CLIMsystems Blog

Observed and Modelled Marine Environmental Change

Authors: Yinpeng Li and Peter Urich


The potentially deadly consequences of the ocean's declining oxygen levels on marine life (and impacts on human life), have been studied for many years. A new study published in the science journal Nature (Schmidtko et al., 2017) found that the ocean's worldwide oxygen content declined by more than 2% between 1960 and 2010. That number may seem small but its significance is not.

The research team found that their models predicted a drop in oceanic oxygen by up to 7% by the year 2100. This may not sound like a huge decrease but it would be enough to kill or seriously harm most sea life, including fish species that humans rely on for food. The ocean already contains just enough oxygen to support marine wildlife. Further exacerbating the problem is the fact that warm water cannot hold onto as much dissolved oxygen as cooler water. This means that as the ocean heats up, it will lose oxygen without being able to replace it. The team predicts that this will cause serious problems in coastal regions that rely on marine tourism and fishing for income, assuming no actions are taken to mitigate global warming.

The authors concluded by noting that "far-reaching implications for marine ecosystems and fisheries can be expected."

An algae-covered public beach in Qingdao, China, 2013. France-Presse — Getty Images

Climate change drivers of marine environmental change

Four principal climate drivers -- pH, temperature, oxygen concentration and food availability -- as identified by the Intergovernmental Panel on Climate Change (IPCC) have wide r4angin impacts on the structure of marine ecosystems, their functioning and capacity to adapt. Climate change is likely to perturb all four with potentially very serious consequences.

Each of these variables can cause substantial change in the physical, chemical, and biological environment, affecting the ocean’s biogeochemical cycles and ecosystems in ways that we are only beginning to fathom (Gruber 2011; Doney et al. 2011; Bopp et al. 2013).

Ocean temperature warming is the result of the heat transfer between the warming air and the ocean’s surfaces. It affects organisms and biogeochemical cycles directly as well as through an increase of upper-ocean stratification.

Ocean temperature warming is the result of the heat transfer between the warming air and the ocean’s surfaces. It affects organisms and biogeochemical cycles directly as well as through an increase of upper-ocean stratification.

Ocean deoxygenation is the decrease of dissolved oxygen levels in the ocean. It is bound to occur in a warming ocean and is exacerbated by stratification, causing stress to macro-organisms that critically depend on sufficient levels of oxygen.

Primary production: although regionally variable, the combined effect of these changes is predicted to be an overall global decrease in primary production (PP), which is the ultimate determinant of food availability to marine ecosystems

Marine ecosystems provide primary protein and regulation of Earth’s climate via the uptake and storage of atmospheric carbon dioxide. One in seven people derive socio-economic value from these ecosystems.

Henson et al. (2017) demonstrated that the exposure of marine ecosystems to climate change-induced stress can be drastically reduced via climate mitigation. They achieved this result by analysing CMIP5 model results. If within the next 15 years there is a concerted effort to mitigate greenhouse gases, susceptibility could be reduced by 34%. Effectively this would buy an additional 20 years for adaptation in marine ecological and socioeconomic systems.

Figure1. Ocean climate change responses and feedbacks
Figure 2: Effect of mitigation on the global emergence in drivers of ecosystem stress. The proportion of the ocean in each year 1900–2100 affected by multiple stress (>1 driver) and quadruple stress (all 4 drivers) in the ‘business-as-usual’ scenario (RCP8.5) and a mitigation scenario (RCP4.5). Shaded areas represent ±1 inter-model s.d. (Henson et al., 2017, figure 4)
Figure 3: Amount of dissolved oxygen and changes per decade since 1960. a, Global oxygen inventory (dissolved oxygen, colour coded). Lines indicate boundaries of oxygen-minimum zones (OMZs): dashed-dotted, regions with less than 80μmol kg-1 oxygen anywhere within the water column; dashed lines and solid lines similarly represent regions with less than 40μmol kg-1 oxygen and 20μmol kg-1 oxygen, respectively. b, Change in dissolved oxygen per decade (colour coded). Lines show OMZs as in a. (From Schmidtko et al., 2017 figure 1)
Figure 4: Dissolved oxygen concentration at surface change projected by CMIP5 models (ensemble median), % per degree global warming (Generated using SimCLIM for ArcGIS Marine).
Figure 5: global mean dissolved oxygen concentration at surface changes in different RCP scenarios, GCM ensemble median (Applied SimCLIM for ArcGIS Marine).

SimCLIM for ArcGIS Marine data represented the CMIP5 model results and the findings of Henson et al., The data and software generated consistent results with observed trends in marine dissolved oxygen concentration change and the variations across ocean basins. As an Esri add-in, the SimCLIM for ArcGIS Marine tool proved to very useful and convenient for modelling marine environment change.

SimCLIM for ArcGIS Marine - data and methodology

The SimCLIM for ArcGIS Marine add-in is an ArcGIS tool developed by CLIMsystems, for the purpose of exploring the impact of climate change on marine biogeochemical cycles. This add-in incorporates output from the CMIP5 (IPCC AR5) GCM data into a simple tool where users can select any year from 1995 up to 2100 to investigate the projections and changes in nine biogeochemical related variables for the whole global ocean area. The data generated from SimCLIM for ArcGIS Marine add-in can help marine ecosystem researchers, marine resource managers, nature conservationists, managers, planners, policy makers, and general public by providing high resolution maps and state of the art scientific information.

Pattern scaling approach data processing

The ocean biogeochemical variables change patterns were processed using the monthly output of GCMs.

