Impact model: H08

Sector
Water (global)
Region
global

H08 is a grid-cell based global hydrological model. It consists of six sub-models, namely land surface hydrology, river routing, reservoir operation, crop growth, environmental flow, and water abstraction. The formulations of sub-models are described in detail in Hanasaki et al. (2008a,b, 2010). In the standard simulation settings, H08 spatially covers the whole globe at a resolution of 0.5°×0.5° in order to assess geographical heterogeneity of hydrology and water use. Simulation period is typically for several decades and calculation interval is a day. The six sub-models exchange water fluxes and updates water storage at each calculation interval with the complete closure of water balance (the error is less than 0.01% of the total input precipitation). These characteristics enable us to explicitly simulate the major interaction between natural water cycle and major human activities of the globe. Source code and the manuals of H08 is open to public, available at http://h08.nies.go.jp.

In 2016, the water abstraction schemes of H08 has been substantially enhanced. In addition, a simple groundwater scheme was added to the land surface hydrology sub-model. It enabled us to estimate water abstraction from six major water sources, namely, streamflow regulated by global reservoirs (i.e. reservoirs regulating the flow of main channel of the world major rivers), aqueduct water transfer, local reservoirs, seawater desalination, renewable groundwater, and non-renewable groundwater. A model description paper is available at https://www.hydrol-earth-syst-sci.net/22/789/2018/.

H08 is one of the 13 global hydrology models following the ISIMIP2a protocol which form the base of simulations for the ISIMIP2a global water sector outputs; for a full technical description of the ISIMIP2a Simulation Data from Water (global) Sector, see this DOI link: http://doi.org/10.5880/PIK.2017.010

For ISIMIP2b, the new H08 (Hanasaki et al., 2018) was used. The model is substantially different from “classic H08” (Hanasaki et al. 2008a, 2008b, 2010) by mainly six points.
- Groundwater scheme was added.
- Groundwater abstraction scheme was added.
- Aqueduct water transfer scheme was added
- Scheme for return flow and delivery loss was added
- Reservoir scheme was updated
- Seawater desalination scheme was added

For ISIMIP3a/3b, the hydrological parameters of H08 have been tuned based on the study of Yoshida et al. (2022; https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2021WR030660 ). We believe now the overall water balance is better estimated compared to previous versions. Due to some technical reasons, the parameters for subarctic climate (Dc) vary by longitudinal sections. This may cause sudden jumps in hydrological variables (e.g. soil moisture) at the borders of sections. See Yoshida et al. (2022) for further details. The tuned parameters include soil depth, which requires special attention in analyzing H08 simulation results. The soil in the H08 global hydrologic model has long been configured as a uniform single layer of 1 m (Hanasaki et al. 2008). In ISIMIP Phase 3, that configuration was changed by adopting the climate zone-specific parameter optimization presented in Yoshida et al. (2022). Now, H08 has climate-zone-specific soil depth and maximum soil moisture holding capacity. These spatial distributions are shown in SoilDepth.nc and SoilMoisMax.nc. There are two caveats here. First, there is still only one soil layer in H08. Second, as mentioned above, based on the discussion in Yoshida et al. (2022), the climatic classification of Dc (subarctic) was further separated into six subregions. Because they are separated by meridians (e.g., 60°E or 90°E), the distribution of soil moisture and water fluxes at high latitudes in the Northern Hemisphere shows an artificial pattern.

For validation of the ISIMIP3a run, see Boulange et al. (2023).

Information for the model H08 is provided for the simulation rounds shown in the tabs below. Click on the appropriate tab to get the information for the simulation round you are interested in.

