models of crop growth combined models of the soil carbon, nitrogen and water cycle. The measures identified include optimizing pH, and precision farming which are operating by increasing the nutrient supply to the crop. In general, the models have not been validated against data from crops receiving low levels of fertilizer nitrogen, and therefore there is a tendency for the yield predictions to be less reliable.

MO6 Incorporating legumes in grass mixes/ crop rotations: Many of the dynamic and deterministic crop models (e.g. APSIM, DSSAT etc.) can be used to model crop rotations, and do simulate the sequence effects where one crop influences the environment under which the following crop grows and hence affects the yield. However, these models do not consider the effects of the accumulation of soil borne diseases and weeds and thus the impact these will have on yield. The approaches that consider the break crop effect of legume on the rotation are either rule-based (Rule based rotation generator) or a combination of models which describe the effects of the sub-components (e.g. LUSO).

A2.8.1 APSIM

Table 51 Model description: APSIM

Model/tool name APSIM  
    References
Impacts assessed (1) The yield loss based on expected yield in the absence of disease and expected disease effects on leaf area duration
(2) The effect of eyespot on green leaf area and the yield (eyespot model being developed)
Poole & Arnaudin 2014
Al-Azri et al. 2015
Sectors covered Crop production
Geographical scope World wide
Modelling approach Dynamic deterministic model of crop and soil processes
Main model outputs Yield, green leaf area, leaf area duration, N 2O, leaching changes in soil carbon https://www.apsim.info/Documentation.aspx
Main data needs Daily weather data, soils characteristics, management of the crop, (1) expected effect on the leaf area duration, (2) data required to predict of disease development in relation to crop growth stages Poole & Arnaudin 2014, Al-Azri et al. 2015
Main limitations The model has been validated for typical management practices
Validation/ Robustness APSIM has been used extensively across the world to predict yields
Scottish/ UK case study examples Development of eyespot model Al-Azri et al. 2015
Examples of integrated use Green leaf retention calculator Poole & Arnaudin 2014

A2.8.2 DSSAT

Table 52 Model description: DSSAT

Model/tool name DSSAT  
    References
Impacts assessed The effect of disease on the crop is a required input to the model. Therefore the model assesses the effect on yield from a level of disease severity. http://abe.ufl.edu/jjones/ABE_5646/Week%207/Pest%20Module%20from%20DSSAT4%20Volume%204.pdf
Sectors covered Crop production
Geographical scope World wide
Modelling approach Dynamic deterministic model of crop and soil processes
Main model outputs Yield, green leaf, N 2O, leaching changes in soil carbon http://dssat.net/
Main data needs Daily weather data, soils characteristics, management of the crop, impact of the disease on the green leaf area http://dssat.net/ http://abe.ufl.edu/jjones/ABE_5646/Week%207/Pest%20Module%20from%20DSSAT4%20Volume%204.pdf
Main limitations The model has been validated for typical management practices
Validation/
robustness
DSSAT and the family of crop models embedded in the framework have been used extensively across the world to predict yields, soil carbon & nitrogen flows http://dssat.net/
Scottish/ UK case study examples Used to predict potato yield under climate change Daccache et al. 2011a & 2011b
Examples of integrated use No information

A2.8.3 LUSO

Table 53 Model description: LUSO

Model/tool name LUSO (The Land Use Sequence Optimiser)  
    References
Impacts assessed Optimizes the crop rotation, based on any expected seasonal and price situation. The model describes the effects of weeds and diseases on the crop rotation. It also describes the nitrogen contribution the legume makes to the following crop as a fertilizer equivalent. Lawes & Renton 2010
Sectors covered Crop rotations that include cereals and legumes
Geographical scope Developed for Australian farming conditions
Modelling approach Nitrogen - rule based Weeds - based on the RIM model that describes seedbank dynamics.
Main model outputs The effect of the nitrogen cost, weeds and diseases on the profitability of the cropping sequence.
Main data needs Length of the sequence, details on the weed seedbank, and the weed population dynamics, nitrogen costs, soil nitrogen status, soil disease population and details on the costs.
Main limitations Developed for Australian systems, and would need UK specific data (like nitrogen application to crops, weed prevalence and disease burden).
Validation/ robustness No information
Scottish/ UK case study examples No information
Examples of integrated use No information

A2.8.4 ROTOR

Table 54 Model description: ROTOR

Model/tool name ROTOR (Rule based rotation generator)  
    References
Impacts assessed Evaluates a range of feasible rotations on gross margins, leaching losses, fertilizer requirements and N 2O emissions. Reckling et al. 2016a, 2016b
Sectors covered Crop rotations.
Geographical scope EU
Modelling approach IPCC assessment of the leaching losses and N 2O emissions.
Main model outputs leaching losses, fertilizer requirements and N 2O emissions
Main data needs Agronomists define input variables such as crops, restriction values, and describe environmental, economic and phytosanitary indicators of the crops within the rotations.
Main limitations Based on expert opinion.
Validation/
robustness
Based on the judgment of experts. At this stage the inputs may be revised and the model re-run.
Scottish/ UK case study examples Used to assess Scottish rotations as part of the EU project Legume Futures.
Examples of integrated use No information

