FEATURE ARTICLE - Cloudy with passing schools of tuna–why don’t we have forecasts of fish?

Our latest feature looks at fish forecasting. Is it fanciful and far-fetched?
Published: 18 February 2019

Mark R. Payne, Technical University of Denmark (DTU Aqua) and Chair of ICES Working Group on Seasonal-to-Decadal Prediction of Marine Ecosystems (WGS2D) 

*Originally published in Polar Prediction Matters on 1 February 2019. 

“And now the long-range forecast to midnight Monday, 26th July 2031. The Situation. Positive sea surface anomalies to the North are driving an intrusion of warm water onto the shelf, bringing with it Lusitanian[i] species and saline conditions. For Dogger, Forties and Fisher, a chance of sprat today. Passing schools of tuna at times on Friday, dense in some places. Visibility clearing up over the weekend, with good chances for herring and mackerel. Cod poor. Salinity 35.3 increasing to 4/5. For Foveaux, Chalmers and Rangitata…"

Weather forecasts have become such an integral part of our lives that we instantly recognize their language and structure, even when, as in this case, the nouns are replaced by something (seemingly) nonsensical. But is it that far-fetched? And have you ever considered why we don't have forecasts of fish?

There can be little doubt about the need for such forecasts. Most people have a sense that, much like farmers, fishers are highly dependent of the whims of nature: in some years, mother Earth provides a feast, in other years a famine, with seemingly little connection between the two. In some societies and communities, particularly in tropical countries, small island developing states (SIDS), and Arctic communities, the ocean provides a critical source of not just protein, but also employment, foreign revenue, micronutrients and cultural nourishment. If it were possible to foresee the fluctuations in these resources, these communities might be able to both plan for and adapt to the changing and variable nature of our planet.

Ocean Forecast on Longer Timescales

Fortunately, the last decade has seen a revolution in our ability to predict the state of the climate, and the ocean in particular. Compared to the atmosphere, changes in the ocean take place much more slowly and predictably (due to the high density of water), giving rise to the ability to make forecasts on much longer timescales. Whereas forecasts of conditions on land are measured in days, weeks and months into the future, in the ocean we can start to think about years: indeed, multi-annual-to-decadal forecasts of sea-surface temperature have already been demonstrated in the northern North Atlantic.

Nevertheless, there are many challenges that remain before regular fish forecasts become a reality. New observations, better models and the regular production of forecasts are all required: the Year of Polar Prediction has an important role to play in advancing each of these elements. Similarly, within the Horizon 2020 project Blue Action, we are developing improved forecast systems based on an improved understanding of the links between the Arctic and lower-latitudes.

Marine Ecological Forecasts

However, the critical bottlenecks to developing marine ecological forecasts currently lie further downstream. Marine biologists have yet to achieve the degree of mechanistic understanding and predictive capacity that their climate scientist colleagues are familiar with. Many complex and difficult-to-measure processes play into the dynamics of a species, including, for example, interactions between individuals of both the same species (e.g. competition, cannibalism) or between different species (e.g. predation). As a consequence, a predictive understanding of the relationship between the physical and biological environment remains difficult to achieve.

Nevertheless, by working with what knowledge is available, researchers are starting to produce the first generation of forecast products of fish (Figure 1). The first of these products appeared a decade ago in Australia, forecasting the distribution of southern bluefin tuna. More recently, researchers in the United States have developed multiple products along the Pacific coast. Europe, however, lags behind these countries: it is only in the last year that first products have emerged, predicting the spatial distribution of blue whiting (Figure 2) and bluefin tuna (Figure 3) in the North Atlantic. To the best of our knowledge, there are no such forecast products being produced regularly that forecast living marine resources in the developing world, where ocean-dependency is highest and there is the greatest need for it.

A Good Forecast is Used to Make Decisions

How can we improve this? Developing a successful forecast product is not just a question of forecast skill: indeed, there is a solid body of work that shows that the relationship between the skill and value of a forecast to the end-user is neither strong nor linear. Instead, developing a forecast requires striking the balance between what is scientifically feasible, on the one hand, and what is useful to end-users, on the other. The co-development approach, which we are following in the Blue Action project, is a particularly valuable way to facilitate this by bringing scientists and end-users together to specify, develop and produce the products that are needed. Scientists have an important role to play in tailoring and translating the available climate data into information, but ultimately the design process is a collaborative one.

