Impact model: LPJmL5-7-10-fire

This model version combines recent changes to the default LPJmL version (5.7.10) with recent updates to the Spitfire model, that are not yet part of the default version of LPJmL. Compared to earlier rounds of ISIMIP, this version includes a representation of the nitrogen cycle. Note that this version differs from the LPJmL version used for the agriculture sector in ISIMIP 3a & 3b.

ISIMIP3b
ISIMIP3a

Person responsible for model simulations in this simulation round

Sebastian Ostberg: ostberg@pik-potsdam.de, 0000-0002-2368-7015, Potsdam Institute for Climate Impact Research (Germany)

Additional persons involved: Luke Oberhagemann, Maik Billing & Markus Drüke (Spitfire update)

Basic information

Model Version: 5.7.10-fire

Reference Paper: Main Reference: Wirth S, Braun J, Heinke J, Ostberg S, Rolinski S, Schaphoff S, Stenzel F, von Bloh W, Müller C et al. Biological nitrogen fixation of natural and agricultural vegetation simulated with LPJmL 5.7.9. EGUsphere [preprint],None,,2024

Reference Paper: Other References: Oberhagemann L, Billing M, von Bloh W, Drüke M, Forrest M, Bowring S, Hetzer J, Ribalaygua Batalla J, Thonicke K et al. Sources of Uncertainty in the Global Fire Model SPITFIRE: Development of LPJmL-SPITFIRE1.9 and Directions for Future Improvements. EGUsphere [preprint],None,,2024

Resolution

Spatial aggregation: regular grid

Horizontal resolution: 0.5°x0.5°

Vertically resolved: Yes

Number of vertical layers: 5 hydrologically active soil layers

Additional spatial aggregation & resolution information: We use a mask prescribing the continental fraction of each cell. Use provided "cellarea" and "contfrac" outputs for aggregation over multiple cells.

Temporal resolution of input data: climate variables: daily

Temporal resolution of input data: co2: annual

Temporal resolution of input data: land use/land cover: annual

Temporal resolution of input data: soil: constant

Person responsible for model simulations in this simulation round

Sebastian Ostberg: ostberg@pik-potsdam.de, 0000-0002-2368-7015, Potsdam Institute for Climate Impact Research (Germany)

Additional persons involved: Luke Oberhagemann, Maik Billing & Markus Drüke (Spitfire update)

Output Data

Experiments: (*) obsclim_1901soc_1901co2, obsclim_2015soc_1901co2, obsclim_nat_default, obsclim_2015soc_default, counterclim_histsoc_default, counterclim_histsoc_obsco2, obsclim_histsoc_1901co2, counterclim_nat_default, counterclim_1901soc_default, counterclim_2015soc_default, obsclim_1901soc_default, obsclim_histsoc_default, obsclim_histsoc_nofire
Climate Drivers: 20CRV3, 20CRV3-ERA5, 20CRV3-W5E5, GSWP3-W5E5
Date: 2024-01-23

Basic information

Model Version: 5.7.10-fire

Model Output License: CC0

Model License: AGPL-3.0 license

Simulation Round Specific Description: Please cite both reference papers listed below since there is no single paper describing the model version used in ISIMIP3a.

Resolution

Spatial aggregation: regular grid

Horizontal resolution: 0.5°x0.5°

Vertically resolved: Yes

Number of vertical layers: 5 hydrologically active soil layers

Temporal resolution of input data: climate variables: daily

Temporal resolution of input data: co2: annual

Temporal resolution of input data: land use/land cover: annual

Temporal resolution of input data: soil: constant

Input data

Observed atmospheric climate data sets used: GSWP3-W5E5 (ISIMIP3a), 20CRv3, 20CRv3-ERA5, 20CRv3-W5E5

Emissions data sets used: Atmospheric composition (ISIMIP3a)

Socio-economic data sets used: National, gridded historical population

Land use data sets used: Historical, gridded land use

Other human influences data sets used: Reservoirs & dams, Nitrogen deposition (ISIMIP3), N-Fertilizer (ISIMIP3a), Water abstraction for domestic and industrial uses

Other data sets used: Lightning, Drainage direction map for river routing

Additional input data sets: Lake area fraction based on GLWD, modified version of GGCMI phase 3 crop calendar, CRU land-sea mask (corresponds to ISIMIP landseamask_water-global except for 1 cell), continental fraction based on GADM, soil data based on HWSD

Climate variables: huss, sfcWind, tasmax, tas, tasmin, rsds, pr

Additional information about input variables: long wave net radiation calculated from long wave downwelling radiation and mean temperature

Exceptions to Protocol

Exceptions: Landuse input provided by ISIMIP was extended further back in time for model spinup in "histsoc" experiments to avoid artefacts of long-running constant landuse. "1901soc" and "2015oc" experiments used constant landuse also for spinup as prescribed by protocol.

