The variables sea surface temperature, surface salinity, surface pressure of CO2 and pH were obtained from fully coupled models in the Coupled Model Intercomparison Project 5 (CMIP5; http://pcmdi9.llnl.gov/ esgf-web-fe/). All model outputs were regridded to the same 1x1º regular grid using bilinear interpolation. If multiple runs were available for a model, these runs were averaged first. Then a multi model ensemble was created Missing values over land were filled in using a Poisson grid filter in the zonal direction. Values where sea ice was present in the GFDL-ESM2M_rcp85 model were masked out. Monthly values were averaged to get yearly or decadal averages. Aragonite saturation state was calculated from SST, surface pressure of CO2, pH, and salinity
References:
van Hooidonk, R. J., Maynard, J. A., Manzello, D., & Planes, S. (2014). Opposite latitudinal gradients in projected ocean acidification and bleaching impacts on coral reefs. Global Change Biology, 103–112. doi:10.1111/gcb.12394).

Sea ice area is computed from the ensemble mean of CMIP5 models projections for variable "sit" (Sea Ice Thickness). Models output are averaged by models (for each r0i0p0, r0i0p1, etc.), then regridded to a common grid, then models averages are averaged together. A grid point is considering having sea ice if the average thickness is not 0.

Ensemble mean computation: https://github.com/IOC-CODE/cmip5_projections/blob/master/make_ensembleM...
DHM and DHM-level 2 frequency computation: https://github.com/IOC-CODE/cmip5_projections/blob/master/make_dhm.py

Seafloor organisms are vital for healthy marine ecosystems, contributing to elemental cycling, benthic remineralisation and ultimately sequestration of carbon. Deep-sea life is primarily reliant on the export flux of particulate organic carbon from the surface ocean for food, but most ocean biogeochemistry models predict global decreases in export flux resulting from 21st century anthropogenically-induced warming. Here we show that decadal-to-century scale changes in carbon export associated with climate change lead to an estimated 5.2% decrease in future (2091-2100) global open ocean benthic biomass under RCP8.5 (reduction of 5.2 Mt C) compared with contemporary conditions (2006-2015). Our projections use multi-model mean export flux estimates from eight fully-coupled earth system models, which contributed to the Coupled Model Intercomparison Project Phase 5, that have been forced by high representative concentration pathways (RCP8.5). These export flux estimates are used in conjunction with published empirical relationships to predict changes in benthic biomass. The polar oceans and some upwelling areas may experience increases in benthic biomass, but most other regions show decreases, with up to 38% reductions in parts of the northeast Atlantic. Our analysis projects a future ocean with smaller sized infaunal benthos, potentially reducing energy transfer rates though benthic multicellular food webs. More than 80% of potential deep-water biodiversity hotspots known around the world, including canyons, seamounts and cold-water coral reefs, are projected to experience negative changes in biomass. These major reductions in biomass may lead to widespread change in benthic ecosystems and the functions and services they provide.

Measurement units: % change
Cell size: 1 degree

Seafloor organisms are vital for healthy marine ecosystems, contributing to elemental cycling, benthic remineralisation and ultimately sequestration of carbon. Deep-sea life is primarily reliant on the export flux of particulate organic carbon from the surface ocean for food, but most ocean biogeochemistry models predict global decreases in export flux resulting from 21st century anthropogenically-induced warming. Here we show that decadal-to-century scale changes in carbon export associated with climate change lead to an estimated 5.2% decrease in future (2091-2100) global open ocean benthic biomass under RCP8.5 (reduction of 5.2 Mt C) compared with contemporary conditions (2006-2015). Our projections use multi-model mean export flux estimates from eight fully-coupled earth system models, which contributed to the Coupled Model Intercomparison Project Phase 5, that have been forced by high representative concentration pathways (RCP8.5). These export flux estimates are used in conjunction with published empirical relationships to predict changes in benthic biomass. The polar oceans and some upwelling areas may experience increases in benthic biomass, but most other regions show decreases, with up to 38% reductions in parts of the northeast Atlantic. Our analysis projects a future ocean with smaller sized infaunal benthos, potentially reducing energy transfer rates though benthic multicellular food webs. More than 80% of potential deep-water biodiversity hotspots known around the world, including canyons, seamounts and cold-water coral reefs, are projected to experience negative changes in biomass. These major reductions in biomass may lead to widespread change in benthic ecosystems and the functions and services they provide.

Measurement units: % change
Cell size: 1 degree

For a given RCP, some models can have different sets of input parameters (called input ensemble), numbered r1i1p1, r1i1p2, etc., corresponding to different settings, resulting is an output for each rXiYpZ input. Variable ‘tos’ is provided by 86 combinations of models and input ensembles (see list in Annex). To compute an ensemble mean with equal weight for each model, the different outputs of a single model are first averaged. The resulting averaged models outputs, 1 average per model, are then regridded to a common grid, defined as a regular grid, with a spatial resolution of ½ ° in latitude per ½ ° in longitude, from 0° to 360° in longitude, and -85° to 85° in latitude. Then, the regridded averages are averaged all together with the same weight.
The averaging operations are grid-cell and time independent, which means that the averaging operator is not applied along the space and time dimensions, only in-between the different models values for the same place and time.
The result of the operation is a time series of ocean surface temperature, from 2020 to 2050, at a grid resolution of 0.5°. Because of the difference in the spatial gridding, and difference in the land mass representation, some grid points did not used the same number of models averages to compute the final average: the number of model averages per grid cell is given in the final product, as well as the min-max amplitude between model averages.
Details are given in the technical chapter 4.3.
Processing software is found at: https://github.com/IOC-CODE/cmip5_projections

The CMIP5 Sea Surface Temperature models projections correspond to the Coupled atmosphere-ocean general circulation models output named ‘TOS’ (Temperature Of Surface) with a monthly time-step (12 values per year, from 2006 to 2100), for RCP 4.5 and RCP 8.5.
For a given RCP, some models can have different sets of input parameters (called input ensemble), numbered r1i1p1, r1i1p2, etc., corresponding to different input settings, resulting is an output for each rXiYpZ input. Variable ‘TOS’ is provided by 86 combinations of models and input ensembles.
To compute an ensemble mean with equal weight for each model, the different outputs of a single model are first averaged. The resulting averaged models outputs, 1 average per model, are then regridded to a common grid, defined as a regular grid, with a spatial resolution of ½ ° in latitude per ½ ° in longitude, from 0° to 360° in longitude, and -85° to 85° in latitude. Because of the difference in the spatial gridding, and difference in the land mass representation, some grid points did not used the same number of models averages to compute the final average. Then, the regridded averages are averaged all together with the same weight (Oldenborgh et al, 2013).
The averaging operations are grid-cell and time independent, which means that the averaging operator is not applied along the space and time dimensions, only in-between the different models values for the same place and time.
The result of the operation is a time series of ocean surface temperature, grouped by decades, from 2010 to 2059 (which is the last year of decade 2050): the number of model averages per grid cell is given in the final product, as well as the min-max amplitude between model averages.
Details are given in the technical chapter 4.3.
Processing software is found at: https://github.com/IOC-CODE/cmip5_projections

This model climatology is computed from CMIP5 models outputs, for the period 1971 to 2000, which pre-dates the starting dates of the RCP scenarios. The climatology is computed as an ensemble mean of the models outputs, which is then corrected from models bias by comparing to "Reynolds" climatology. Details are given in the technical chapter 4.3. Processing software found at: https://github.com/IOC-CODE/cmip5_projections
Use file "run.sh" to see examples of usage.

The projection of the Pacific and Indian Ocean warm pool is estimated by using projections of sea surface temperatures (SST).
The projected SST are obtained by averaging models outputs (see data source section below), for IPCC scenario RCP 8.5. Areas with an annual average temperature above 28.5°C are flagged as belonging to the global warmpool (file warmpool_extent_global_rcp85.txt). From this global warmpool detection, areas between 35°E and 290°E (70°W) correspond to the Pacific and Indian Ocean warmpool (file warmpool_extent_indopacific_rcp85.txt).

Source codes:
Processing of SST projection: make_warmpool.py
Example for execution in: run.sh

To create the indicators, experimental data on the capacity of pteropods to produce their shell under ocean acidification scenarios were coupled with models describing chemical conditions (aragonite saturation state) of the oceans at present, in 2030 and 2050, under an optimistic carbon dioxide emission scenario (Representative Concentration Pathway 4.5) and a pessimistic projection (Representative Concentration Pathway 8.5).
Indicators were classified in five risk levels:
1: low, 2: medium, 3: high, 4: very high, and 5: critical corresponding to a decrease in shell production of 0-20%, 20-40%, 40-60%, 60-80%, and superior to 80% respectively, compared to the present average specific shell production. For temperature, risk levels were assessed as a function of average surface yearly temperature deviation from the maximum temperature at which organisms from the three species are now found. Finally, the two indicators were combined to calculate the cumulative effect of ocean acidification and global warming.

To create the indicators, experimental data on the capacity of pteropods to produce their shell under ocean acidification scenarios were coupled with models describing chemical conditions (aragonite saturation state) of the oceans at present, in 2030 and 2050, under an optimistic carbon dioxide emission scenario (Representative Concentration Pathway 4.5) and a pessimistic projection (Representative Concentration Pathway 8.5).
Indicators were classified in five risk levels:
1: low, 2: medium, 3: high, 4: very high, and 5: critical corresponding to a decrease in shell production of 0-20%, 20-40%, 40-60%, 60-80%, and superior to 80% respectively, compared to the present average specific shell production. For temperature, risk levels were assessed as a function of average surface yearly temperature deviation from the maximum temperature at which organisms from the three species are now found. Finally, the two indicators were combined to calculate the cumulative effect of ocean acidification and global warming.

To create the indicators, experimental data on the capacity of pteropods to produce their shell under ocean acidification scenarios were coupled with models describing chemical conditions (aragonite saturation state) of the oceans at present, in 2030 and 2050, under an optimistic carbon dioxide emission scenario (Representative Concentration Pathway 4.5) and a pessimistic projection (Representative Concentration Pathway 8.5).
Indicators were classified in five risk levels:
1: low, 2: medium, 3: high, 4: very high, and 5: critical corresponding to a decrease in shell production of 0-20%, 20-40%, 40-60%, 60-80%, and superior to 80% respectively, compared to the present average specific shell production. For temperature, risk levels were assessed as a function of average surface yearly temperature deviation from the maximum temperature at which organisms from the three species are now found. Finally, the two indicators were combined to calculate the cumulative effect of ocean acidification and global warming.

To create the indicators, experimental data on the capacity of pteropods to produce their shell under ocean acidification scenarios were coupled with models describing chemical conditions (aragonite saturation state) of the oceans at present, in 2030 and 2050, under an optimistic carbon dioxide emission scenario (Representative Concentration Pathway 4.5) and a pessimistic projection (Representative Concentration Pathway 8.5).
Indicators were classified in five risk levels:
1: low, 2: medium, 3: high, 4: very high, and 5: critical corresponding to a decrease in shell production of 0-20%, 20-40%, 40-60%, 60-80%, and superior to 80% respectively, compared to the present average specific shell production. For temperature, risk levels were assessed as a function of average surface yearly temperature deviation from the maximum temperature at which organisms from the three species are now found. Finally, the two indicators were combined to calculate the cumulative effect of ocean acidification and global warming.