Person responsible for model simulations in this simulation round
Naota Hanasaki: hanasaki@nies.go.jp, 0000-0002-5092-7563, National Institute for Environmental Studies (Japan)
Kedar Otta: otta.kedar@nies.go.jp, 0000-0002-2540-9879, National Institute for Environmental Studies (NIES)) (Japan)
Output Data
Experiments: picontrol_1850soc_default, ssp370_1850soc_default, historical_2015soc_default, ssp126_2015soc_default, ssp585_1850soc_default, ssp126_1850soc_default, historical_histsoc_default, picontrol_2015soc_default, ssp585_2015soc_default, historical_1850soc_default, picontrol_histsoc_default, ssp370_2015soc_default
Climate Drivers: GFDL-ESM4, IPSL-CM6A-LR, MPI-ESM1-2-HR, MRI-ESM2-0, UKESM1-0-LL
Date: 2021-11-02
Basic information
Model Output License: CC0
Model Homepage: http://h08.nies.go.jp/
Reference Paper: Main Reference: Hanasaki, N., Yoshikawa, S., Pokhrel, Y., Kanae, S. et al. A global hydrological simulation to specify the sources of water used by humans. Hydrol. Earth Syst. Sci.,22,789–817,2018
Reference Paper: Other References:
Resolution
Spatial aggregation: regular grid
Horizontal resolution: 0.5°x0.5°
Vertically resolved: No
Number of vertical layers: One layer but the depth varies by climate zone. As of ISIMIP Phase 3, the soil layer configuration was changed by adopting the climate zone-specific parameter optimization presented in Yoshida et al. (2022). H08 now has one climate-zone-specific soil depth and maximum soil moisture holding capacity. Based on the discussion in Yoshida et al. (2022), the climatic classification of Dc (subarctic) was further separated into six subregions. Because they are separated by meridians (e.g., 60°E or 90°E), the distribution at high latitudes in the Northern Hemisphere shows an artificial pattern. The spatial soil depth distributions are given in the attachment below.
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
Simulated atmospheric climate data sets used: MRI-ESM2-0, IPSL-CM6A-LR, MPI-ESM1-2-HR, UKESM1-0-LL, GFDL-ESM4
Land use data sets used: Historical, gridded land use
Climate variables: huss, sfcWind, tas, rlds, rsds, prsn, ps, pr
Exceptions to Protocol
Exceptions: H08 does not have “2015soc-from-histsoc” simulations as 2015soc simulations start with spinoffs. This is because the current version of H08 doesn’t have a “restart” function, hence we cannot start a simulation using the end values of another simulation. Similarly, all future period simulations start with spin-up in 2015 instead of continuing from the historical period. However, for pi-controlled simulations, the future period is in continuation from the historical period.
Spin-up
Was a spin-up performed?: Yes
Spin-up design: Iterate the first year of simulation until the hydrological state variables (i.e. soil moisture) reach equilibrium.
Soil
Soil layers: 1 layer climate-zone-specific soil depth. Check reference paper for further information.
Water Use
Water-use types: water withdrawal, consumption (i.e. evaporation), and return flow. Surface water abstraction (river, reservoir, inter-grid-cell transfer, etc.) Groundwater abstraction (renewable and non-renewable).
Water-use sectors: Irrigation, Industrial, and Domestic
Routing
Runoff routing: Based on the TRIP routing scheme (Oki et al. 1999; Journal of the Meteorological Society of Japan)
Routing data: DDM30
Land Use
Land-use change effects: Irrigation area expansion.
Dams & Reservoirs
Dam and reservoir implementation: Reservoir operation scheme of Hanasaki et al. (2006).
Calibration
Was the model calibrated?: Yes
Which years were used for calibration?: 1961-1970
Which dataset was used for calibration?: W5E5
How many catchments were callibrated?: 777. Then optimal parameter set was determined for each climatic zone (see Yoshida et al. 2022, WRR)
Vegetation
Is co2 fertilisation accounted for?: No
How is vegetation represented?: We assume the Earth's land surface was covered by grassland uniformly (NB this is an assumption of the standard Bucket Land Surface Model). Then, different hydrological parameters were set for each climatic zone. This treatment implicitly reflects the dominant vegetation coverage of each climatic zone on the model. For further discussion, please see Yoshida et al. (2022, WRR)
Methods
Potential evapotranspiration: Solving the energy balance at the surface.
Snow melt: Solving the energy balance at the surface and estimate the snow melt rate.
Attachments
666-H08_SoilDepth.nc (1017.3 KB) H08 Soil depths in ISIMIP3
666-H08_SoilMoisMax.nc (1017.2 KB) H08 Maximum soil moisture in ISIMIP3
Person responsible for model simulations in this simulation round
Naota Hanasaki: hanasaki@nies.go.jp, 0000-0002-5092-7563, National Institute for Environmental Studies (Japan)