(1) Global area weighted means of each year from 2006 to 2100 were calculated for all the available GCM’s RCP runs.

(2) The GCM ensemble means of each RCP were calculated.

(3) The 95 year values were fitted to smooth curved polynomial lines, that is the global annual mean change value (GV). The difference between 1995 and a projection year is called

Global curves for sea surface temperature.

(4) For a given climate variable V, its anomaly for a particular grid cell (i), month (j) and year or period (y) under an RCP45: where being the annual global mean change of variable V.

(5) The local change pattern value () was calculated from the GCM simulation anomaly () using linear least squares regression, that is, the slope of the fitted linear line.

where m is the number of future sample periods used, 5 year average as a period, 19 period from 2006 to 2100.

The global curves and local change patterns were stored in the SimCLIM for ArcGIS Marine add-in. When the user selects a variable and a projection year, an equation (1) is applied to calculate the changes between the baseline period and the projection year. The projection in the SimCLIM for ArcGIS Marine add-in is the baseline plus the changes.

This number is determined by GCM data availability (not all the GCMs have marine biogeochemical components, and not every model has all the RCP experiments), and data quality (some GCMs have missing data, and other GCMs have data quality problems). Table 1 lists the complete data stored in the SimCLIM for ArcGIS Marine add-in.

Table 1 SimCLIM for ArcGIS Marine add-in GCM variable availability (all at a resolution of 0.25 degrees (approximately 27.75 kilometres)

CANESM2 vv v vv
CNRM-CM5vvvv vvvv
CSIRO-MK3-6-0 v
HADGEM2-CCvvvvv vvv
HADGEM2-ESvvvvv vvv

Global mean changes in SST, pH, O2 and PP projected by CMIP5 models, ensemble median of all available models (smoothed curves).

Ocean pH

About a third of the CO2 released in the atmosphere dissolves in the oceans, where it slightly lowers the pH. This effect is known as ocean acidification.

Ocean acidification is of great concern: small changes in pH impact the CaCO3-CO2 equilibrium thus slowing coral growth and weakening the coral that does grow under such conditions.

The image shows the result from 12 models from the CMIP5 data for changes in pH. Because pH is a log-scale unit, the ratio of pH for 1995 and 2035 is presented.

The redder colour shows a stronger change. This is mostly occurring in shallower areas, as there is an effect from temperature as well.

As the model-data (using grid-cells of 0.25°x0.25°) does not cover partial cells close to the coastline, the FILTER function of ArcGIS was used to fill in the gaps.

The pH-information is combined with other data:

  • country data (grey land and black boarders)
  • coral reef locations (purple)
  • ocean bathymetry (from ETOPO1, through ArcGIS-online) (this layer is visible because the pH layer has been made 25% transparent)

The image is a beautiful example of how a toolbar (the CLIMsystems SimCLIM for ArcGIS Marine toolbar, in addition to the “Climate” toolbar), combines perfectly with the functionality of ArcGIS.

Net Primary Production

Changes in the surface chemistry and temperature of the oceans because of climate change, impact the primary production from phytoplankton. The image shows the distribution of the relative change in net primary production. In the blue areas production decreases, and in the yellow/red areas it increases (more than a factor of 2, by 2100 under the RCP8.5 emission pathway).

The contour lines separate the areas of increase from those decreasing.

The increase is mostly situated in the colder regions, while the decrease is primarily in the areas that are already warmer i.e. the Gulf Stream from the Gulf of Mexico to Europe is clearly visible.

The Future

The year 2016 made history, with a record global temperature, exceptionally low sea ice, and unabated sea level rise and ocean heat, according to the World Meteorological Organization (WMO). Extreme weather and climate conditions have continued into 2017.

During its 45th Session (Guadalajara, Mexico, 28 - 31 March 2017), the Panel approved the outline of the Special Report on Oceans and Cryosphere in a Changing Climate (SROCC) to be finalized in September 2019.

Following topics will be articulated in that report:

  • Changes in key physical and biogeochemical properties and processes, including the deep ocean and relevant ocean regions, modes of variability, teleconnections and their feedbacks on the climate system.
  • Specific and combined effects of changes in climate related variables (e.g., warming, acidification, oxygen loss, dust inputs) on e.g., productivity, species distribution and exclusion, habitat compression, food webs.
  • Impacts of ecosystem changes on key ecosystem services (e.g., carbon uptake, biodiversity, coastal protection, fisheries, food security and tourism)


Bopp, L., Resplandy, L., Orr, J. C., Doney, S. C., Dunne, J. P., Gehlen, M., & Tjiputra, J. (2013). Multiple stressors of ocean ecosystems in the 21st century: projections with CMIP5 models. Biogeosciences, 10, 6225-6245.

Doney, S. C., Ruckelshaus, M., Duffy, J. E., Barry, J. P., Chan, F., English, C. A., & Polovina, J. (2011). Climate change impacts on marine ecosystems.

Gruber, N. (2011). Warming up, turning sour, losing breath: ocean biogeochemistry under global change. Philosophical Transactions of the Royal Society of London A: Mathematical, Physical and Engineering Sciences, 369(1943), 1980-1996.

Henson, S. A., Beaulieu, C., Ilyina, T., John, J. G., Long, M., Séférian, R., & Sarmiento, J. L. (2017). Rapid emergence of climate change in environmental drivers of marine ecosystems. Nature Communications.

Schmidtko, S., Stramma, L., Visbeck, M. (2017). Decline in global oceanic oxygen content during the past five decades. Nature 542, 335-339. Doi:10.1038/nature21399

Return to the blog index