Periodic Reporting for period 1 - NeTNPPAO (Near-term predictability of net primary production in the Atlantic Ocean)
Berichtszeitraum: 2017-01-09 bis 2019-05-08
With new predictions of future world population reaching 9.6 billion by 2050 and the related spectre of a global food crisis, the necessity to improve our ability to manage world's fisheries has never been more pressing. A fundamental step in this direction and the main goal of NeTNPPAO is the improvement of near-term (i.e. seasonal to decadal) predictions of Net Primary Production (NPP) in the Atlantic Ocean. NPP is the rate of production of phytoplankton biomass, the primary source of food for marine animal life and there is close relationship between fish biomass and NPP in the open ocean. According to “The State of World Fisheries and Aquaculture-2012”3 (Food and Agriculture Organization - FAO), in 2010 the Eastern Tropical Atlantic sustained two of the top three fishery industries of the African continent (Morocco and Senegal) with total catches of ~4 million tons while the Northeast and Northwest Atlantic areas together summed ~11 million tons of landed fish with important economic impacts on both Europe and North America. However, all 6 Atlantic fishing areas identified in the report have shown decreasing trends or considerable fluctuations in catches during the past decades resulting from either environmental variability, fishing pressure or a combination of both. Many studies point to a tight connection between fluctuations in fish populations and both natural climate variability and anthropogenic climate change. However, in the past fish stocks management has often relied on the assumption that marine ecosystems were in long-term equilibrium, with fishing pressure being the dominant factor controlling fish populations. This resulted in management strategies based on extrapolations from historical time-series of fish catches, leading in some cases to overestimated sustainable harvest rates that contributed to the decline of fish stocks. Therefore, the improvement of ecosystem-based strategies is becoming a priority for near-future fisheries management.
Over the past few years near-term climate predictions have emerged as rapidly improving tools at the service of society and decision-makers. The 5th Coupled Model Intercomparison Project (CMIP5) that contributed to the 5th Assessment Report (AR5) from the Intergovernmental Expert Panel on Climate Change (IPCC), included a set of near-term climate predictions that proved skillful at regional scales. Near-term climate predictions can be performed using state-of-the-art Earth System Models (ESMs). ESMs are complex tools that attempt to represent all the processes that are of relevance for climate. They are formed by several coupled modules representing the physical state of the atmosphere, ocean and land surface. Furthermore, ESMs usually include a representation of ocean biogeochemical cycles. EC-Earth is the model of choice to develop the activities planned in NeTNPPAO.
The ocean component of EC-Earth was used to produce initial conditions for both the physical and the biogeochemical component to be used in retrospective near-term predictions (i.e. predictions of the past) of the biogeochemical fields of the Atlantic Ocean. These initial conditions were generated by assimilating physical fields (temperature and salinity) from observation-based products over the period 1958-2015. Several simulations were tested where the degree of the constraint and its extension were varied to test the response of the biogeochemical model. Overall, one key result that emerged was that the extension (rather than the strength) of the data-assimilation had a strong control over the response of the biogeochemical model. This is due to the onset of strong spurious vertical velocities at low latitudes that bring large amounts of nutrients to the surface resulting in anomalously high surface chlorophyll when compared to satellite observations.
Moreover, an extensive analysis of the relationship between nutrients and physical properties of sea water in key regions was performed. The main result of this analysis is that in the Southern Ocean there are very strong correlations between density and nutrient concentration in the density space encompassing Subantarctic Mode Water and Antarctic Intermediate Water, two key water masses for the redistribution of nutrients to the low latitude upwelling systems. These correlations can be exploited to partially reconstruct the interannual variability of nutrient distribution (currently missing due to lack of observations). Such time-varying distribution of nutrients can then be used to constraint the solution of the biogeochemical model and thus provide more reliable initial conditions for near-term predictions of ocean biogeochemistry.
The ocean component of EC-Earth was used to produce initial conditions for both the physical and the biogeochemical component to be used in retrospective near-term predictions (i.e. predictions of the past) of the biogeochemical fields of the Atlantic Ocean. These initial conditions were generated by assimilating physical fields (temperature and salinity) from observation-based products over the period 1958-2015. Several simulations were tested where the degree of the constraint and its extension were varied to test the response of the biogeochemical model. Overall, one key result that emerged was that the extension (rather than the strength) of the data-assimilation had a strong control over the response of the biogeochemical model. This is due to the onset of strong spurious vertical velocities at low latitudes that bring large amounts of nutrients to the surface resulting in anomalously high surface chlorophyll when compared to satellite observations.