To create the indicators, experimental data on the capacity of pteropods to produce their shell under ocean acidification scenarios were coupled with models describing chemical conditions (aragonite saturation state) of the oceans at present, in 2030 and 2050, under an optimistic carbon dioxide emission scenario (Representative Concentration Pathway 4.5) and a pessimistic projection (Representative Concentration Pathway 8.5).
Indicators were classified in five risk levels:
1: low, 2: medium, 3: high, 4: very high, and 5: critical corresponding to a decrease in shell production of 0-20%, 20-40%, 40-60%, 60-80%, and superior to 80% respectively, compared to the present average specific shell production. For temperature, risk levels were assessed as a function of average surface yearly temperature deviation from the maximum temperature at which organisms from the three species are now found. Finally, the two indicators were combined to calculate the cumulative effect of ocean acidification and global warming.

To create the indicators, experimental data on the capacity of pteropods to produce their shell under ocean acidification scenarios were coupled with models describing chemical conditions (aragonite saturation state) of the oceans at present, in 2030 and 2050, under an optimistic carbon dioxide emission scenario (Representative Concentration Pathway 4.5) and a pessimistic projection (Representative Concentration Pathway 8.5).
Indicators were classified in five risk levels:
1: low, 2: medium, 3: high, 4: very high, and 5: critical corresponding to a decrease in shell production of 0-20%, 20-40%, 40-60%, 60-80%, and superior to 80% respectively, compared to the present average specific shell production. For temperature, risk levels were assessed as a function of average surface yearly temperature deviation from the maximum temperature at which organisms from the three species are now found. Finally, the two indicators were combined to calculate the cumulative effect of ocean acidification and global warming.

To create the indicators, experimental data on the capacity of pteropods to produce their shell under ocean acidification scenarios were coupled with models describing chemical conditions (aragonite saturation state) of the oceans at present, in 2030 and 2050, under an optimistic carbon dioxide emission scenario (Representative Concentration Pathway 4.5) and a pessimistic projection (Representative Concentration Pathway 8.5).
Indicators were classified in five risk levels:
1: low, 2: medium, 3: high, 4: very high, and 5: critical corresponding to a decrease in shell production of 0-20%, 20-40%, 40-60%, 60-80%, and superior to 80% respectively, compared to the present average specific shell production. For temperature, risk levels were assessed as a function of average surface yearly temperature deviation from the maximum temperature at which organisms from the three species are now found. Finally, the two indicators were combined to calculate the cumulative effect of ocean acidification and global warming.

To create the indicators, experimental data on the capacity of pteropods to produce their shell under ocean acidification scenarios were coupled with models describing chemical conditions (aragonite saturation state) of the oceans at present, in 2030 and 2050, under an optimistic carbon dioxide emission scenario (Representative Concentration Pathway 4.5) and a pessimistic projection (Representative Concentration Pathway 8.5).
Indicators were classified in five risk levels:
1: low, 2: medium, 3: high, 4: very high, and 5: critical corresponding to a decrease in shell production of 0-20%, 20-40%, 40-60%, 60-80%, and superior to 80% respectively, compared to the present average specific shell production. For temperature, risk levels were assessed as a function of average surface yearly temperature deviation from the maximum temperature at which organisms from the three species are now found. Finally, the two indicators were combined to calculate the cumulative effect of ocean acidification and global warming.

To create the indicators, experimental data on the capacity of pteropods to produce their shell under ocean acidification scenarios were coupled with models describing chemical conditions (aragonite saturation state) of the oceans at present, in 2030 and 2050, under an optimistic carbon dioxide emission scenario (Representative Concentration Pathway 4.5) and a pessimistic projection (Representative Concentration Pathway 8.5).
Indicators were classified in five risk levels:
1: low, 2: medium, 3: high, 4: very high, and 5: critical corresponding to a decrease in shell production of 0-20%, 20-40%, 40-60%, 60-80%, and superior to 80% respectively, compared to the present average specific shell production. For temperature, risk levels were assessed as a function of average surface yearly temperature deviation from the maximum temperature at which organisms from the three species are now found. Finally, the two indicators were combined to calculate the cumulative effect of ocean acidification and global warming.

To create the indicators, experimental data on the capacity of pteropods to produce their shell under ocean acidification scenarios were coupled with models describing chemical conditions (aragonite saturation state) of the oceans at present, in 2030 and 2050, under an optimistic carbon dioxide emission scenario (Representative Concentration Pathway 4.5) and a pessimistic projection (Representative Concentration Pathway 8.5).
Indicators were classified in five risk levels:
1: low, 2: medium, 3: high, 4: very high, and 5: critical corresponding to a decrease in shell production of 0-20%, 20-40%, 40-60%, 60-80%, and superior to 80% respectively, compared to the present average specific shell production. For temperature, risk levels were assessed as a function of average surface yearly temperature deviation from the maximum temperature at which organisms from the three species are now found. Finally, the two indicators were combined to calculate the cumulative effect of ocean acidification and global warming.

To create the indicators, experimental data on the capacity of pteropods to produce their shell under ocean acidification scenarios were coupled with models describing chemical conditions (aragonite saturation state) of the oceans at present, in 2030 and 2050, under an optimistic carbon dioxide emission scenario (Representative Concentration Pathway 4.5) and a pessimistic projection (Representative Concentration Pathway 8.5).
Indicators were classified in five risk levels:
1: low, 2: medium, 3: high, 4: very high, and 5: critical corresponding to a decrease in shell production of 0-20%, 20-40%, 40-60%, 60-80%, and superior to 80% respectively, compared to the present average specific shell production. For temperature, risk levels were assessed as a function of average surface yearly temperature deviation from the maximum temperature at which organisms from the three species are now found. Finally, the two indicators were combined to calculate the cumulative effect of ocean acidification and global warming.

To create the indicators, experimental data on the capacity of pteropods to produce their shell under ocean acidification scenarios were coupled with models describing chemical conditions (aragonite saturation state) of the oceans at present, in 2030 and 2050, under an optimistic carbon dioxide emission scenario (Representative Concentration Pathway 4.5) and a pessimistic projection (Representative Concentration Pathway 8.5).
Indicators were classified in five risk levels:
1: low, 2: medium, 3: high, 4: very high, and 5: critical corresponding to a decrease in shell production of 0-20%, 20-40%, 40-60%, 60-80%, and superior to 80% respectively, compared to the present average specific shell production. For temperature, risk levels were assessed as a function of average surface yearly temperature deviation from the maximum temperature at which organisms from the three species are now found. Finally, the two indicators were combined to calculate the cumulative effect of ocean acidification and global warming.

To create the indicators, experimental data on the capacity of pteropods to produce their shell under ocean acidification scenarios were coupled with models describing chemical conditions (aragonite saturation state) of the oceans at present, in 2030 and 2050, under an optimistic carbon dioxide emission scenario (Representative Concentration Pathway 4.5) and a pessimistic projection (Representative Concentration Pathway 8.5).
Indicators were classified in five risk levels:
1: low, 2: medium, 3: high, 4: very high, and 5: critical corresponding to a decrease in shell production of 0-20%, 20-40%, 40-60%, 60-80%, and superior to 80% respectively, compared to the present average specific shell production. For temperature, risk levels were assessed as a function of average surface yearly temperature deviation from the maximum temperature at which organisms from the three species are now found. Finally, the two indicators were combined to calculate the cumulative effect of ocean acidification and global warming.

To create the indicators, experimental data on the capacity of pteropods to produce their shell under ocean acidification scenarios were coupled with models describing chemical conditions (aragonite saturation state) of the oceans at present, in 2030 and 2050, under an optimistic carbon dioxide emission scenario (Representative Concentration Pathway 4.5) and a pessimistic projection (Representative Concentration Pathway 8.5).
Indicators were classified in five risk levels:
1: low, 2: medium, 3: high, 4: very high, and 5: critical corresponding to a decrease in shell production of 0-20%, 20-40%, 40-60%, 60-80%, and superior to 80% respectively, compared to the present average specific shell production. For temperature, risk levels were assessed as a function of average surface yearly temperature deviation from the maximum temperature at which organisms from the three species are now found. Finally, the two indicators were combined to calculate the cumulative effect of ocean acidification and global warming.

To create the indicators, experimental data on the capacity of pteropods to produce their shell under ocean acidification scenarios were coupled with models describing chemical conditions (aragonite saturation state) of the oceans at present, in 2030 and 2050, under an optimistic carbon dioxide emission scenario (Representative Concentration Pathway 4.5) and a pessimistic projection (Representative Concentration Pathway 8.5).
Indicators were classified in five risk levels:
1: low, 2: medium, 3: high, 4: very high, and 5: critical corresponding to a decrease in shell production of 0-20%, 20-40%, 40-60%, 60-80%, and superior to 80% respectively, compared to the present average specific shell production. For temperature, risk levels were assessed as a function of average surface yearly temperature deviation from the maximum temperature at which organisms from the three species are now found. Finally, the two indicators were combined to calculate the cumulative effect of ocean acidification and global warming.

To create the indicators, experimental data on the capacity of pteropods to produce their shell under ocean acidification scenarios were coupled with models describing chemical conditions (aragonite saturation state) of the oceans at present, in 2030 and 2050, under an optimistic carbon dioxide emission scenario (Representative Concentration Pathway 4.5) and a pessimistic projection (Representative Concentration Pathway 8.5).
Indicators were classified in five risk levels:
1: low, 2: medium, 3: high, 4: very high, and 5: critical corresponding to a decrease in shell production of 0-20%, 20-40%, 40-60%, 60-80%, and superior to 80% respectively, compared to the present average specific shell production. For temperature, risk levels were assessed as a function of average surface yearly temperature deviation from the maximum temperature at which organisms from the three species are now found. Finally, the two indicators were combined to calculate the cumulative effect of ocean acidification and global warming.

To create the indicators, experimental data on the capacity of pteropods to produce their shell under ocean acidification scenarios were coupled with models describing chemical conditions (aragonite saturation state) of the oceans at present, in 2030 and 2050, under an optimistic carbon dioxide emission scenario (Representative Concentration Pathway 4.5) and a pessimistic projection (Representative Concentration Pathway 8.5).
Indicators were classified in five risk levels:
1: low, 2: medium, 3: high, 4: very high, and 5: critical corresponding to a decrease in shell production of 0-20%, 20-40%, 40-60%, 60-80%, and superior to 80% respectively, compared to the present average specific shell production. For temperature, risk levels were assessed as a function of average surface yearly temperature deviation from the maximum temperature at which organisms from the three species are now found. Finally, the two indicators were combined to calculate the cumulative effect of ocean acidification and global warming.

To create the indicators, experimental data on the capacity of pteropods to produce their shell under ocean acidification scenarios were coupled with models describing chemical conditions (aragonite saturation state) of the oceans at present, in 2030 and 2050, under an optimistic carbon dioxide emission scenario (Representative Concentration Pathway 4.5) and a pessimistic projection (Representative Concentration Pathway 8.5).
Indicators were classified in five risk levels:
1: low, 2: medium, 3: high, 4: very high, and 5: critical corresponding to a decrease in shell production of 0-20%, 20-40%, 40-60%, 60-80%, and superior to 80% respectively, compared to the present average specific shell production. For temperature, risk levels were assessed as a function of average surface yearly temperature deviation from the maximum temperature at which organisms from the three species are now found. Finally, the two indicators were combined to calculate the cumulative effect of ocean acidification and global warming.