Spin-up

Was a spin-up performed?: Yes

Spin-up design: First spinup without any human forcing for 3500 years followed by transitional run of 400 years with human forcing (either constant (1901soc, 2015soc), or transient (histsoc) or no human forcing (nat)). No-fire sensitivity experiment also does not use fire during spinup and transitional run.

Natural Vegetation

Natural vegetation partition: Dynamic vegetation composition for natural vegetation

Natural vegetation dynamics: Bioclimatic limits determine whether PFTs can establish. Established PFTs compete for light, water and nitrogen and are also affected by disturbance, e.g. fire.

Soil layers: From top to bottom: 20 cm, 30 cm, 50 cm, 100 cm, 100 cm.

Management & Adaptation Measures

Management: Prescribed (constant) sowing dates and prescribed PHU requirements (transient over time) for annual crops. N fertilizer application according to ISIMIP input. Irrigation possible on irrigated areas according to landuse input and based on dynamically calculated irrigation requirements and water availability. Effects of tillage on croplands included in simulations. Prescribed constant livestock density on pastures. Note: Landuse fractions in ISIMIP input refer to the grid area and have been rescaled based on our continental area fraction mask.

Extreme Events & Disturbances

Key challenges: No effects of pests or water logging. Fire disturbance simulated on natural vegetation and pastures. No fires on cropland. Frost damage possible for annual crops.

Fire-specific input data sets

What input datasets are used in the fire model and what are they used for?: population density and an input describing human ignitions per individual per day to calculate human-based ignition events; lightning ignitions

What is the time step of the fire model?: daily

What is the time step of the exchange between fire and vegetation model? e.g. are carbon pools and cover fractions updated every day?: Fire mortality and PFT fractions are updated annually; fuels are updated daily.

Burnt Area

What are the main components of burned area computation?: Number of ignition events, fire danger (based on water vapour pressure deficit), rate of spread (based on fuel load characteristics and wind speed)

Ignition

Which sources of ignition are included?: lightning; human-caused

Is fire ignition implemented as a random process?: no

How are natural ignitions implemented? which data is used and how is it scaled?: daily fire danger index multiplied by daily lightning ignitions, assuming a cloud-to-ground ratio of lightning flashes of 0.2 and a lightning efficiency in starting fires of 0.04

Is human influence on fire ignition and/or suppression included? how?: No human fire suppression, local population density influences human-caused ignitions

If human ignitions are included for which conditions are the ignitions highest/lowest?: Human ignitions are lowest for very high and very low population density, human ignitions are highest for medium population densities

Spread and duration

How does fire spread?: Fire spread is calculated according to the Rothermel equation. Fires spread in elliptic shapes, depending on average wind speed.

How is fire duration computed?: Fire duration depends on model-internal fire danger index (sigmoid shape dependency), minimum and maximum are fixed model parameters

Fuel load and combustion

How does the model compute fuel load?: Fuel load calculations are based on the dynamic vegetation; dead fuels comprise carbon of dead trees, litter and cured grass; live fuel comprise living grasses

List of fuel classes (full names and abbreviations): 1hr (1h fuels), 10hr (10h fuels), 100 hr (100h fuels), 1000hr (1000h fuels), live grass (live grass)

Is fuel moisture linked to soil moisture/air humidity/precip?: Dead fuel moisture is linked to litter moisture; live fuel moisture (grasses) depends on a growing season index.

Which carbon pools are combusted?: all fuel classes

Is the combustion completeness constant or depends on what (fuel type, moisture?): Combustion depends on fuel type and surface fire intensity

Landcover

What is the minimum/maximum burned area fraction at grid cell level? over which time period? : maximum burned area fraction = 100% - inland water body fraction - cropland fraction

Land-cover classes allowed to burn: natural vegetation, pastures

Is burned area computed separately for each pft? if not how is burned area separated into the pft-burned area? : PFTs grow together on a stand/tile with a shared litter/dead fuel layer, and burned area is calculated for that stand/tile. Burned area is not separated into PFT-specific burned area.

Are peatland fires included?: no

Are deforestation or land clearing fires included?: no

How are pastures represented?: Only grass PFTs grow on prescribed pastures. Livestock grazing as pasture management. No specific fire management on pastures

If croplands burn, does the fire model differ for this pft? if yes please describe.: no fires on cropland

If pastures burn, does the fire model differ for his pft? if yes, please describe: no, same fire model as for natural vegetation

Fire mortality

Vegetation fire mortality: is it constant/constant per pft/depends on (for instance fire intensity, bark thickness, veg height): Post fire mortality depends on fire intensity, tree height and bark thickness