- Since the 1960s, the global oceanic oxygen content has declined by more than 2%
- Sustained observations in the North Sea reveal the recent onset of oxygen deficiency in late summer, partly due to ocean warming. The intensity and extent of oxygen deficiency has also increased over time.
- Short-term measurements for the Celtic Sea also indicate the onset of oxygen deficiency in late summer.
Low evidence, medium consensus
On a global scale, there is a high level of confidence that the oceans are losing oxygen due to ocean warming. In UK coastal waters, there is a high level of understanding of the seasonal and interannual variability in oxygen dynamics in the permanently mixed and seasonally stratifying waters in the North Sea due to the relatively extensive sampling regime for dissolved oxygen concentrations in this region over the past four to five decades. Repeat sampling at specific sites has provided insight into the occurrence and onset of oxygen deficiency but the spatial extent of oxygen deficiency outside of these specific regions within the North Sea is uncertain due to the paucity of direct observations. Approximately one third of the historical seasonally focused depletion of oxygen in the North Sea has been attributed to warming but the remaining two thirds are thought to be due to enhanced oxygen consumption. The relative importance of the processes that drive enhanced oxygen consumption (e.g. more organic matter, decreased ventilation of bottom waters), however, remain poorly understood. Finally, whereas the North Sea is well sampled in time and space, the rest of the Northwest European Shelf waters, especially the Celtic and Irish seas, are relatively poorly sampled and offer low levels of confidence in the occurrence or risk of oxygen depletion. Nevertheless, both observations and models agree that the UK coastal and shelf seas are losing oxygen and thus there is a medium level of confidence on the direction of change.
- Annual mean dissolved oxygen concentration in the global ocean is projected to decline by 1.5 to 4% by 2090 for all RCPs. This decline will be most acute below the thermocline.
- For UK shelf waters, models project that annual mean oxygen concentration will decline most strongly in North Sea regions and the Celtic Sea (5.6 to 5.9% by 2100, RCP 8.5). Deeper regions exposed to exchange with the open ocean (the Irish shelf and Shetland shelf) are expected to be less affected, decreasing by 2.9 to 3.1%.
- The predicted increase in temperature over this century for UK shelf seas will lead to a decrease in dissolved oxygen through the whole water column because of reduced solubility.
- The risk of oxygen deficiency in summer will increase because of lower oxygen levels experienced during the preceding winter and spring.
- Increased rainfall and runoff would increase the risk of eutrophication and cause oxygen concentrations to locally decrease.
- Continued warming and reduced oxygen availability will affect the metabolism, health, and reproduction of many marine organisms, which could have major impacts on ecosystems and commercial fisheries.
Medium evidence, medium consensus
At the global scale, there is a high level of confidence that an increase in temperature will continue to reduce the solubility of oxygen and enhance stratification and thus lead to the ongoing decline in dissolved oxygen concentrations, especially below the thermocline. On a regional scale appropriate for coastal and shelf seas, there is a general consensus that the ocean will lose oxygen. Model simulations can provide estimates of the magnitude and causes of the decline in dissolved oxygen but there is still uncertainty in how well they represent the coupling between physical and biogeochemical processes, biological processes specifically, strong seasonality in nutrient supply in a shallow water column and interaction with the sediment. Therefore, there is a medium level of confidence on the future of dissolved oxygen dynamics on a regional scale relevant to UK marine waters.
- We need to be able to determine the mechanisms driving spatial and temporal trends in dissolved oxygen and confidently identify when and where changes in dissolved oxygen are being driven by human-induced activity such as ocean warming or nutrient enrichment relative to background natural variability.
- Assessing the occurrence, frequency and spatial extent of oxygen deficiency in UK coastal and shelf waters is hampered by the lack of long-term data in regions outside of the North Sea. The poor resolution of dissolved oxygen data also hampers the ability to confidently test coastal and shelf-sea models.
- There is still uncertainty surrounding the ability of models to simulate the individual processes and coupling between processes that control dissolved oxygen dynamics. To accurately predict dissolved oxygen, models need to simulate each contributing process correctly, in isolation but also coupled to other processes. The lack of understanding is particularly acute within coastal and shelf sea sediments.