For low emission scenarios (SSP1 and SSP4 with RCP4.5 sea level rise): 2100 Country Sea Level Rise Risk Index = (LN Total Area + LN Population)_normalized × (RCP 4.5 Maximum Sea Level Rise) × (HDI Gap)
For high emission scenarios (SSP2, SSP3 and SSP5 with RCP8.5 sea level rise): 2100 Country Sea Level Rise Risk Index = (LN Total Area + LN Population)_normalized × (RCP 8.5 Maximum Sea Level Rise) × (HDI Gap)
Country scale risk indices are weighted by the ratios of the total area within the 10m elevation and 10km coast to the regional total, and summed to obtain the regional risk index for each of 18 regions of the world.

DHM projections are computed by comparing the ensemble mean of SST projections from CMIP5 models outputs (for IPCC scenarios RCP 4.5 and RCP 8.5) to Reynolds climatology.
The DHM is defined as the accumulation, over a time period of 4 months, of temperature differences between models projections and Reynolds climatology (only positive difference, i.e. projection temperature above climatology, are accounted).
A yearly DHM corresponds to the maximum 4-month-DHM observed in the year. A level 2 DHM corresponds the 4-month-DHM equal to or above 2°C. The level 2 frequency corresponds to the number of occurrence of yearly level 2 in 10 years.

Ensemble mean computation: https://github.com/IOC-CODE/cmip5_projections/blob/master/make_ensembleM...
DHM and DHM-level 2 frequency computation: https://github.com/IOC-CODE/cmip5_projections/blob/master/make_dhm.py

DHM projections are computed by comparing the ensemble mean of SST projections from CMIP5 models outputs (for IPCC scenarios RCP 4.5 and RCP 8.5) to Reynolds climatology.
The DHM is defined as the accumulation, over a time period of 4 months, of temperature differences between models projections and Reynolds climatology (only positive difference, i.e. projection temperature above climatology, are accounted).
A yearly DHM corresponds to the maximum 4-month-DHM observed in the year. A level 2 DHM corresponds the 4-month-DHM equal to or above 2°C. The level 2 frequency corresponds to the number of occurrence of yearly level 2 in 10 years.

Ensemble mean computation: https://github.com/BrunoCombal/climate/blob/master/make_ensembleMean_tyx.py
DHM and DHM-level 2 frequency computation: https://github.com/BrunoCombal/climate/blob/master/make_dhm.py

DHM projections are computed by comparing the ensemble mean of SST projections from CMIP5 models outputs (for IPCC scenarios RCP 4.5 and RCP 8.5) to Reynolds climatology.
The DHM is defined as the accumulation, over a time period of 4 months, of temperature differences between models projections and Reynolds climatology (only positive difference, i.e. projection temperature above climatology, are accounted).
A yearly DHM corresponds to the maximum 4-month-DHM observed in the year. A level 2 DHM corresponds the 4-month-DHM equal to or above 2°C. The level 2 frequency corresponds to the number of occurrence of yearly level 2 in 10 years.

Ensemble mean computation: https://github.com/BrunoCombal/climate/blob/master/make_ensembleMean_tyx.py
DHM and DHM-level 2 frequency computation: https://github.com/BrunoCombal/climate/blob/master/make_dhm.py

DHM projections are computed by comparing the ensemble mean of SST projections from CMIP5 models outputs (for IPCC scenarios RCP 4.5 and RCP 8.5) to Reynolds climatology.
The DHM is defined as the accumulation, over a time period of 4 months, of temperature differences between models projections and Reynolds climatology (only positive difference, i.e. projection temperature above climatology, are accounted).
A yearly DHM corresponds to the maximum 4-month-DHM observed in the year. A level 2 DHM corresponds the 4-month-DHM equal to or above 2°C. The level 2 frequency corresponds to the number of occurrence of yearly level 2 in 10 years.

Ensemble mean computation: https://github.com/BrunoCombal/climate/blob/master/make_ensembleMean_tyx.py
DHM and DHM-level 2 frequency computation: https://github.com/BrunoCombal/climate/blob/master/make_dhm.py

DHM projections are computed by comparing the ensemble mean of SST projections from CMIP5 models outputs (for IPCC scenarios RCP 4.5 and RCP 8.5) to Reynolds climatology.
The DHM is defined as the accumulation, over a time period of 4 months, of temperature differences between models projections and Reynolds climatology (only positive difference, i.e. projection temperature above climatology, are accounted).
A yearly DHM corresponds to the maximum 4-month-DHM observed in the year. A level 2 DHM corresponds the 4-month-DHM equal to or above 2°C. The level 2 frequency corresponds to the number of occurrence of yearly level 2 in 10 years.

Ensemble mean computation: https://github.com/BrunoCombal/climate/blob/master/make_ensembleMean_tyx.py
DHM and DHM-level 2 frequency computation: https://github.com/BrunoCombal/climate/blob/master/make_dhm.py

DHM projections are computed by comparing the ensemble mean of SST projections from CMIP5 models outputs (for IPCC scenarios RCP 4.5 and RCP 8.5) to Reynolds climatology.
The DHM is defined as the accumulation, over a time period of 4 months, of temperature differences between models projections and Reynolds climatology (only positive difference, i.e. projection temperature above climatology, are accounted).
A yearly DHM corresponds to the maximum 4-month-DHM observed in the year. A level 2 DHM corresponds the 4-month-DHM equal to or above 2°C. The level 2 frequency corresponds to the number of occurrence of yearly level 2 in 10 years.

Ensemble mean computation: https://github.com/BrunoCombal/climate/blob/master/make_ensembleMean_tyx.py
DHM and DHM-level 2 frequency computation: https://github.com/BrunoCombal/climate/blob/master/make_dhm.py

DHM projections are computed by comparing the ensemble mean of SST projections from CMIP5 models outputs (for IPCC scenarios RCP 4.5 and RCP 8.5) to Reynolds climatology.
The DHM is defined as the accumulation, over a time period of 4 months, of temperature differences between models projections and Reynolds climatology (only positive difference, i.e. projection temperature above climatology, are accounted).
A yearly DHM corresponds to the maximum 4-month-DHM observed in the year. A level 2 DHM corresponds the 4-month-DHM equal to or above 2°C. The level 2 frequency corresponds to the number of occurrence of yearly level 2 in 10 years.

Ensemble mean computation: https://github.com/BrunoCombal/climate/blob/master/make_ensembleMean_tyx.py
DHM and DHM-level 2 frequency computation: https://github.com/BrunoCombal/climate/blob/master/make_dhm.py

DHM projections are computed by comparing the ensemble mean of SST projections from CMIP5 models outputs (for IPCC scenarios RCP 4.5 and RCP 8.5) to Reynolds climatology.
The DHM is defined as the accumulation, over a time period of 4 months, of temperature differences between models projections and Reynolds climatology (only positive difference, i.e. projection temperature above climatology, are accounted).
A yearly DHM corresponds to the maximum 4-month-DHM observed in the year. A level 2 DHM corresponds the 4-month-DHM equal to or above 2°C. The level 2 frequency corresponds to the number of occurrence of yearly level 2 in 10 years.

Ensemble mean computation: https://github.com/BrunoCombal/climate/blob/master/make_ensembleMean_tyx.py
DHM and DHM-level 2 frequency computation: https://github.com/BrunoCombal/climate/blob/master/make_dhm.py

DHM projections are computed by comparing the ensemble mean of SST projections from CMIP5 models outputs (for IPCC scenarios RCP 4.5 and RCP 8.5) to Reynolds climatology.
The DHM is defined as the accumulation, over a time period of 4 months, of temperature differences between models projections and Reynolds climatology (only positive difference, i.e. projection temperature above climatology, are accounted).
A yearly DHM corresponds to the maximum 4-month-DHM observed in the year. A level 2 DHM corresponds the 4-month-DHM equal to or above 2°C. The level 2 frequency corresponds to the number of occurrence of yearly level 2 in 10 years.

Ensemble mean computation: https://github.com/BrunoCombal/climate/blob/master/make_ensembleMean_tyx.py
DHM and DHM-level 2 frequency computation: https://github.com/BrunoCombal/climate/blob/master/make_dhm.py

DHM projections are computed by comparing the ensemble mean of SST projections from CMIP5 models outputs (for IPCC scenarios RCP 4.5 and RCP 8.5) to Reynolds climatology.
The DHM is defined as the accumulation, over a time period of 4 months, of temperature differences between models projections and Reynolds climatology (only positive difference, i.e. projection temperature above climatology, are accounted).
A yearly DHM corresponds to the maximum 4-month-DHM observed in the year. A level 2 DHM corresponds the 4-month-DHM equal to or above 2°C. The level 2 frequency corresponds to the number of occurrence of yearly level 2 in 10 years.

Ensemble mean computation: https://github.com/BrunoCombal/climate/blob/master/make_ensembleMean_tyx.py
DHM and DHM-level 2 frequency computation: https://github.com/BrunoCombal/climate/blob/master/make_dhm.py

Data are published in DarwinCore format via the OBIS nodes and harvested by the central OBIS node hosted at the UNESCO/IOC project office for IODE in Oostende (Belgium). Data are processed, quality controlled, integrated and indexed into a single consoldidated OBIS database.

Data are published in DarwinCore format via the OBIS nodes and harvested by the central OBIS node hosted at the UNESCO/IOC project office for IODE in Oostende (Belgium). Data are processed, quality controlled, integrated and indexed into a single consoldidated OBIS databae

Data are published in DarwinCore format via the OBIS nodes and harvested by the central OBIS node hosted at the UNESCO/IOC project office for IODE in Oostende (Belgium). Data are processed, quality controlled, integrated and indexed into a single consoldidated OBIS databae

Data are published in DarwinCore format via the OBIS nodes and harvested by the central OBIS node hosted at the UNESCO/IOC project office for IODE in Oostende (Belgium). Data are processed, quality controlled, integrated and indexed into a single consoldidated OBIS database.

Data are published in DarwinCore format via the OBIS nodes and harvested by the central OBIS node hosted at the UNESCO/IOC project office for IODE in Oostende (Belgium). Data are processed, quality controlled, integrated and indexed into a single consoldidated OBIS database.

Data are published in DarwinCore format via the OBIS nodes and harvested by the central OBIS node hosted at the UNESCO/IOC project office for IODE in Oostende (Belgium). Data are processed, quality controlled, integrated and indexed into a single consolidated OBIS database.

Data are published in DarwinCore format via the OBIS nodes and harvested by the central OBIS node hosted at the UNESCO/IOC project office for IODE in Oostende (Belgium). Data are processed, quality controlled, integrated and indexed into a single consoldidated OBIS database.

Data are published in DarwinCore format via the OBIS nodes and harvested by the central OBIS node hosted at the UNESCO/IOC project office for IODE in Oostende (Belgium). Data are processed, quality controlled, integrated and indexed into a single consoldidated OBIS database.

Data are published in DarwinCore format via the OBIS nodes and harvested by the central OBIS node hosted at the UNESCO/IOC project office for IODE in Oostende (Belgium). Data are processed, quality controlled, integrated and indexed into a single consoldidated OBIS database.

Data are published in DarwinCore format via the OBIS nodes and harvested by the central OBIS node hosted at the UNESCO/IOC project office for IODE in Oostende (Belgium). Data are processed, quality controlled, integrated and indexed into a single consoldidated OBIS database.

Data are published in DarwinCore format via the OBIS nodes and harvested by the central OBIS node hosted at the UNESCO/IOC project office for IODE in Oostende (Belgium). Data are processed, quality controlled, integrated and indexed into a single consoldidated OBIS database.

Data are published in DarwinCore format via the OBIS nodes and harvested by the central OBIS node hosted at the UNESCO/IOC project office for IODE in Oostende (Belgium). Data are processed, quality controlled, integrated and indexed into a single consoldidated OBIS database.