Kedar Otta: otta.kedar@nies.go.jp, 0000-0002-2540-9879, National Institute for Environmental Studies (NIES)) (Japan)
Output Data
Experiments: obsclim_histsoc_default, counterclim_histsoc_default, obsclim_histsoc_1901irr, counterclim_2015soc_default, obsclim_2015soc_default, counterclim_1901soc_default, obsclim_1901soc_default
Climate Drivers: 20CRV3, 20CRV3-ERA5, 20CRV3-W5E5, GSWP3-W5E5
Date: 2021-12-01
Basic information
Model Output License: CC0
Model Homepage: https://h08.nies.go.jp/
Model License: Apache License, Version 2.0
Reference Paper: Main Reference: Hanasaki, N., Yoshikawa, S., Pokhrel, Y., Kanae, S. et al. et al. A global hydrological simulation to specify the sources of water used by humans. 2017
Reference Paper: Other References:
Resolution
Spatial aggregation: regular grid
Horizontal resolution: 0.5°x0.5°
Vertically resolved: No
Number of vertical layers: One layer but the depth varies by climate zone. As of ISIMIP Phase 3, the soil layer configuration was changed by adopting the climate zone-specific parameter optimization presented in Yoshida et al. (2022). H08 now has one climate-zone-specific soil depth and maximum soil moisture holding capacity. Based on the discussion in Yoshida et al. (2022), the climatic classification of Dc (subarctic) was further separated into six subregions. Because they are separated by meridians (e.g., 60°E or 90°E), the distribution at high latitudes in the Northern Hemisphere shows an artificial pattern. The spatial soil depth distributions are given in the attachment below.
Temporal resolution of input data: climate variables: daily
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)
Land use data sets used: Historical, gridded land use
Climate variables: huss, sfcWind, tas, rlds, rsds, prsn, ps, pr
Exceptions to Protocol
Exceptions: The current version of H08 doesn’t have a “restart” function, hence we cannot start a simulation using the end values of another simulation. Therefore, spin-up is provided for the first year of each "Experiment"
Spin-up
Was a spin-up performed?: Yes
Spin-up design: Iterate the first year of simulation until the hydrological state variables (i.e. soil moisture) reach equilibrium.
Natural Vegetation
Soil layers: 1 layer
Soil
Soil layers: 1 layer climate-zone-specific soil depth. Check reference paper for further information.
Water Use
Water-use types: water withdrawal, consumption (i.e. evaporation), and return flow. Surface water abstraction (river, reservoir, inter-grid-cell transfer, etc.) Groundwater abstraction (renewable and non-renewable).
Water-use sectors: Irrigation, Industrial, and Domestic
Routing
Runoff routing: Based on the TRIP routing scheme (Oki et al. 1999; Journal of the Meteorological Society of Japan)
Routing data: DDM30
Land Use
Land-use change effects: Irrigation area expansion.
Dams & Reservoirs
Dam and reservoir implementation: Reservoir operation scheme of Hanasaki et al. (2006).
Calibration
Was the model calibrated?: Yes
Which years were used for calibration?: 1961-1970
Which dataset was used for calibration?: W5E5
How many catchments were callibrated?: 777. Then optimal parameter set was determined for each climatic zone (see Yoshida et al. 2022, WRR)
Vegetation
Is co2 fertilisation accounted for?: No
How is vegetation represented?: We assume the Earth's land surface was covered by grassland uniformly (NB this is an assumption of the standard Bucket Land Surface Model). Then, different hydrological parameters were set for each climatic zone. This treatment implicitly reflects the dominant vegetation coverage of each climatic zone on the model. For further discussion, please see Yoshida et al. (2022, WRR)
Methods
Potential evapotranspiration: Solving the energy balance at the surface.
Snow melt: Solving the energy balance at the surface and estimating the snow melt rate
Attachments
709-H08_SoilDepth.nc (1017.3 KB) H08 Soil depths in ISIMIP3
709-H08_SoilMoisMax.nc (1017.2 KB) H08 Maximum soil moisture in ISIMIP3
Person responsible for model simulations in this simulation round
Naota Hanasaki: hanasaki@nies.go.jp, 0000-0002-5092-7563, National Institute for Environmental Studies (Japan)
Yoshimitsu Masaki: ymasaki.tskb.ibrk.305@gmail.com, Hirosaki University (Japan)
Output Data
Experiments: I, II, III, IV, V, VI, VII, VIII
Climate Drivers: None
Date: 2017-10-10
Basic information
Model Version: H08 (Hanasaki et al., 2018)
Model Output License: CC BY 4.0
Reference Paper: Main Reference: Hanasaki, N., Yoshikawa, S., Pokhrel, Y., Kanae, S. et al. A global hydrological simulation to specify the sources of water used by humans. Hydrol. Earth Syst. Sci.,22,789–817,2018
Reference Paper: Other References:
Resolution