Moreover, an extensive analysis of the relationship between nutrients and physical properties of sea water in key regions was performed. The main result of this analysis is that in the Southern Ocean there are very strong correlations between density and nutrient concentration in the density space encompassing Subantarctic Mode Water and Antarctic Intermediate Water, two key water masses for the redistribution of nutrients to the low latitude upwelling systems. These correlations can be exploited to partially reconstruct the interannual variability of nutrient distribution (currently missing due to lack of observations). Such time-varying distribution of nutrients can then be used to constraint the solution of the biogeochemical model and thus provide more reliable initial conditions for near-term predictions of ocean biogeochemistry.
Producing near-term ocean biogeochemistry predictions is an emerging and promising idea but it has been applied only to two models so far. The research executed within NeTNPPAO will allow us to move beyond these first attempts by using different and more sophisticated initialization techniques for the physical state of the ocean and, in particular by investigating the improvement in predictability skill derived from the concurrent initialization of reconstructed 3D nutrient fields.
The added value of initializing nutrient fields is still to be tested but the robustness of the correlations found between density and nutrient distributions in key regions of the Southern Ocean are encouraging. Two set of retrospective near-term predictions will be executed within one year from the end of NeTNPPAO: one initialized from a reconstruction where only physical fields have been assimilated and one initialized from a reconstruction where both physical fields and nutrients have been assimilated. This has never been attempted before and it requires a reconstruction of time-varying 3D nutrient fields because only climatological fields are available at global scale. This reconstruction is based on the exploitation of the correlations found between density (for which there are interannualy variable records) and nutrients.
Results from this project will represent a fundamental step towards the development of the new generation of fishery management tools. The possibility to predict near-term variability in NPP will open new possibilities in the development of fish population models which will need to be able to resolve population dynamics on these shorter timescales. Furthermore, this project will contribute substantially to push an emerging topic such as the near-term prediction of ocean biogeochemical variables at the forefront of the international agenda of the climate modelling community. Existing long-term projections, although useful to understand the impact of climate change on fish population dynamics, do not place on policy makers the necessary pressure to act because of a diffused short-term bias in risk perception intrinsic in politics (and in the human nature, in general). Providing reliable near-term predictions on fish population dynamics – and this project is a necessary step in that direction – will create the needed pressure for policy makers to act, while giving them, at the same time, the proper tools to take informed decisions. Furthermore, the evidence of the ability of science to solve immediate and concrete problems will stimulate the increase in the allocation of funding to develop further the scientific knowledge and the modelling capability of both near-term climate predictions and fish population dynamics.
The added value of initializing nutrient fields is still to be tested but the robustness of the correlations found between density and nutrient distributions in key regions of the Southern Ocean are encouraging. Two set of retrospective near-term predictions will be executed within one year from the end of NeTNPPAO: one initialized from a reconstruction where only physical fields have been assimilated and one initialized from a reconstruction where both physical fields and nutrients have been assimilated. This has never been attempted before and it requires a reconstruction of time-varying 3D nutrient fields because only climatological fields are available at global scale. This reconstruction is based on the exploitation of the correlations found between density (for which there are interannualy variable records) and nutrients.
Results from this project will represent a fundamental step towards the development of the new generation of fishery management tools. The possibility to predict near-term variability in NPP will open new possibilities in the development of fish population models which will need to be able to resolve population dynamics on these shorter timescales. Furthermore, this project will contribute substantially to push an emerging topic such as the near-term prediction of ocean biogeochemical variables at the forefront of the international agenda of the climate modelling community. Existing long-term projections, although useful to understand the impact of climate change on fish population dynamics, do not place on policy makers the necessary pressure to act because of a diffused short-term bias in risk perception intrinsic in politics (and in the human nature, in general). Providing reliable near-term predictions on fish population dynamics – and this project is a necessary step in that direction – will create the needed pressure for policy makers to act, while giving them, at the same time, the proper tools to take informed decisions. Furthermore, the evidence of the ability of science to solve immediate and concrete problems will stimulate the increase in the allocation of funding to develop further the scientific knowledge and the modelling capability of both near-term climate predictions and fish population dynamics.