The empirical NASA OC4v6 algorithm was used on water-leaving reflectances at four wavelengths.

The phenology indicators were estimated through the fitting of an equation, constructed from two Gaussian peaks and a sloping baseline chlorophyll concentration, to the chlorophyll time series for each year, resolved at five-day intervals.

The threat index combines the percentage in different threat categories into a single index: Local threat index = % under Medium threat + 2 x % High threat + 3x % Very high threat. The Integrated Local Threat Index combines four local threats: overfishing (including destructive fishing), coastal development, watershed-based sediment and pollution, and marine based threats (physical damage and pollution). The integrated local threat index was developed by summing the four individual local threats, where reefs were categorized into low (0), medium (1), or high (2) in each case. The summed threats were then categorized into the index as follows: Low: 0 points (scored low for all local threats), Medium: 1–2 points (scored medium on one or two local threats or high on a single threat), High: 3–4 points (scored medium on at least three threats, or medium on one threat and high on another threat, or high on two threats), Very high: 5 points or higher (scored medium or higher on at least three threats, and scored high on at least one).

The Integrated Local Threat Index combines four local threats: overfishing (including destructive fishing), coastal development, watershed-based sediment and pollution, and marine based threats (physical damage and pollution). The integrated local threat index was developed by summing the four individual local threats, where reefs were categorized into low (0), medium (1), or high (2) in each case. The summed threats were then categorized into the index as follows: Low: 0 points (scored low for all local threats), Medium: 1–2 points (scored medium on one or two local threats or high on a single threat), High: 3–4 points (scored medium on at least three threats, or medium on one threat and high on another threat, or high on two threats), Very high: 5 points or higher (scored medium or higher on at least three threats, and scored high on at least one).

tbd

tbd

tbd

tbd

The sampling region is the Benguela Current. CPR surveys off the west coast of South Africa, Namibia and Angola are the most recent routine survey to get underway, with samples first collected in 2011. Data have not been included in this assessment, though it is expected that future assessments will contain CPR data, but there is a lengthy time series of zooplankton net sampling from the region dating back to 1951. Sampling took place in St Helena Bay, South Africa (latitude 32-33°S) and more recently also off Walvis bay in Namibia, 23°S, 4°E, during austral autumn, at the peak of pelagic fish recruitment. See www.sahfos.org for further details. The data is derived from zooplankton sampled by Continuous Plankton Recorders (CPRs). CPRs are typically deployed from commercial ships on their normal routes of passage. The CPR is a robust mechanical device towed behind ships in the near-surface waters (usually commercial ships but also research, military and fishing vessels) that filters plankton from the water along the ship’s path onto a length of 270 µm mesh. The mesh is divided into separate 18.5 km samples after the sampling (9.25 km in the Southern ocean Survey), with time, date and position of each sample calculated from ship’s log information or recorded underway GPS/environmental data. The plankton are identified with a microscope and counted. Full details of the methodology can be found in Batten et al. (2003) and Hosie et al. (2003), and details on data analysis and utility in Richardson et al. (2006).
Batten, S.D., Clarke, R.A., Flinkman, J., Hays, G.C., John, E.H., John, A.W.G., Jonas, T.J., Lindley, J.A., Stevens, D.P., Walne, A.W. (2003) CPR sampling – The technical background, materials and methods, consistency and comparability. Progress in Oceanography, 58, 193-215.
Hosie, G.W., Fukuchi, M., Kawaguchi, S. (2003) Development of the Southern Ocean Continuous Plankton Recorder Survey. Progress in Oceanography 58 (2-4), 263–283
Richardson, A.J., Walne, A.W., John, A.W.G.J., Jonas, T.D., Lindley, J.A, Sims, D.W., Stevens, D., and Witt, M. (2006) Using continuous plankton recorder data. Progress in Oceanography, 68, 27-74.

The sampling region is the Northeast Atlantic, which has the greatest temporal (over 60 years) and spatial coverage. Because of the large amount of data it has been subdivided into 41 “Standard Areas”, usually along major lines of latitude and longitude (see www.sahfos.org for further details).
The data is derived from zooplankton sampled by Continuous Plankton Recorders (CPRs). CPRs are typically deployed from commercial ships on their normal routes of passage. The CPR is a robust mechanical device towed behind ships in the near-surface waters (usually commercial ships but also research, military and fishing vessels) that filters plankton from the water along the ship’s path onto a length of 270 µm mesh. The mesh is divided into separate 18.5 km samples after the sampling (9.25 km in the Southern ocean Survey), with time, date and position of each sample calculated from ship’s log information or recorded underway GPS/environmental data. The plankton are identified with a microscope and counted. Full details of the methodology can be found in Batten et al. (2003) and Hosie et al. (2003), and details on data analysis and utility in Richardson et al. (2006).
Batten, S.D., Clarke, R.A., Flinkman, J., Hays, G.C., John, E.H., John, A.W.G., Jonas, T.J., Lindley, J.A., Stevens, D.P., Walne, A.W. (2003) CPR sampling – The technical background, materials and methods, consistency and comparability. Progress in Oceanography, 58, 193-215.
Hosie, G.W., Fukuchi, M., Kawaguchi, S. (2003) Development of the Southern Ocean Continuous Plankton Recorder Survey. Progress in Oceanography 58 (2-4), 263–283
Richardson, A.J., Walne, A.W., John, A.W.G.J., Jonas, T.D., Lindley, J.A, Sims, D.W., Stevens, D., and Witt, M. (2006) Using continuous plankton recorder data. Progress in Oceanography, 68, 27-74.

Australia is unique globally in being bounded by two poleward-flowing warm-water currents and the CPR survey samples the east, west and south coasts of Australia. Longhurst provinces were used to divide up the Australian marine domain into ecoregions viz. AUSE (Eastern Australia Coastal), ARCH (Western Pacific Archipelagic Deep Basins), TASM (Tasman Sea), AUSW (Australia-Indonesia Coastal), MONS (Indian Ocean Monsoon Gyres), SSTC (South Subtropical Convergence). See www.sahfos.org for further details. The data is derived from zooplankton sampled by Continuous Plankton Recorders (CPRs). CPRs are typically deployed from commercial ships on their normal routes of passage. The CPR is a robust mechanical device towed behind ships in the near-surface waters (usually commercial ships but also research, military and fishing vessels) that filters plankton from the water along the ship’s path onto a length of 270 µm mesh. The mesh is divided into separate 18.5 km samples after the sampling (9.25 km in the Southern ocean Survey), with time, date and position of each sample calculated from ship’s log information or recorded underway GPS/environmental data. The plankton are identified with a microscope and counted. Full details of the methodology can be found in Batten et al. (2003) and Hosie et al. (2003), and details on data analysis and utility in Richardson et al. (2006).
Batten, S.D., Clarke, R.A., Flinkman, J., Hays, G.C., John, E.H., John, A.W.G., Jonas, T.J., Lindley, J.A., Stevens, D.P., Walne, A.W. (2003) CPR sampling – The technical background, materials and methods, consistency and comparability. Progress in Oceanography, 58, 193-215.
Hosie, G.W., Fukuchi, M., Kawaguchi, S. (2003) Development of the Southern Ocean Continuous Plankton Recorder Survey. Progress in Oceanography 58 (2-4), 263–283
Richardson, A.J., Walne, A.W., John, A.W.G.J., Jonas, T.D., Lindley, J.A, Sims, D.W., Stevens, D., and Witt, M. (2006) Using continuous plankton recorder data. Progress in Oceanography, 68, 27-74.

The sampling region is the Northeast Pacific, which is subdivided into the coastal regions of the British Columbian shelf, the shelf south of Prince William Sound, Cook Inlet, the shelf SE of Cook Inlet, and the shelf around Unimak Pass in the Aleutian Islands (see www.sahfos.org for further details).
The data is derived from zooplankton sampled by Continuous Plankton Recorders (CPRs). CPRs are typically deployed from commercial ships on their normal routes of passage. The CPR is a robust mechanical device towed behind ships in the near-surface waters (usually commercial ships but also research, military and fishing vessels) that filters plankton from the water along the ship’s path onto a length of 270 µm mesh. The mesh is divided into separate 18.5 km samples after the sampling (9.25 km in the Southern ocean Survey), with time, date and position of each sample calculated from ship’s log information or recorded underway GPS/environmental data. The plankton are identified with a microscope and counted. Full details of the methodology can be found in Batten et al. (2003) and Hosie et al. (2003), and details on data analysis and utility in Richardson et al. (2006).
Batten, S.D., Clarke, R.A., Flinkman, J., Hays, G.C., John, E.H., John, A.W.G., Jonas, T.J., Lindley, J.A., Stevens, D.P., Walne, A.W. (2003) CPR sampling – The technical background, materials and methods, consistency and comparability. Progress in Oceanography, 58, 193-215.
Hosie, G.W., Fukuchi, M., Kawaguchi, S. (2003) Development of the Southern Ocean Continuous Plankton Recorder Survey. Progress in Oceanography 58 (2-4), 263–283
Richardson, A.J., Walne, A.W., John, A.W.G.J., Jonas, T.D., Lindley, J.A, Sims, D.W., Stevens, D., and Witt, M. (2006) Using continuous plankton recorder data. Progress in Oceanography, 68, 27-74.

The sampling region is the Northwest Atlantic, which which is subdivided into the Gulf of Maine and mid-Atlantic Bight regions. (see www.sahfos.org for further details). The data is derived from zooplankton sampled by Continuous Plankton Recorders (CPRs). CPRs are typically deployed from commercial ships on their normal routes of passage. The CPR is a robust mechanical device towed behind ships in the near-surface waters (usually commercial ships but also research, military and fishing vessels) that filters plankton from the water along the ship’s path onto a length of 270 µm mesh. The mesh is divided into separate 18.5 km samples after the sampling (9.25 km in the Southern ocean Survey), with time, date and position of each sample calculated from ship’s log information or recorded underway GPS/environmental data. The plankton are identified with a microscope and counted. Full details of the methodology can be found in Batten et al. (2003) and Hosie et al. (2003), and details on data analysis and utility in Richardson et al. (2006).
Batten, S.D., Clarke, R.A., Flinkman, J., Hays, G.C., John, E.H., John, A.W.G., Jonas, T.J., Lindley, J.A., Stevens, D.P., Walne, A.W. (2003) CPR sampling – The technical background, materials and methods, consistency and comparability. Progress in Oceanography, 58, 193-215.
Hosie, G.W., Fukuchi, M., Kawaguchi, S. (2003) Development of the Southern Ocean Continuous Plankton Recorder Survey. Progress in Oceanography 58 (2-4), 263–283
Richardson, A.J., Walne, A.W., John, A.W.G.J., Jonas, T.D., Lindley, J.A, Sims, D.W., Stevens, D., and Witt, M. (2006) Using continuous plankton recorder data. Progress in Oceanography, 68, 27-74.