A2.9 Models and tools for economic impacts ( WI14- WI16)

In assessing both the farm and off-farm wider economic effects (both co-benefits and adverse side effects) multi-sectoral economic models of Scotland exist which can quantify all of these effects, and separately distinguish by sector and activity where relevant. In principle the models below can explore and quantify the qualitative and quantitative consequences of a host of MOs - and the details below give examples of such uses. In each case however, it is appropriate to ensure that the modelling system being used is able to reflect important aspects of the economic question being addressed in that use. For example, models of a single "Agriculture" sector - but with multiple non-agricultural activities identified - can be useful for qualitative descriptions of within and outwith agriculture effects, but are unable to capture what might be important heterogeneity within that sector; models which appropriate consider land use and competing uses would be appropriate for exploring cases where there might be alternative uses of this factor of production.

A2.9.1 CGE models

Table 55 Model description: CGE models

Model/tool name CGE models (e.g. AMOS)  
    References
Impacts assessed System-wide consequences of exogenously determined policy/non-policy options and disturbances Harrigan et al. 1991
Sectors covered All industrial sectors of economy, which could be separately identified at level of policy interest (with sufficient data and disaggregation). Current IO accounts for Scotland provide, for example, 98 sectors using Standard Industrial Classification 2007, mapping to national economic accounts. Scottish Government 2016
Geographical scope AMOS model framework has been applied to single region/nation analysis and inter-regional analysis. Application framework has been applied based on availability of model inputs, see below. Jersey: Learmonth et al. 2007
Scotland: FAI & Macaulay & Arkleton 2003, Lecca et al. 2014a
UK: Allan et al. 2007a
Inter-regional UK: Gilmartin et al. 2013
Modelling approach CGE model solves for equilibria in all markets for all goods and factors of production simultaneously. Comparative static or dynamic framework can show impacts in conceptual or annual time periods, and trajectory of variables between equilibria. Framework flexible to consider alternative model specifications, and so adapt to specific focus of application. Lecca, McGregor and Swales 2013
Lecca, Swales and Turner 2009
Main model outputs Economic variables (e.g. gross domestic product, aggregate employment, unemployment, household income) as well as sectoral levels of gross output, value-added, intermediate inputs, employment, and capital stocks. Also included are (endogenously determined) energy use (by sector), prices and costs of goods and factor inputs (including wages, return on capital). Energy use by sector is linked to CO 2 emissions, so that production-oriented measures of emissions are automatically tracked. (Consumption-oriented measures can be developed given appropriate trade-related data.)
Main data needs Uses IO and SAM as benchmark dataset for economic and sectoral structures, while behavioural specification and parameters appropriate for spatial scale and economy under consideration are required to configure relationships within and between markets. (These draw on new or existing econometric evidence.) Scottish Government 2016, Emonts-Holley et al. 2016
Main limitations Typically non-stochastic, calibrated to a single year's SAM and focus typically on policy simulation, not forecasting or historical analysis.
Validation/ robustness Tests on calibration accuracy; test simulations to check e.g. homogeneity properties of the model; extensive sensitivity analysis, drawing on statistical estimates where available; outputs subject to peer review.
Scottish/ UK case study examples Impact of onshore wind on rural and urban areas of North East Scotland Impact of expenditures related to establishing renewable energy capacity, including local content
System-wide impact of energy efficiency improvements in production sectors for Scotland and UK

System-wide impact of household energy efficiency improvements
Energy-economy-environmental impacts on Scotland of a carbon tax - "double dividend" and the importance of revenue recycling
Impact on Scotland of foot and mouth outbreak, 2001
Phimister and Roberts 2012
Gilmartin and Allan 2015; Allan et al. 2014a Hanley et al. 2006; Allan et al. 2007a; Anson and Turner 2009, Turner 2009 Lecca et al. 2014b
Allan et al. 2014b
FAI & Macaulay & Arkleton 2003
Examples of integrated use There are examples of CGEs having been combined with energy systems models and with micro-simulation models, but these are at a very early stage of development.