In a well-known article from the 1990s, the meteorologist Allan Murphy argued that a good forecast is one that is used to make decisions. The products that have been developed thus far clearly illustrate that marine ecological forecasting is possible (e.g. Figure 2). We have yet to reach the point where marine ecological forecasts are regularly used in decision-making, but it is also clear that we have many of the pieces required in hand. Involving those that make these important decisions in the development process can help to strike the balance between what is feasible and what is useful. Hopefully future generations will come to see a fish-forecast as something a little less fanciful.

“… And now, the news. A consortium of researchers and fishers are predicting unusually early migrations of Hornfisk this year in response to an anomalously warm winter. Welcoming this advanced warning, Jens Jensen from the Danish Fishing Association said…"

[i] Species adapted to warmer, southern European waters e.g. Anchovy, Sardine


The work of WGS2D contributes to two main science priorities in ICES Science Plan:

Ecosystem science with its focus on developing methods to map and predict the distribution of seabed and pelagic habitats and biodiversity and their sensitivity to environmental variation and change, and

Emerging techniques and technologies with their emphasis on data gathering, processing, and interpretation that have potential to progress the ICES vision and mission. 

Print this pagePrint it Request newsletterSend to Post to Facebook Post to Twitter Post to LinkedIn Share it

Why don't we have forecasts of fish? Passing school of Northern Anchovy at the Monterey Bay Aquarium, California. Photo: Kirstin Werner.

 Figure 1. Map of known examples of operational marine ecological nowcast and forecast products of living marine resources. The first such products were developed in Australia and have been in operation for over a decade, while products are rapidly coming online in North America. Note the absence of products in Europe. (Payne et al. 2017 Front. Mar. Sci.)

Figure 2. Figure 2. Example of an ecological forecast product, showing the forecast spawning distribution for blue whiting (Micromesistius poutassou) in March 2018. Distribution is represented here as the probability of observing blue whiting larvae and is plotted as a) the value and b) the anomaly relative to the climatological probability (1960-2010).

 Figure 3. Figure 3. Bluefin tuna feeding habitat forecast in August 2019. For each pixel, the probability of the sea surface temperatures being thermally suitable for Atlantic bluefin tuna is estimated from the North American Multi-Model Ensemble (NMME). The contour line corresponding to a probability of 0.5, which has been used here to define the most likely limit of suitable habitat, is plotted as a heavy blue line, while the historical minimum and maximum habitat boundaries (1960-2015) are plotted as red lines.

​Author Mark R. Payne's research examines the impacts of climate change and climate variability on life in the ocean. His work is pioneering the development of Climate Services for monitoring and managing life in the ocean in Europe and involves coupling biological knowledge to climate models to produce predictions that are of direct relevance to end-users. Payne is also the leader of the Climate Services work package within the EU H2020 project “Blue Action”. The latest version of the forecasts that he has developed can be found on the website



Murphy, A. H. 1993. What Is a Good Forecast? An Essay on the Nature of Goodness in Weather Forecasting. Weather and Forecasting, 8(2), 281–293.<0281:WIAGFA>2.0.CO;2

Payne, M. R., Hobday, A. J., MacKenzie, B. R., Tommasi, D., Dempsey, D. P., Fässler, S. M. M., Haynie, A. C., Rubao, J., Gang, L., Lynch, P. D., Mateil, D., Miesner, A. K., Mills, K. E. ,Strand, K. O., Villarino, E. 2017. Lessons from the First Generation of Marine Ecological Forecast Products. Frontiers in Marine Science, 4.

c FollowFollow Focus on ContentFocus on Content
HelpGive Feedback

FEATURE ARTICLE - Cloudy with passing schools of tuna–why don’t we have forecasts of fish?

International Council for the Exploration of the Sea (ICES) · Conseil International pour l'Exploration de la Mer (CIEM)
ICES Secretariat · H. C. Andersens Boulevard 44-46, DK 1553 Copenhagen V, Denmark · Tel: +45 3338 6700 · Fax: +45 3393 4215 ·
Disclaimer Privacy policy · © ICES - All Rights Reserved