The sampling region is the Northwest Pacific, which is subdivided into the Northwest Pacific Oceanic Gyre and the Oyashio current region (typically richer in nutrients). See www.sahfos.org for further details.
The data is derived from zooplankton sampled by Continuous Plankton Recorders (CPRs). CPRs are typically deployed from commercial ships on their normal routes of passage. The CPR is a robust mechanical device towed behind ships in the near-surface waters (usually commercial ships but also research, military and fishing vessels) that filters plankton from the water along the ship’s path onto a length of 270 µm mesh. The mesh is divided into separate 18.5 km samples after the sampling (9.25 km in the Southern ocean Survey), with time, date and position of each sample calculated from ship’s log information or recorded underway GPS/environmental data. The plankton are identified with a microscope and counted. Full details of the methodology can be found in Batten et al. (2003) and Hosie et al. (2003), and details on data analysis and utility in Richardson et al. (2006).
Batten, S.D., Clarke, R.A., Flinkman, J., Hays, G.C., John, E.H., John, A.W.G., Jonas, T.J., Lindley, J.A., Stevens, D.P., Walne, A.W. (2003) CPR sampling – The technical background, materials and methods, consistency and comparability. Progress in Oceanography, 58, 193-215.
Hosie, G.W., Fukuchi, M., Kawaguchi, S. (2003) Development of the Southern Ocean Continuous Plankton Recorder Survey. Progress in Oceanography 58 (2-4), 263–283
Richardson, A.J., Walne, A.W., John, A.W.G.J., Jonas, T.D., Lindley, J.A, Sims, D.W., Stevens, D., and Witt, M. (2006) Using continuous plankton recorder data. Progress in Oceanography, 68, 27-74.

With no land masses except Antarctica as the southern boundary, the Southern Ocean is characterised by latitudinal fronts. Four frontal zones are recognised:
- the Sea Ice Zone (SIZ) with the northern limit being defined as the maximum northern winter extent based on the 15% ice cover threshold,
- the Permanent Open Ocean Zone (POOZ) between the SIZ and the Polar Front, which lies within the ACC but displays a marked change in temperature and salinity,
- the Polar Frontal Zone (PFZ), between the Polar Front and the Sub-Antarctic Front, which is the northern boundary of the Antarctic Circum-polar Current (ACC),
- and the Sub-Antarctic Zone (SAZ), north of the Sub-Antarctic Front, which is a noted biogeographic boundary for zooplankton.
Data within each frontal zone have also been subdivided into the eastern Antarctic Sector and the Ross Sea sector.
See www.sahfos.org for further details. The data is derived from zooplankton sampled by Continuous Plankton Recorders (CPRs). CPRs are typically deployed from commercial ships on their normal routes of passage. The CPR is a robust mechanical device towed behind ships in the near-surface waters (usually commercial ships but also research, military and fishing vessels) that filters plankton from the water along the ship’s path onto a length of 270 µm mesh. The mesh is divided into separate 18.5 km samples after the sampling (9.25 km in the Southern ocean Survey), with time, date and position of each sample calculated from ship’s log information or recorded underway GPS/environmental data. The plankton are identified with a microscope and counted. Full details of the methodology can be found in Batten et al. (2003) and Hosie et al. (2003), and details on data analysis and utility in Richardson et al. (2006).
Batten, S.D., Clarke, R.A., Flinkman, J., Hays, G.C., John, E.H., John, A.W.G., Jonas, T.J., Lindley, J.A., Stevens, D.P., Walne, A.W. (2003) CPR sampling – The technical background, materials and methods, consistency and comparability. Progress in Oceanography, 58, 193-215.
Hosie, G.W., Fukuchi, M., Kawaguchi, S. (2003) Development of the Southern Ocean Continuous Plankton Recorder Survey. Progress in Oceanography 58 (2-4), 263–283
Richardson, A.J., Walne, A.W., John, A.W.G.J., Jonas, T.D., Lindley, J.A, Sims, D.W., Stevens, D., and Witt, M. (2006) Using continuous plankton recorder data. Progress in Oceanography, 68, 27-74.

Seafloor organisms are vital for healthy marine ecosystems, contributing to elemental cycling, benthic remineralisation and ultimately sequestration of carbon. Deep-sea life is primarily reliant on the export flux of particulate organic carbon from the surface ocean for food, but most ocean biogeochemistry models predict global decreases in export flux resulting from 21st century anthropogenically-induced warming. Here we show that decadal-to-century scale changes in carbon export associated with climate change lead to an estimated 5.2% decrease in future (2091-2100) global open ocean benthic biomass under RCP8.5 (reduction of 5.2 Mt C) compared with contemporary conditions (2006-2015). Our projections use multi-model mean export flux estimates from eight fully-coupled earth system models, which contributed to the Coupled Model Intercomparison Project Phase 5, that have been forced by high representative concentration pathways (RCP8.5). These export flux estimates are used in conjunction with published empirical relationships to predict changes in benthic biomass. The polar oceans and some upwelling areas may experience increases in benthic biomass, but most other regions show decreases, with up to 38% reductions in parts of the northeast Atlantic. Our analysis projects a future ocean with smaller sized infaunal benthos, potentially reducing energy transfer rates though benthic multicellular food webs. More than 80% of potential deep-water biodiversity hotspots known around the world, including canyons, seamounts and cold-water coral reefs, are projected to experience negative changes in biomass. These major reductions in biomass may lead to widespread change in benthic ecosystems and the functions and services they provide.

Measurement units: % change
Cell size: 1 degree

Seafloor organisms are vital for healthy marine ecosystems, contributing to elemental cycling, benthic remineralisation and ultimately sequestration of carbon. Deep-sea life is primarily reliant on the export flux of particulate organic carbon from the surface ocean for food, but most ocean biogeochemistry models predict global decreases in export flux resulting from 21st century anthropogenically-induced warming. Here we show that decadal-to-century scale changes in carbon export associated with climate change lead to an estimated 5.2% decrease in future (2091-2100) global open ocean benthic biomass under RCP8.5 (reduction of 5.2 Mt C) compared with contemporary conditions (2006-2015). Our projections use multi-model mean export flux estimates from eight fully-coupled earth system models, which contributed to the Coupled Model Intercomparison Project Phase 5, that have been forced by high representative concentration pathways (RCP8.5). These export flux estimates are used in conjunction with published empirical relationships to predict changes in benthic biomass. The polar oceans and some upwelling areas may experience increases in benthic biomass, but most other regions show decreases, with up to 38% reductions in parts of the northeast Atlantic. Our analysis projects a future ocean with smaller sized infaunal benthos, potentially reducing energy transfer rates though benthic multicellular food webs. More than 80% of potential deep-water biodiversity hotspots known around the world, including canyons, seamounts and cold-water coral reefs, are projected to experience negative changes in biomass. These major reductions in biomass may lead to widespread change in benthic ecosystems and the functions and services they provide.

Measurement units: % change
Cell size: 1 degree

Primary production was computed using a spectrally resolved vertical model of light transmission and primary production with variables for the chlorophyll-a profile, phytoplankton photo-physiology, and surface irradiance assigned on a pixel-by-pixel basis.

To create the indicators, experimental data on the capacity of pteropods to produce their shell under ocean acidification scenarios were coupled with models describing chemical conditions (aragonite saturation state) of the oceans at present, in 2030 and 2050, under an optimistic carbon dioxide emission scenario (Representative Concentration Pathway 4.5) and a pessimistic projection (Representative Concentration Pathway 8.5).
Indicators were classified in five risk levels:
1: low, 2: medium, 3: high, 4: very high, and 5: critical corresponding to a decrease in shell production of 0-20%, 20-40%, 40-60%, 60-80%, and superior to 80% respectively, compared to the present average specific shell production. For temperature, risk levels were assessed as a function of average surface yearly temperature deviation from the maximum temperature at which organisms from the three species are now found. Finally, the two indicators were combined to calculate the cumulative effect of ocean acidification and global warming.

To create the indicators, experimental data on the capacity of pteropods to produce their shell under ocean acidification scenarios were coupled with models describing chemical conditions (aragonite saturation state) of the oceans at present, in 2030 and 2050, under an optimistic carbon dioxide emission scenario (Representative Concentration Pathway 4.5) and a pessimistic projection (Representative Concentration Pathway 8.5).
Indicators were classified in five risk levels:
1: low, 2: medium, 3: high, 4: very high, and 5: critical corresponding to a decrease in shell production of 0-20%, 20-40%, 40-60%, 60-80%, and superior to 80% respectively, compared to the present average specific shell production. For temperature, risk levels were assessed as a function of average surface yearly temperature deviation from the maximum temperature at which organisms from the three species are now found. Finally, the two indicators were combined to calculate the cumulative effect of ocean acidification and global warming.

To create the indicators, experimental data on the capacity of pteropods to produce their shell under ocean acidification scenarios were coupled with models describing chemical conditions (aragonite saturation state) of the oceans at present, in 2030 and 2050, under an optimistic carbon dioxide emission scenario (Representative Concentration Pathway 4.5) and a pessimistic projection (Representative Concentration Pathway 8.5).
Indicators were classified in five risk levels:
1: low, 2: medium, 3: high, 4: very high, and 5: critical corresponding to a decrease in shell production of 0-20%, 20-40%, 40-60%, 60-80%, and superior to 80% respectively, compared to the present average specific shell production. For temperature, risk levels were assessed as a function of average surface yearly temperature deviation from the maximum temperature at which organisms from the three species are now found. Finally, the two indicators were combined to calculate the cumulative effect of ocean acidification and global warming.

To create the indicators, experimental data on the capacity of pteropods to produce their shell under ocean acidification scenarios were coupled with models describing chemical conditions (aragonite saturation state) of the oceans at present, in 2030 and 2050, under an optimistic carbon dioxide emission scenario (Representative Concentration Pathway 4.5) and a pessimistic projection (Representative Concentration Pathway 8.5).
Indicators were classified in five risk levels:
1: low, 2: medium, 3: high, 4: very high, and 5: critical corresponding to a decrease in shell production of 0-20%, 20-40%, 40-60%, 60-80%, and superior to 80% respectively, compared to the present average specific shell production. For temperature, risk levels were assessed as a function of average surface yearly temperature deviation from the maximum temperature at which organisms from the three species are now found. Finally, the two indicators were combined to calculate the cumulative effect of ocean acidification and global warming.

To create the indicators, experimental data on the capacity of pteropods to produce their shell under ocean acidification scenarios were coupled with models describing chemical conditions (aragonite saturation state) of the oceans at present, in 2030 and 2050, under an optimistic carbon dioxide emission scenario (Representative Concentration Pathway 4.5) and a pessimistic projection (Representative Concentration Pathway 8.5).
Indicators were classified in five risk levels:
1: low, 2: medium, 3: high, 4: very high, and 5: critical corresponding to a decrease in shell production of 0-20%, 20-40%, 40-60%, 60-80%, and superior to 80% respectively, compared to the present average specific shell production. For temperature, risk levels were assessed as a function of average surface yearly temperature deviation from the maximum temperature at which organisms from the three species are now found. Finally, the two indicators were combined to calculate the cumulative effect of ocean acidification and global warming.

To create the indicators, experimental data on the capacity of pteropods to produce their shell under ocean acidification scenarios were coupled with models describing chemical conditions (aragonite saturation state) of the oceans at present, in 2030 and 2050, under an optimistic carbon dioxide emission scenario (Representative Concentration Pathway 4.5) and a pessimistic projection (Representative Concentration Pathway 8.5).
Indicators were classified in five risk levels:
1: low, 2: medium, 3: high, 4: very high, and 5: critical corresponding to a decrease in shell production of 0-20%, 20-40%, 40-60%, 60-80%, and superior to 80% respectively, compared to the present average specific shell production. For temperature, risk levels were assessed as a function of average surface yearly temperature deviation from the maximum temperature at which organisms from the three species are now found. Finally, the two indicators were combined to calculate the cumulative effect of ocean acidification and global warming.