A2.9.2 IO and SAM models

Table 56 Model description: IO and SAM models

Model/tool name IO and SAM models
    References
Impacts assessed Changes in quantities or prices and system-wide consequences Miller and Blair 2009
Sectors covered All sectors of economy. For example, Scottish Input-Output tables are now (August 2016) available for years 1998 to 2013, covering 98 sectors and consistent with ESA 2010 (Scottish Government, 2016). Single "Agriculture" sector covering SIC2007 sector 01, with four sectors covering forestry and fishing activities. Disaggregation of sectors possible to focus on area of policy interest and address heterogeneity within industrial sectors, while disaggregation of categories of consumption permit examination of impacts across, e.g. household income types or household characteristics. Allan et al. 2007b
Geographical scope Local, regional or national (with inter-regional/inter-national configurations possible)
Modelling approach Static typically, deterministic, using inter-sectoral linkages to quantify system-wide impacts of changes in individual sectors or elements of demand or inputs.
Main model outputs Economic variables (e.g. gross domestic product, employment, household income, the sectoral levels of output, value-added, employment and capital stocks) as well as variables linked to sectoral output, including GHG emissions.
Main data needs IO accounts for regions/nations of interest showing production and consumption linkages between and within sectors and elements of consumption, e.g. households, exports, etc. Non-survey approaches allow estimation of IO accounts for smaller spatial levels, although (more time-consuming) survey-based approaches can capture more refined treatment of local differences in, e.g. linkages.
Main limitations Assumptions in modelling using IO include passive supply curve for all factors of production (no crowding out of activity) and that sectoral production inputs are combined in fixed proportions. (So typically motivated in terms of high unemployment and unused capacity in short-run, but in regional context also by factor mobility in long-run.) Miller and Blair 2009
Validation/ robustness -
Scottish/ UK case study examples Impact of community owned vs. community benefit-paying windfarm on the Shetland Islands, using SAM model for Shetland to show alternative impacts of locally-retained incomes from renewable energy project Impact of new onshore windfarm on farming households in north east Scotland Impacts of community wind power in rural areas in Scotland
Disaggregation of sea fishing sector to address heterogeneity of economic linkages within fishing fleet Review of economic multipliers for Scottish agriculture
Database of disaggregation of household types within SAM for Scotland 2009 IO accounts used to examine economic value of services produced by specific sectors for region/nation
Impacts of changes of forestry and afforestation on Scotland and UK

Uses and approach of IO/ SAM modelling in context of new biofuels production, including treatment of land in such models.
Allan et al. 2011


Phimister and Roberts 2012 Okkonen and Lehtonen 2016 Seafish 2006
Scottish Government 2010 Ross 2016
Cambridge Econometrics 2005, 2008 McGregor and McNicholl 1992, Eiser and Roberts 2002 Allan 2015
Examples of integrated use IO/ SAM database are used as the benchmark datasets and inputs to CGE models, which is a more flexible framework for exploring the range of factor supply assumptions and production structures, of which IO/ SAM are a special case. Extensions of IO/ SAMs to incorporate energy and environmental variables are common.

A2.10 Models and tools for resource efficiency ( WI17)

The energy and material recycling and resource use efficiency impacts arise from the improved utilization of nitrogen, energy and other resources on farm for the agricultural production related MOs ( MO1- MO9). These changes can be captured by the models and tools developed to estimate the GHG emissions and emission intensity of livestock and crop production, like whole-farm models life cycle assessment tools and carbon calculators (e.g. AGRILCA (Williams et al. 2006), AgRECalc ( http://www.agrecalc.com/) or CoolFarmTool ( https://www.coolfarmtool.org/)). The challenge with whole-farm approaches is the derivation of national level assessment from the farm-level models. On the other hand, no national level models were found which could capture the management changes implied by the implementation of MO1- MO9.

The resource use impacts of Reduced livestock product consumption and Afforestation can be estimated via economy wide models (see Section A2.9) if they are capable of tracking biomass, energy and nitrogen flows.

A2.11 Models and tools for human health ( WI18)

The reviewed GHG MOs can affect human health in various ways, from a reduction in water and air pollutants to a change in the diet and exercise level or an increase in antimicrobial resistance. Below is a list of the health impacts based on Section 3:

  • NH 3 emissions: MO1, MO2, MO4, MO5, MO7, MO8
  • NO x emissions: MO1, MO4
  • PM emissions: MO1, MO4, MO5, MO11
  • H 2S emissions: MO8
  • N leaching: MO2, MO7
  • P leaching: MO2
  • Release of pesticides and other chemicals to water: MO2, MO5, MO9, MO10, MO11
  • Heavy metals in the soil: MO3
  • Zoonosis: MO9
  • Antimicrobial resistance: MO9
  • Risk from handling acids: MO8
  • Diet: MO10
  • Exercise and mental health: MO11, MO12
  • Noise: MO1

The effects of air pollution ( NH 3, sulphur dioxide and PM 10) on human health have been explored and monetised, and they are included in the damage costs values used in the UK (Defra 2011a).