Spatial aggregation: regular grid
Horizontal resolution: 0.5°x0.5°
Temporal resolution of input data: climate variables: daily
Temporal resolution of input data: co2: constant
Temporal resolution of input data: land use/land cover: annual
Temporal resolution of input data: soil: NA (H08 doesn't specify soil texture)
Additional temporal resolution information: co2 has no effect on the simualation
Input data
Land use data sets used: Historical, gridded land use (HYDE 3.2)
Additional input data sets: Following the formulation of new H08, only the reservoirs larger than 5000km2 of catchment area regulate river flow. The remaining reservoirs are aggregated into one hypothetical reservoir for each grid cell which acts as a storage pond independent from rivers.
Climate variables: tas, rlds, rsds, prsn, ps, pr
Exceptions to Protocol
Exceptions: - Since H08 does not separate industrial water into electricity and manufacturing, total industrial water is reported in manufacturing file. - Due to limitation in research resources, we have simplified the protocol as follows - PI Control: We took Option 2 (fix at 2005) for the pre-industrial and historical periods. - PI Control: Accidentally the original simulation terminated at the end of 2290/1/1. There might some gaps in fluxes and state variables between 2290 and 2291 since two simulation are not perfectly consistent. - RCP2.6 and RCP6.0: We took Option 2 (fix at 2005) for the historical periods - RCP2.6 and RCP6.0: We calculated historical and future periods independently: there might some gaps between the results of 12/31/1860 and 1/1/1861 and those of 12/31/2005 and 1/1/2006.
Spin-up
Was a spin-up performed?: Yes
Spin-up design: We repeated the simulation of 1661 until it reaches soil moisture equilibrium.
Management & Adaptation Measures
Management: Planting date was determined to obtain maximum yield under meteorological conditions for 1961-1990. The planting date was fixed throughout the simulation period.  Harvesting date was calculated in the model and changed with years according to meteorological conditions.
Extreme Events & Disturbances
Additional comments: Historical run of IPSL: At the end of Apr 2017, we found a major problem in the simulation set up (we fed wind speed for shortwave radiation!), and re-run all the simulation. We re-uploaded the simulation results on May 4 2017 Water balance problem: For some simulation runs, the water balance is not closed (~100km3/yr). As of May 2017, we are investigating the cause of this problem.
Technological Progress
Technological progress: No
Soil
Soil layers: 1-layerSoil information was not incorporated, but runoff properties vary with climate zones (see Appendix C of Hanasaki et al. (2008a) for details)
Water Use
Water-use types: irrigation (computed internally), domestic water (prescribed), industrial water (prescribed)
Water-use sectors: irrigation (computed internally), domestic water (prescribed), industrial water (prescribed)
Routing
Runoff routing: Based on DDM30.
Land Use
Land-use change effects: For historical run, temporal variations in cultivars and irrigated/rainfed crop land area were included in the simulation.
Dams & Reservoirs
Dam and reservoir implementation: Dams and reservoirs were implemented based on the GranD, but the new H08 subdivides reservoirs into two groups. One is 'global reservoir' which regulates the flow of major rivers. The other is 'local reservoir' which is a storage pond, not regulating river flow. For histoircal run, the construction year of global reservoir was taken into account.
Calibration
Was the model calibrated?: No
Vegetation
Is co2 fertilisation accounted for?: No
How is vegetation represented?: Each land cell is subdivided into four land use (double-cropping irrigated cropland, single-cropping irrigated cropland, rain-fed cropland and natural use). Crop growth depends on given meteorological conditions. [For nosoc run] Natural use: Globally uniform. No-specific land type is assigned, as known as Manabe's bucket. [For pressoc run] Croplands: Fixed cultivars according to HYDE/MIRCA at 2000 with fixed planting dates. Geographical distribution of irrigated/rainfed cropland was based on HYDE/MIRCA at 2000. We assumed the geographical distribution of double-cropping cultivation was based on Siebert et al. (2005). Natural use: Globally uniform. No-specific land type is assigned, as known as Manabe's bucket. [For varsoc run] Croplands: Changing cultivars according to HYDE/MIRCA with fixed planting dates. Geographical distribution of irrigated/rainfed cropland was based on HYDE/MIRCA at a given year. We assumed the geographical distribution of double-cropping cultivation was fixed throughout the period and based on Siebert et al. (2005). Natural use: Globally uniform. No-specific land type is assigned, as known as Manabe's bucket.