To create the indicators, experimental data on the capacity of pteropods to produce their shell under ocean acidification scenarios were coupled with models describing chemical conditions (aragonite saturation state) of the oceans at present, in 2030 and 2050, under an optimistic carbon dioxide emission scenario (Representative Concentration Pathway 4.5) and a pessimistic projection (Representative Concentration Pathway 8.5).
Indicators were classified in five risk levels:
1: low, 2: medium, 3: high, 4: very high, and 5: critical corresponding to a decrease in shell production of 0-20%, 20-40%, 40-60%, 60-80%, and superior to 80% respectively, compared to the present average specific shell production. For temperature, risk levels were assessed as a function of average surface yearly temperature deviation from the maximum temperature at which organisms from the three species are now found. Finally, the two indicators were combined to calculate the cumulative effect of ocean acidification and global warming.

To create the indicators, experimental data on the capacity of pteropods to produce their shell under ocean acidification scenarios were coupled with models describing chemical conditions (aragonite saturation state) of the oceans at present, in 2030 and 2050, under an optimistic carbon dioxide emission scenario (Representative Concentration Pathway 4.5) and a pessimistic projection (Representative Concentration Pathway 8.5).
Indicators were classified in five risk levels:
1: low, 2: medium, 3: high, 4: very high, and 5: critical corresponding to a decrease in shell production of 0-20%, 20-40%, 40-60%, 60-80%, and superior to 80% respectively, compared to the present average specific shell production. For temperature, risk levels were assessed as a function of average surface yearly temperature deviation from the maximum temperature at which organisms from the three species are now found. Finally, the two indicators were combined to calculate the cumulative effect of ocean acidification and global warming.

To create the indicators, experimental data on the capacity of pteropods to produce their shell under ocean acidification scenarios were coupled with models describing chemical conditions (aragonite saturation state) of the oceans at present, in 2030 and 2050, under an optimistic carbon dioxide emission scenario (Representative Concentration Pathway 4.5) and a pessimistic projection (Representative Concentration Pathway 8.5).
Indicators were classified in five risk levels:
1: low, 2: medium, 3: high, 4: very high, and 5: critical corresponding to a decrease in shell production of 0-20%, 20-40%, 40-60%, 60-80%, and superior to 80% respectively, compared to the present average specific shell production. For temperature, risk levels were assessed as a function of average surface yearly temperature deviation from the maximum temperature at which organisms from the three species are now found. Finally, the two indicators were combined to calculate the cumulative effect of ocean acidification and global warming.

To create the indicators, experimental data on the capacity of pteropods to produce their shell under ocean acidification scenarios were coupled with models describing chemical conditions (aragonite saturation state) of the oceans at present, in 2030 and 2050, under an optimistic carbon dioxide emission scenario (Representative Concentration Pathway 4.5) and a pessimistic projection (Representative Concentration Pathway 8.5).
Indicators were classified in five risk levels:
1: low, 2: medium, 3: high, 4: very high, and 5: critical corresponding to a decrease in shell production of 0-20%, 20-40%, 40-60%, 60-80%, and superior to 80% respectively, compared to the present average specific shell production. For temperature, risk levels were assessed as a function of average surface yearly temperature deviation from the maximum temperature at which organisms from the three species are now found. Finally, the two indicators were combined to calculate the cumulative effect of ocean acidification and global warming.

To create the indicators, experimental data on the capacity of pteropods to produce their shell under ocean acidification scenarios were coupled with models describing chemical conditions (aragonite saturation state) of the oceans at present, in 2030 and 2050, under an optimistic carbon dioxide emission scenario (Representative Concentration Pathway 4.5) and a pessimistic projection (Representative Concentration Pathway 8.5).
Indicators were classified in five risk levels:
1: low, 2: medium, 3: high, 4: very high, and 5: critical corresponding to a decrease in shell production of 0-20%, 20-40%, 40-60%, 60-80%, and superior to 80% respectively, compared to the present average specific shell production. For temperature, risk levels were assessed as a function of average surface yearly temperature deviation from the maximum temperature at which organisms from the three species are now found. Finally, the two indicators were combined to calculate the cumulative effect of ocean acidification and global warming.

To create the indicators, experimental data on the capacity of pteropods to produce their shell under ocean acidification scenarios were coupled with models describing chemical conditions (aragonite saturation state) of the oceans at present, in 2030 and 2050, under an optimistic carbon dioxide emission scenario (Representative Concentration Pathway 4.5) and a pessimistic projection (Representative Concentration Pathway 8.5).
Indicators were classified in five risk levels:
1: low, 2: medium, 3: high, 4: very high, and 5: critical corresponding to a decrease in shell production of 0-20%, 20-40%, 40-60%, 60-80%, and superior to 80% respectively, compared to the present average specific shell production. For temperature, risk levels were assessed as a function of average surface yearly temperature deviation from the maximum temperature at which organisms from the three species are now found. Finally, the two indicators were combined to calculate the cumulative effect of ocean acidification and global warming.

To create the indicators, experimental data on the capacity of pteropods to produce their shell under ocean acidification scenarios were coupled with models describing chemical conditions (aragonite saturation state) of the oceans at present, in 2030 and 2050, under an optimistic carbon dioxide emission scenario (Representative Concentration Pathway 4.5) and a pessimistic projection (Representative Concentration Pathway 8.5).
Indicators were classified in five risk levels:
1: low, 2: medium, 3: high, 4: very high, and 5: critical corresponding to a decrease in shell production of 0-20%, 20-40%, 40-60%, 60-80%, and superior to 80% respectively, compared to the present average specific shell production. For temperature, risk levels were assessed as a function of average surface yearly temperature deviation from the maximum temperature at which organisms from the three species are now found. Finally, the two indicators were combined to calculate the cumulative effect of ocean acidification and global warming.

To create the indicators, experimental data on the capacity of pteropods to produce their shell under ocean acidification scenarios were coupled with models describing chemical conditions (aragonite saturation state) of the oceans at present, in 2030 and 2050, under an optimistic carbon dioxide emission scenario (Representative Concentration Pathway 4.5) and a pessimistic projection (Representative Concentration Pathway 8.5).
Indicators were classified in five risk levels:
1: low, 2: medium, 3: high, 4: very high, and 5: critical corresponding to a decrease in shell production of 0-20%, 20-40%, 40-60%, 60-80%, and superior to 80% respectively, compared to the present average specific shell production. For temperature, risk levels were assessed as a function of average surface yearly temperature deviation from the maximum temperature at which organisms from the three species are now found. Finally, the two indicators were combined to calculate the cumulative effect of ocean acidification and global warming.

To create the indicators, experimental data on the capacity of pteropods to produce their shell under ocean acidification scenarios were coupled with models describing chemical conditions (aragonite saturation state) of the oceans at present, in 2030 and 2050, under an optimistic carbon dioxide emission scenario (Representative Concentration Pathway 4.5) and a pessimistic projection (Representative Concentration Pathway 8.5).
Indicators were classified in five risk levels:
1: low, 2: medium, 3: high, 4: very high, and 5: critical corresponding to a decrease in shell production of 0-20%, 20-40%, 40-60%, 60-80%, and superior to 80% respectively, compared to the present average specific shell production. For temperature, risk levels were assessed as a function of average surface yearly temperature deviation from the maximum temperature at which organisms from the three species are now found. Finally, the two indicators were combined to calculate the cumulative effect of ocean acidification and global warming.

To create the indicators, experimental data on the capacity of pteropods to produce their shell under ocean acidification scenarios were coupled with models describing chemical conditions (aragonite saturation state) of the oceans at present, in 2030 and 2050, under an optimistic carbon dioxide emission scenario (Representative Concentration Pathway 4.5) and a pessimistic projection (Representative Concentration Pathway 8.5).
Indicators were classified in five risk levels:
1: low, 2: medium, 3: high, 4: very high, and 5: critical corresponding to a decrease in shell production of 0-20%, 20-40%, 40-60%, 60-80%, and superior to 80% respectively, compared to the present average specific shell production. For temperature, risk levels were assessed as a function of average surface yearly temperature deviation from the maximum temperature at which organisms from the three species are now found. Finally, the two indicators were combined to calculate the cumulative effect of ocean acidification and global warming.

To create the indicators, experimental data on the capacity of pteropods to produce their shell under ocean acidification scenarios were coupled with models describing chemical conditions (aragonite saturation state) of the oceans at present, in 2030 and 2050, under an optimistic carbon dioxide emission scenario (Representative Concentration Pathway 4.5) and a pessimistic projection (Representative Concentration Pathway 8.5).
Indicators were classified in five risk levels:
1: low, 2: medium, 3: high, 4: very high, and 5: critical corresponding to a decrease in shell production of 0-20%, 20-40%, 40-60%, 60-80%, and superior to 80% respectively, compared to the present average specific shell production. For temperature, risk levels were assessed as a function of average surface yearly temperature deviation from the maximum temperature at which organisms from the three species are now found. Finally, the two indicators were combined to calculate the cumulative effect of ocean acidification and global warming.

To create the indicators, experimental data on the capacity of pteropods to produce their shell under ocean acidification scenarios were coupled with models describing chemical conditions (aragonite saturation state) of the oceans at present, in 2030 and 2050, under an optimistic carbon dioxide emission scenario (Representative Concentration Pathway 4.5) and a pessimistic projection (Representative Concentration Pathway 8.5).
Indicators were classified in five risk levels:
1: low, 2: medium, 3: high, 4: very high, and 5: critical corresponding to a decrease in shell production of 0-20%, 20-40%, 40-60%, 60-80%, and superior to 80% respectively, compared to the present average specific shell production. For temperature, risk levels were assessed as a function of average surface yearly temperature deviation from the maximum temperature at which organisms from the three species are now found. Finally, the two indicators were combined to calculate the cumulative effect of ocean acidification and global warming.

DHM projections are computed by comparing the ensemble mean of SST projections from CMIP5 models outputs (for IPCC scenarios RCP 4.5 and RCP 8.5) to Reynolds climatology.
The DHM is defined as the accumulation, over a time period of 4 months, of temperature differences between models projections and Reynolds climatology (only positive difference, i.e. projection temperature above climatology, are accounted).
A yearly DHM corresponds to the maximum 4-month-DHM observed in the year. A level 2 DHM corresponds the 4-month-DHM equal to or above 2°C. The level 2 frequency corresponds to the number of occurrence of yearly level 2 in 10 years.

Ensemble mean computation: https://github.com/IOC-CODE/cmip5_projections/blob/master/make_ensembleM...
DHM and DHM-level 2 frequency computation: https://github.com/IOC-CODE/cmip5_projections/blob/master/make_dhm.py

DHM projections are computed by comparing the ensemble mean of SST projections from CMIP5 models outputs (for IPCC scenarios RCP 4.5 and RCP 8.5) to Reynolds climatology.
The DHM is defined as the accumulation, over a time period of 4 months, of temperature differences between models projections and Reynolds climatology (only positive difference, i.e. projection temperature above climatology, are accounted).
A yearly DHM corresponds to the maximum 4-month-DHM observed in the year. A level 2 DHM corresponds the 4-month-DHM equal to or above 2°C. The level 2 frequency corresponds to the number of occurrence of yearly level 2 in 10 years.

Ensemble mean computation: https://github.com/BrunoCombal/climate/blob/master/make_ensembleMean_tyx.py
DHM and DHM-level 2 frequency computation: https://github.com/BrunoCombal/climate/blob/master/make_dhm.py

DHM projections are computed by comparing the ensemble mean of SST projections from CMIP5 models outputs (for IPCC scenarios RCP 4.5 and RCP 8.5) to Reynolds climatology.
The DHM is defined as the accumulation, over a time period of 4 months, of temperature differences between models projections and Reynolds climatology (only positive difference, i.e. projection temperature above climatology, are accounted).
A yearly DHM corresponds to the maximum 4-month-DHM observed in the year. A level 2 DHM corresponds the 4-month-DHM equal to or above 2°C. The level 2 frequency corresponds to the number of occurrence of yearly level 2 in 10 years.