The human health impacts from nitrate pollution of watercourses and eventually drinking water consist of risk of metheamoglobinaemia and risk of cancer from nitrite-derived carcinogenic compounds. Though some estimates for these effects are available (van Grinsven et al. 2010), no model was found to assess this risk.

Though models exist to predict the risk of high-pesticide exposure of agricultural workers (Mage et al. 2000), no model was found which could assess the pesticide exposure of the general population.

As pH can affect plant-absorbable metal concentrations (e.g. lead, copper, zinc, nickel, aluminium) in soils ( Section A1.3), maintaining an optimal soil pH ( MO3) might decrease the risk of excessive consumption of these materials from crops. The CLEA software (Jeffries 2009) is a tool used by the Environment Agency to assess soil contamination risks; however, as it only covers home-grown produce it was not included in the assessed models. A tool suitable for assessing the risk of metal exposure as depending on soil pH for commercial agricultural land was not found. Similarly no models or tools were found for assessing the health risk arising from exposure to strong acids and H 2S (related to MO8: Low emission storage and application of organic fertiliser).

Improving animal health ( MO9) might decrease zoonosis incidents but could contribute to the prevalence of antimicrobial resistance ( Section A1.9). Though the literature on the various vectors' prevalence, their control mechanisms and the human health risk is wide (Lloyd-Smith et al. 2009), and estimates to the total aggregate human health effects and costs of selected pathogens exists for some countries (Lake et al. 2010, Scallane et al. 2015, Scharff 2012) an integrated tool linking livestock management and human illness prevalence was not found. As for the use of antimicrobials in the livestock sector and the potential effects on human health currently available data do not allow the quantification of these relationships (Rushton et al. 2014).

Assessing the health impacts of a change in diet ( MO10) is possible and already done by comparative risk assessment models ( Section A2.11.1).

Finally, the potential effects of afforestation ( MO11) on human health (arising from increased exercise levels and benefits to mental health - as opposed to the air purification effects of trees) have been explored in England, deriving per ha values for woodlands (based on woodland quality and proximity to urban areas) (Bateman et al. 2011). Nevertheless, a tool to assess these impacts was not found.

A2.11.1 DIETRON and PRIME

Table 57 Model description: DIETRON and PRIME

Model/tool name DIETRON and PRIME (Preventable Risk Integrated ModEl)  
    References
Impacts assessed Impact of diet on cardiovascular disease and cancer mortality (the models are being expanded to include physical activity, smoking and alcohol consumption) Scarborough et al. 2012b, Smed et al. 2016
Sectors covered Agriculture
Geographical scope UK
Modelling approach Comparative risk assessment: association between food components and coronary heart disease, stroke, cancer derived from individual meta-analyses (sugars not included as meta-analysis were not available) Scarborough et al. 2012b
Main model outputs Mortality and costs to NHS Scarborough et al. 2010
Main data needs Baseline and alternative diet composition
Main limitations The correlation between health effects are not included (e.g. serum cholesterol and BMI or fruit and vegetables and dietary fibre), therefore some overestimation is possible; assumes a linear dose-response relations; a shift in an average diet is modelled (no disaggregation allows for dietary groups) Scarborough et al. 2012b
Validation/ robustness No information
Scottish/ UK case study examples UK GHG emission based food taxes UK healthy diets Briggs et al. 2013, Scarborough et al. 2010
Examples of integrated use Similar comparative risk assessment models are linked a detailed agricultural modeling framework ( IMPACT (the International Model for Policy Analysis of Agricultural Commodities and Trade)) and to a life cycle GHG model Springmann et al. 2016a & 2016b

A2.12 Models and tools for social and cultural impacts ( WI19, WI20)

Tools or models to assess the social impacts of the MOs were not found.

Cultural impacts can be classified following the ecosystem services approach, whereby cultural ecosystem services are usually grouped as aesthetic, spiritual, educational and recreational services (Millennium Ecosystem Assessment 2005). Recreational impacts are the most studies of these, particularly in relation to greenspaces. This is of relevance to the MOs Afforestation ( MO11) and Peatland restoration ( MO12), in some cases possibly to MO5 (Agroforestry) as well.

A2.12.1 ORVal

Table 58 Model description: ORVal

Model/tool name ORVal (The outdoor recreation valuation tool)  
    References
Impacts assessed Recreational benefits - afforestation, peatland restoration http://leep.exeter.ac.uk/orval/
Sectors covered Recreation sites
Geographical scope England
Modeling approach A map-based tool using a statistical model of recreational demand (person-level model aggregated to England) Day and Smith 2016
Main model outputs Welfare values of currently accessible and proposed greenspaces (individual site level or aggregated by regions)
Main data needs Map of proposed recreation sites, data on their characteristics
Main limitations

Validation/ robustness No available information
Scottish/ UK case study examples Developed for England
Examples of integrated use


Contact

Email: Debbie Sagar