Input data set: N-Fertilizer 15crops (ISIMIP4)

Description:

Annual nitrogen fertilizer application rate for each of the fifteen ISIMIP crop groups — temperate cereals, rice, maize, the oil crops, pulses and the remaining groups — in kg N ha⁻¹ yr⁻¹ over the crop season, on the ISIMIP 0.25° and 0.5° grids, with an annual time step.

The rate applies per hectare of the respective crop group, not per hectare of grid cell. To obtain a nitrogen mass, multiply by that crop group's area, which is provided by the companion landuse-15crops dataset. Grid cells that contain no area of a crop group carry a missing value for that group's rate.

LUH3 provides one fertilizer rate per crop functional type and no distinction between the crops within a type. Applying that single rate to every crop in a type would erase real agronomic variation, since nitrogen-fixing legumes need substantially less synthetic nitrogen than cereals, so this dataset redistributes the functional-type rate across the crop groups it contains using a crop-specific relative intensity derived from NPKGRIDS. The relative intensity is normalised so that its area-weighted average across a functional type is exactly one, which makes the redistribution nitrogen-conserving. The two perennial functional types are themselves single crop groups and are passed through unchanged, identical to n-fertilizer-5crops.

Users needing the nitrogen mass per crop functional type to close exactly should use n-fertilizer-5crops instead.

A method description covering both n-fertilizer-5crops and n-fertilizer-15crops can be found here: https://docs.google.com/document/d/1Q77KZniheAL9G8Iyegcjk-WU5GDlmKeB3EMwLF9z9Ng/edit?usp=sharing

Derived from the CMIP7 input4MIPs UofMD Land-Use Harmonization (LUH3) dataset, using the multiple-management fertilizer rates together with the multiple-states crop-type area fractions.
Project page: https://luh.umd.edu/

Chini, L. and Hurtt, G.: Land-Use Harmonization 3 v1.2 for CMIP7 Historical (LUH3 v1.2), source_id UofMD-landState-3-1-2, https://doi.org/10.5281/zenodo.19261724 — used for histsoc, 1850soc, 1901soc and 2021soc

Chini, L. and Hurtt, G.: Land-Use Harmonization 3 Future Scenario VL v1.1 for ScenarioMIP (LUH3-VL v1.1), source_id UofMD-landState-vl-3-1-1, https://doi.org/10.5281/zenodo.19353397 — used for ssp1vlsoc-noadapt

Chini, L. and Hurtt, G.: Land-Use Harmonization 3 Future Scenario H v1.1 for ScenarioMIP (LUH3-H v1.1), source_id UofMD-landState-h-3-1-1, https://doi.org/10.5281/zenodo.19354835 — used for ssp3hsoc-noadapt

All three were taken at input4MIPs publication version v20260408.

The crop-specific relative intensity between crop groups is derived from NPKGRIDS v1.08 — nitrogen, phosphorus and potassium application rates for 173 crops at 0.05°, representing conditions around 2015–2020 — area-weighted with crop areas from the companion CROPGRIDS v1.08 dataset, which shares its crop taxonomy.

Nguyen, Tang, Conchedda et al. (2024), NPKGRIDS, Scientific Data, https://doi.org/10.1038/s41597-024-04030-4
Tang et al. (2024), CROPGRIDS, Scientific Data, https://doi.org/10.1038/s41597-024-03247-7

References for the LUH methodology:
Hurtt et al. (2020), Geoscientific Model Development, https://doi.org/10.5194/gmd-13-5425-2020
Chini et al. (2021), Earth System Science Data, https://doi.org/10.5194/essd-13-4175-2021

The rates are not resolved sub-nationally. As in n-fertilizer-5crops, the underlying functional-type rate carries one value per country, crop functional type and year, so the total nitrogen this dataset assigns to a given country, functional type and year is that same national number. What NPKGRIDS adds is the allocation of that total between crop groups, which does vary within a country.

Check nitrogen conservation at 0.25°, not at 0.5°. Within each crop functional type, the area-weighted sum of the crop-group rates reproduces the functional-type rate of n-fertilizer-5crops to float precision on the native 0.25° grid. On the remapped 0.5° product it is only approximate, about −0.09 % globally, because this dataset and n-fertilizer-5crops each remap mass-conservatively against their own crop areas (group area here, functional-type area there) and the two weightings do not commute with the regridding.

This dataset holds slightly less nitrogen than n-fertilizer-5crops. The 500 kg N ha⁻¹ yr⁻¹ bound is applied a second time after redistribution, since a group's intensity can legitimately exceed the functional-type average once redistributed. Where it binds, the excess nitrogen is clipped rather than reallocated. The resulting deficit is below 0.8 % of global nitrogen mass in every ISIMIP4b scenario and year, and below 0.3 % under ISIMIP4a. The second bound avoids unrealistic rates in edge cases.

The two resolutions do not give the same global totals. The 0.5° product is conservatively remapped from the native 0.25° grid and renormalized over the land area of each coarser cell, so a global nitrogen total computed from a 30arcmin file exceeds the same total from the 15arcmin file by about 1 %, concentrated in coastal cells. Use the 15arcmin files when the global mass matters.

NPKGRIDS' own global nitrogen total lies about 10 % below FAO/IFA figures, a limitation reported by its authors. It affects the relative allocation between crop groups, not the functional-type total, which comes from LUH3.

The pre-industrial scenarios are zero everywhere. 1850soc (ISIMIP4b) and 1901soc (ISIMIP4a) contain zero for all fifteen variables in every year. This is intended: synthetic nitrogen fertilizer did not exist at either base year (Haber-Bosch process only from ~ 1913), so holding the forcing at that level gives a no-fertilizer counterfactual. In the LUH3 source, application rates are exactly zero through 1910 and first become non-zero between 1910 and 1920.

2021soc means different things in the two rounds. Under ISIMIP4b it holds the 2021 rate fixed across the 2022–2100 future window; under ISIMIP4a it holds the same rate fixed across the full 1850–2025 historical period, as a present-day baseline.

Under ISIMIP4a the value for 2025 repeats 2024, the last observed year.