Ensemble mean computation: https://github.com/BrunoCombal/climate/blob/master/make_ensembleMean_tyx.py
DHM and DHM-level 2 frequency computation: https://github.com/BrunoCombal/climate/blob/master/make_dhm.py

DHM projections are computed by comparing the ensemble mean of SST projections from CMIP5 models outputs (for IPCC scenarios RCP 4.5 and RCP 8.5) to Reynolds climatology.
The DHM is defined as the accumulation, over a time period of 4 months, of temperature differences between models projections and Reynolds climatology (only positive difference, i.e. projection temperature above climatology, are accounted).
A yearly DHM corresponds to the maximum 4-month-DHM observed in the year. A level 2 DHM corresponds the 4-month-DHM equal to or above 2°C. The level 2 frequency corresponds to the number of occurrence of yearly level 2 in 10 years.

Ensemble mean computation: https://github.com/BrunoCombal/climate/blob/master/make_ensembleMean_tyx.py
DHM and DHM-level 2 frequency computation: https://github.com/BrunoCombal/climate/blob/master/make_dhm.py

DHM projections are computed by comparing the ensemble mean of SST projections from CMIP5 models outputs (for IPCC scenarios RCP 4.5 and RCP 8.5) to Reynolds climatology.
The DHM is defined as the accumulation, over a time period of 4 months, of temperature differences between models projections and Reynolds climatology (only positive difference, i.e. projection temperature above climatology, are accounted).
A yearly DHM corresponds to the maximum 4-month-DHM observed in the year. A level 2 DHM corresponds the 4-month-DHM equal to or above 2°C. The level 2 frequency corresponds to the number of occurrence of yearly level 2 in 10 years.

Ensemble mean computation: https://github.com/BrunoCombal/climate/blob/master/make_ensembleMean_tyx.py
DHM and DHM-level 2 frequency computation: https://github.com/BrunoCombal/climate/blob/master/make_dhm.py

DHM projections are computed by comparing the ensemble mean of SST projections from CMIP5 models outputs (for IPCC scenarios RCP 4.5 and RCP 8.5) to Reynolds climatology.
The DHM is defined as the accumulation, over a time period of 4 months, of temperature differences between models projections and Reynolds climatology (only positive difference, i.e. projection temperature above climatology, are accounted).
A yearly DHM corresponds to the maximum 4-month-DHM observed in the year. A level 2 DHM corresponds the 4-month-DHM equal to or above 2°C. The level 2 frequency corresponds to the number of occurrence of yearly level 2 in 10 years.

Ensemble mean computation: https://github.com/BrunoCombal/climate/blob/master/make_ensembleMean_tyx.py
DHM and DHM-level 2 frequency computation: https://github.com/BrunoCombal/climate/blob/master/make_dhm.py

DHM projections are computed by comparing the ensemble mean of SST projections from CMIP5 models outputs (for IPCC scenarios RCP 4.5 and RCP 8.5) to Reynolds climatology.
The DHM is defined as the accumulation, over a time period of 4 months, of temperature differences between models projections and Reynolds climatology (only positive difference, i.e. projection temperature above climatology, are accounted).
A yearly DHM corresponds to the maximum 4-month-DHM observed in the year. A level 2 DHM corresponds the 4-month-DHM equal to or above 2°C. The level 2 frequency corresponds to the number of occurrence of yearly level 2 in 10 years.

Ensemble mean computation: https://github.com/BrunoCombal/climate/blob/master/make_ensembleMean_tyx.py
DHM and DHM-level 2 frequency computation: https://github.com/BrunoCombal/climate/blob/master/make_dhm.py

DHM projections are computed by comparing the ensemble mean of SST projections from CMIP5 models outputs (for IPCC scenarios RCP 4.5 and RCP 8.5) to Reynolds climatology.
The DHM is defined as the accumulation, over a time period of 4 months, of temperature differences between models projections and Reynolds climatology (only positive difference, i.e. projection temperature above climatology, are accounted).
A yearly DHM corresponds to the maximum 4-month-DHM observed in the year. A level 2 DHM corresponds the 4-month-DHM equal to or above 2°C. The level 2 frequency corresponds to the number of occurrence of yearly level 2 in 10 years.

Ensemble mean computation: https://github.com/BrunoCombal/climate/blob/master/make_ensembleMean_tyx.py
DHM and DHM-level 2 frequency computation: https://github.com/BrunoCombal/climate/blob/master/make_dhm.py

DHM projections are computed by comparing the ensemble mean of SST projections from CMIP5 models outputs (for IPCC scenarios RCP 4.5 and RCP 8.5) to Reynolds climatology.
The DHM is defined as the accumulation, over a time period of 4 months, of temperature differences between models projections and Reynolds climatology (only positive difference, i.e. projection temperature above climatology, are accounted).
A yearly DHM corresponds to the maximum 4-month-DHM observed in the year. A level 2 DHM corresponds the 4-month-DHM equal to or above 2°C. The level 2 frequency corresponds to the number of occurrence of yearly level 2 in 10 years.

Ensemble mean computation: https://github.com/BrunoCombal/climate/blob/master/make_ensembleMean_tyx.py
DHM and DHM-level 2 frequency computation: https://github.com/BrunoCombal/climate/blob/master/make_dhm.py

DHM projections are computed by comparing the ensemble mean of SST projections from CMIP5 models outputs (for IPCC scenarios RCP 4.5 and RCP 8.5) to Reynolds climatology.
The DHM is defined as the accumulation, over a time period of 4 months, of temperature differences between models projections and Reynolds climatology (only positive difference, i.e. projection temperature above climatology, are accounted).
A yearly DHM corresponds to the maximum 4-month-DHM observed in the year. A level 2 DHM corresponds the 4-month-DHM equal to or above 2°C. The level 2 frequency corresponds to the number of occurrence of yearly level 2 in 10 years.

Ensemble mean computation: https://github.com/BrunoCombal/climate/blob/master/make_ensembleMean_tyx.py
DHM and DHM-level 2 frequency computation: https://github.com/BrunoCombal/climate/blob/master/make_dhm.py

Annual landings by bottom-impacting gear types, including dredges and bottom trawls, were extracted from the Sea Around Us database from 1950 to 2006. As the Sea Around Us extended catch data from 2007 to 2010 are not aggregated by gear type, we estimated the catch of bottom-impacting gear types (i.e., trawling and dredging gears) by calculating the proportions of these gear types to the total catch by each fishing country and LME combination in 2006. We then used these proportions to estimate the catch by bottom-impacting gear types from 2007 to 2010. The undeveloped and the developing categories will by definition have disappeared by the end of the time series since the scale is based on the year with maximum catch (which has to occur somewhere on the time line)
References:
Froese, R. and K. Kesner-Reyes (2002). Impact of Fishing on the Abundance of Marine Species. Rainer. ICES CM 2002/L:12, 15 p.
Grainger, R.J.R. and S. Garcia (1996). Chronicles of marine fisheries landings (1950-1994): trend analysis and fisheries potential. FAO Fish. Tech. Pap. 359, 51 p.
Pauly, D., J. Alder, S. Booth, W.W.L. Cheung, V. Christensen, C. Close, U.R. Sumaila, W. Swartz, A. Tavakolie, R. Watson, L. Wood and D. Zeller. 2008. Fisheries in Large Marine Ecosystems: Descriptions and Diagnoses. p. 23-40. In: K. Sherman and G. Hempel (eds.) The UNEP Large Marine Ecosystem Report: a Perspective on Changing Conditions in LMEs of the World’s Regional Seas. UNEP Regional Seas Reports and Studies No. 182.

Spatio-temporal trends in fishing power globally, and by continents, countries, gross registered tonnage class, and vessel/gear types.
References:
Anticamara, J.A., R. Watson, A. Gelchu, J. Beblow, D. Pauly. MS. Global fishing effort (1950-2010): Trends, gaps, and implications.
FAO, 2009. State of the world's fisheries and aquaculture 2008. Food and Agriculture Organization of the United Nations, Rome.
Gelchu, A., Pauly, D., 2007. Growth and distribution of port-based global fishing effort within countries' EEZ from 1970-1995. Fisheries Centre Research Report, Vancouver, Canada.
Watson, R., W.W.L. Cheung, J. Anticamara, R.U. Sumaila, D. Zeller and D. Pauly. 2013. Global marine yield halved as fishing intensity redoubles. Fish and Fisheries, 14: 493-503.

Calculation details are given above. Limitations are similar to the MTI, with the added uncertainty caused by the assumption of energy transfer efficiency of 10% between trophic levels (Pauly and Christensen 1995). Supplements the MTI and should preferably be interpreted in connection with this index.

Based on analysis of 1066 species of marine fish and invertebrates, representing the major commercially exploited species, as reported in the FAO fisheries statistics (see http://www.fao.org). Future distributions of these species are projected using a dynamic bioclimate envelope model under the SRES A1B scenario (see Cheung and others 2008b, 2009) while primary production is projected by empirical models (Behrenfeld and Falkowski 1997, Carr 2002, Marra and others 2003, Sarmiento and others 2004). The annual maximum catch potential for ½˚ by ½˚ spatial cells is calculated based on the model of Cheung and others (2008). The empirical model estimates a species’ maximum catch potential (MSY) based on the total primary production within its exploitable range, the area of its geographic range, its trophic level, and includes terms correcting the biases from the observed catch potential. Development of alternative methods should be encouraged.
References:
Bakun, A. 1990. Global climate change and intensification of coastal ocean upwelling. Science, 247, 198-201
Behrenfeld, M.J. and Falkowski, P.G. 1997. Photosynthetic rates derived from satellite-based chlorophyll concentration. Limnology and Oceangraphy, 42(1): 1-20.
Carr, M.E., 2002. Estimation of potential productivity in Eastern Boundary Currents using remote sensing. Deep Sea Research Part II, 49: 59-80.
Cheung, W.W.L., Close, C., Lam, V.W.Y., Watson, R. and Pauly, D. 2008. Application of macroecological theory to predict effects of climate change on global fisheries potential. Marine Ecology Progress Series, 365: 187-197
Cheung, W.W.L., Lam, V.W.Y., Sarmiento, J.L., Kearney, K., Watson, R., Zeller, D. and Pauly, D. 2010. Large-scale redistribution of maximum catch potential in the global ocean under climate change. Global Change Biology, 16: 24-35
Diaz, R.J., Rosenberg, R. 2008. Spreading dead zones and consequences for marine ecosystems. Science, 321, 926-929
Marra, J., Ho, C., Trees, C.C. 2003. An Algorithm for the Calculation of Primary Productivity from Remote Sensing Data. Lamont-Doherty Earth Obs., Palisades, New York. 27 p.
IPCC 2007 Summary for Policymakers. In: Climate Change 2007: The Physical Science Basis. Working Group I Contribution to the Fourth Assessment Report of the IPCC (eds S. Solomon, D. Qin, M. Manning et al.), pp. 1-18. Cambridge University Press, Cambridge
Sarmiento, J.L. et al. 2004. Response of ocean ecosystems to climate warming. Global Biogeochemical Cycles, 18(3): GB3003.1-GB3004.23.