Methods
Potential evapotranspiration: Bulk formula
Snow melt: Energy balance
Person responsible for model simulations in this simulation round
Naota Hanasaki: hanasaki@nies.go.jp, 0000-0002-5092-7563, National Institute for Environmental Studies (Japan)
Yoshimitsu Masaki: ymasaki.tskb.ibrk.305@gmail.com, Hirosaki University (Japan)
Output Data
Experiments: historical
Climate Drivers: None
Date: 2016-05-02
Basic information
Model Version: H08
Model Output License: CC BY 4.0
Reference Paper: Main Reference: Hanasaki N, Kanae S, Oki T, Masuda K, Motoya K, Shirakawa N, Shen Y, Tanaka K et al. An integrated model for the assessment of global water resources – Part 1: Model description and input meteorological forcing. Hydrol. Earth Syst. Sci.,12,1007-1025,2008
Reference Paper: Other References:
Resolution
Spatial aggregation: regular grid
Horizontal resolution: 0.5°x0.5°
Temporal resolution of input data: climate variables: daily
Temporal resolution of input data: co2: constant
Temporal resolution of input data: land use/land cover: annual
Temporal resolution of input data: soil: NA (H08 doesn't specify soil texture)
Additional temporal resolution information: co2 has no effect on the simualation
Input data
Observed atmospheric climate data sets used: GSWP3, PGMFD v2.1 (Princeton), WATCH (WFD), WATCH-WFDEI
Climate variables: tas, rlds, wind, rsds, prsn, ps, pr
Additional information about input variables: For WATCH and WFDEI forcings, precipitation was separated into rainfall and snowfall according to Yasutomi et al. (2011).
Spin-up
Was a spin-up performed?: Yes
Spin-up design: We started hydrological simulation since 1901, long enough for stabilizing initial conditions.
Management & Adaptation Measures
Management: Planting date was determined to obtain maximum yield under meteorological conditions for 1960-1999. The planting date was fixed throughout the simulation period.  Harvesting date was calculated in the model and changed with years according to meteorological conditions.
Technological Progress
Technological progress: No
Soil
Soil layers: 1-layerSoil information was not incorporated, but runoff properties vary with climate zones (see Appendix C of Hanasaki et al. (2008a) for details)
Water Use
Water-use types: irrigation (computed internally), domestic water (prescribed), industrial water (prescribed)
Water-use sectors: [pressoc/varsoc runs submitted before Dec.2015] irrigation (computed internally), domestic water (prescribed), industrial water (prescribed)
Routing
Runoff routing: Based on DDM30.
Land Use
Land-use change effects: For varsoc run, time-varying cultivars and irrigated/rain-fed cultivation were included for croplands.
Dams & Reservoirs
Dam and reservoir implementation: Dams and reservoirs were implemented based on the protocol (GranD). For varsoc run, the construction year of each dam/reservoir was taken into accounted.
Calibration
Was the model calibrated?: No
Vegetation
Is co2 fertilisation accounted for?: No
How is vegetation represented?: Each land cell is subdivided into four land use (double-cropping irrigated cropland, single-cropping irrigated cropland, rain-fed cropland and natural use). Crop growth depends on given meteorological conditions. [For nosoc run] Natural use: Globally uniform. No-specific land type is assigned, as known as Manabe's bucket. [For pressoc run] Croplands: Fixed cultivars according to HYDE/MIRCA at 2000 with fixed planting dates. Geographical distribution of irrigated/rainfed cropland was based on HYDE/MIRCA at 2000. We assumed the geographical distribution of double-cropping cultivation was based on Siebert et al. (2005). Natural use: Globally uniform. No-specific land type is assigned, as known as Manabe's bucket. [For varsoc run] Croplands: Changing cultivars according to HYDE/MIRCA with fixed planting dates. Geographical distribution of irrigated/rainfed cropland was based on HYDE/MIRCA at a given year. We assumed the geographical distribution of double-cropping cultivation was fixed throughout the period and based on Siebert et al. (2005). Natural use: Globally uniform. No-specific land type is assigned, as known as Manabe's bucket.
Methods
Potential evapotranspiration: Bulk formula
Snow melt: Energy balance
Person responsible for model simulations in this simulation round
Naota Hanasaki: hanasaki@nies.go.jp, 0000-0002-5092-7563, National Institute for Environmental Studies (Japan)
Yoshimitsu Masaki: ymasaki.tskb.ibrk.305@gmail.com, Hirosaki University (Japan)
Output Data
Experiments: historical, rcp26, rcp45, rcp60, rcp85
Climate Drivers: None
Date: 2013-12-13