Based on analysis of 1066 species of marine fish and invertebrates, representing the major commercially exploited species, as reported in the FAO fisheries statistics (see http://www.fao.org). Future distributions of these species are projected using a dynamic bioclimate envelope model under the SRES A1B scenario (see Cheung and others 2008b, 2009) while primary production is projected by empirical models (Behrenfeld and Falkowski 1997, Carr 2002, Marra and others 2003, Sarmiento and others 2004). The annual maximum catch potential for ½˚ by ½˚ spatial cells is calculated based on the model of Cheung and others (2008). The empirical model estimates a species’ maximum catch potential (MSY) based on the total primary production within its exploitable range, the area of its geographic range, its trophic level, and includes terms correcting the biases from the observed catch potential. Development of alternative methods should be encouraged.
References:
Bakun, A. 1990. Global climate change and intensification of coastal ocean upwelling. Science, 247, 198-201
Behrenfeld, M.J. and Falkowski, P.G. 1997. Photosynthetic rates derived from satellite-based chlorophyll concentration. Limnology and Oceangraphy, 42(1): 1-20.
Carr, M.E., 2002. Estimation of potential productivity in Eastern Boundary Currents using remote sensing. Deep Sea Research Part II, 49: 59-80.
Cheung, W.W.L., Close, C., Lam, V.W.Y., Watson, R. and Pauly, D. 2008. Application of macroecological theory to predict effects of climate change on global fisheries potential. Marine Ecology Progress Series, 365: 187-197
Cheung, W.W.L., Lam, V.W.Y., Sarmiento, J.L., Kearney, K., Watson, R., Zeller, D. and Pauly, D. 2010. Large-scale redistribution of maximum catch potential in the global ocean under climate change. Global Change Biology, 16: 24-35
Diaz, R.J., Rosenberg, R. 2008. Spreading dead zones and consequences for marine ecosystems. Science, 321, 926-929
Marra, J., Ho, C., Trees, C.C. 2003. An Algorithm for the Calculation of Primary Productivity from Remote Sensing Data. Lamont-Doherty Earth Obs., Palisades, New York. 27 p.
IPCC 2007 Summary for Policymakers. In: Climate Change 2007: The Physical Science Basis. Working Group I Contribution to the Fourth Assessment Report of the IPCC (eds S. Solomon, D. Qin, M. Manning et al.), pp. 1-18. Cambridge University Press, Cambridge
Sarmiento, J.L. et al. 2004. Response of ocean ecosystems to climate warming. Global Biogeochemical Cycles, 18(3): GB3003.1-GB3004.23.

Based on analysis of 1066 species of marine fish and invertebrates, representing the major commercially exploited species, as reported in the FAO fisheries statistics (see http://www.fao.org). Future distributions of these species are projected using a dynamic bioclimate envelope model under the SRES A1B scenario (see Cheung and others 2008b, 2009) while primary production is projected by empirical models (Behrenfeld and Falkowski 1997, Carr 2002, Marra and others 2003, Sarmiento and others 2004). The annual maximum catch potential for ½˚ by ½˚ spatial cells is calculated based on the model of Cheung and others (2008). The empirical model estimates a species’ maximum catch potential (MSY) based on the total primary production within its exploitable range, the area of its geographic range, its trophic level, and includes terms correcting the biases from the observed catch potential. Development of alternative methods should be encouraged.
References:
Bakun, A. 1990. Global climate change and intensification of coastal ocean upwelling. Science, 247, 198-201
Behrenfeld, M.J. and Falkowski, P.G. 1997. Photosynthetic rates derived from satellite-based chlorophyll concentration. Limnology and Oceangraphy, 42(1): 1-20.
Carr, M.E., 2002. Estimation of potential productivity in Eastern Boundary Currents using remote sensing. Deep Sea Research Part II, 49: 59-80.
Cheung, W.W.L., Close, C., Lam, V.W.Y., Watson, R. and Pauly, D. 2008. Application of macroecological theory to predict effects of climate change on global fisheries potential. Marine Ecology Progress Series, 365: 187-197
Cheung, W.W.L., Lam, V.W.Y., Sarmiento, J.L., Kearney, K., Watson, R., Zeller, D. and Pauly, D. 2010. Large-scale redistribution of maximum catch potential in the global ocean under climate change. Global Change Biology, 16: 24-35
Diaz, R.J., Rosenberg, R. 2008. Spreading dead zones and consequences for marine ecosystems. Science, 321, 926-929
Marra, J., Ho, C., Trees, C.C. 2003. An Algorithm for the Calculation of Primary Productivity from Remote Sensing Data. Lamont-Doherty Earth Obs., Palisades, New York. 27 p.
IPCC 2007 Summary for Policymakers. In: Climate Change 2007: The Physical Science Basis. Working Group I Contribution to the Fourth Assessment Report of the IPCC (eds S. Solomon, D. Qin, M. Manning et al.), pp. 1-18. Cambridge University Press, Cambridge
Sarmiento, J.L. et al. 2004. Response of ocean ecosystems to climate warming. Global Biogeochemical Cycles, 18(3): GB3003.1-GB3004.23.

Based on analysis of 1066 species of marine fish and invertebrates, representing the major commercially exploited species, as reported in the FAO fisheries statistics (see http://www.fao.org). Future distributions of these species are projected using a dynamic bioclimate envelope model under the SRES A1B scenario (see Cheung and others 2008b, 2009) while primary production is projected by empirical models (Behrenfeld and Falkowski 1997, Carr 2002, Marra and others 2003, Sarmiento and others 2004). The annual maximum catch potential for ½˚ by ½˚ spatial cells is calculated based on the model of Cheung and others (2008). The empirical model estimates a species’ maximum catch potential (MSY) based on the total primary production within its exploitable range, the area of its geographic range, its trophic level, and includes terms correcting the biases from the observed catch potential. Development of alternative methods should be encouraged.
References:
Bakun, A. 1990. Global climate change and intensification of coastal ocean upwelling. Science, 247, 198-201
Behrenfeld, M.J. and Falkowski, P.G. 1997. Photosynthetic rates derived from satellite-based chlorophyll concentration. Limnology and Oceangraphy, 42(1): 1-20.
Carr, M.E., 2002. Estimation of potential productivity in Eastern Boundary Currents using remote sensing. Deep Sea Research Part II, 49: 59-80.
Cheung, W.W.L., Close, C., Lam, V.W.Y., Watson, R. and Pauly, D. 2008. Application of macroecological theory to predict effects of climate change on global fisheries potential. Marine Ecology Progress Series, 365: 187-197
Cheung, W.W.L., Lam, V.W.Y., Sarmiento, J.L., Kearney, K., Watson, R., Zeller, D. and Pauly, D. 2010. Large-scale redistribution of maximum catch potential in the global ocean under climate change. Global Change Biology, 16: 24-35
Diaz, R.J., Rosenberg, R. 2008. Spreading dead zones and consequences for marine ecosystems. Science, 321, 926-929
Marra, J., Ho, C., Trees, C.C. 2003. An Algorithm for the Calculation of Primary Productivity from Remote Sensing Data. Lamont-Doherty Earth Obs., Palisades, New York. 27 p.
IPCC 2007 Summary for Policymakers. In: Climate Change 2007: The Physical Science Basis. Working Group I Contribution to the Fourth Assessment Report of the IPCC (eds S. Solomon, D. Qin, M. Manning et al.), pp. 1-18. Cambridge University Press, Cambridge
Sarmiento, J.L. et al. 2004. Response of ocean ecosystems to climate warming. Global Biogeochemical Cycles, 18(3): GB3003.1-GB3004.23.

TLs are assigned to all catches from a given area, typically based on information in FishBase or SeaLifeBase. The weighted TL of the catch is then calculated by weighting the species/group TL with the corresponding catch level. Among the limitations are: Diagnosing fishing down from the mean trophic level of landings is problematic as landings reflect abundances only crudely; The trophic level is typically assumed constant for a given species/group, but may change over time, notably if the size of individuals in the catches change; Changes in MTI may reflect spatial expansion of fisheries, which can cause temporary increases in the index.
The MTI is closely related to the Fishing-in-Balance (FiB) index, and should preferably be interpreted in connection with this index.
References:
CBD (2004) Annex I, decision VII/30. The 2020 biodiversity target: a framework for implementation, p. 351. Decisions from the Seventh Meeting of the Conference of the Parties of the Convention on Biological Diversity, Kuala Lumpur, 9–10 and 27 February 2004. Montreal: Secretariat of the CBD.
Cheung, W., R. Watson, T. Morato, T. Pitcher and D. Pauly. (2007) Change of intrinsic vulnerability in the global fish catch. Marine Ecology Progress Series 333: 1-12.
Pauly, D., V. Christensen, J. Dalsgaard, R. Froese and F.C. Torres Jr. (1998). Fishing down marine food webs. Science 279: 860-863.
Pauly, D. and R. Watson (2005). Background and interpretation of the ‘Marine Trophic Index’ as a measure of biodiversity. Philosophical Transactions of the Royal Society: Biological Sciences 360: 415-423.

1) Assemble the catch data to be mapped, here consisting mainly of the catch data reported by member countries to FAO, and distributed via the Fishstat database after their assignment to FAO statistical areas, complemented by data from the FAO’s ‘Atlas of Tuna and Billfish Statistics’ (www.fao.org/fishery/statistics/tuna-atlas/4/en);

2) Create, for each taxon (species, genus or family) for which at least one country reports landing, distributions range map (for tuna mainly based on FishBase; www.fishbase.org);

3) Allocate the catch reported in (1) to the distribution range in (2) subject to the constraint that an access agreement (or traditional access in pre-EEZ times) must exist for the a catch to be allocated to cells that are part of an EEZ other than that of the reporting country;

4) When necessary, identify the reason(s) why a catch cannot be allocated, which may be due to (a) a faulty distribution map, (b) the non-availability of an access agreement, or (c) one or several other constraints – omitted here - not being met;

5) Aggregate the half degree cells (and the catch assigned to them) into a large area of interest, e.g., the EEZs of maritime countries, large marine ecosystems, or here, the Open Ocean part of FAO statistical areas (Watson et al. 2004; Pauly et al. 2008, and see www.seaaroundus.org).

References:

Watson, R., A. Kitchingman, A. Gelchu and D. Pauly. 2004. Mapping global fisheries: sharpening our focus. Fish and Fisheries 5: 168-177.

Pauly, D., J. Alder, S. Booth, W.W.L. Cheung, V. Christensen, C. Close, U.R. Sumaila, W. Swartz, A. Tavakolie, R. Watson, L. Wood and D. Zeller. 2008. Fisheries in Large Marine Ecosystems: Descriptions and Diagnoses. p. 23-40. In: K. Sherman and G. Hempel (eds.) The UNEP Large Marine Ecosystem Report: a Perspective on Changing Conditions in LMEs of the World’s Regional Seas. UNEP Regional Seas Reports and Studies No. 182.

A data scoping exercise was undertaken to gather secondary information and primary data from all available sources (Internet, ocean governance organisations, NGOs) that provided coverage of the global ocean. All existing spatial data and corresponding metadata were imported and examined using GIS software.
Several ocean governance agreements did not exist in a spatial data format and therefore had to be created from secondary information.
Details on the methodology are found in:
Baldwin, K. and R. Mahon. 2014. Spatial analysis of ocean governance at the global level: Spatial data availability, accuracy, coordinate systems and considerations for practice. Centre for Resource Management and Environmental Studies, The University of the West Indies, Cave Hill Campus, Barbados, CERMES Technical Report No 74: 19pp, http://www.cavehill.uwi.edu/cermes/docs/technical_reports/baldwin_mahon_2014_twap_global_gis_analysis_ctr_74.aspx