An introduction to scientific seasonal forecasts

Seasonal forecasting is becoming an increasingly useful tool to support long term decision making in farming. There is a great potential in utilizing seasonal forecasts for farming practices if it is done properly. This can include supporting decisions about sowing, harvesting and irrigation as well as the choices of planting a specific crop variety.

Long term weather predictions have been documented as early as 1793[1], although in rather different form than the seasonal forecasts we are speaking of in this blog. Then, using common weather patterns and indicators like the thickness of an onion skin while we are now looking at scientific seasonal forecasts using climate models, it is safe to say that there have been some developments. The WaterApps project aims to combine ancient and local forecasting knowledge with these scientific forecasts in Bangladesh and Ghana. This blog gives a quick insight into scientific seasonal forecasts, their limits and their potential illustrated by two experts of Wageningen University and Research.

Current scientific seasonal forecasts are based on climate models that are tested on historical data and then used to make a projection for the coming season. This is often presented as a deviation from the long term mean from historical data. For example a 40% chance of higher than average precipitation or a 20% change of lower than average temperatures. There is a variety of seasonal forecasts, for climatic indicators like precipitation, temperature, radiation (important for photosynthesis and evapotranspiration) but also hydrological indicators like river runoff and soil moisture. You can see how this kind of information can be useful for a farmer who is deciding on a long term farming strategy.

Dr. Ronald Hutjes is an associate professor at Wageningen University and Research who specifically works with seasonal forecasts for agriculture practices. “It would be great if we could combine seasonal forecasts with advice on sowing time and crop variety. In a good season a farmer can invest in a high yield crop and in a bad season in a more robust crop variety”. This advice would be dependent on more than just the seasonal forecasts, the availability of seeds for example. On this Researcher Dr. Wouter Gruell (also of Wageningen University and Research) agrees.

Dr. Gruell is working on hydrological forecasts, he explains that “A very important aspect of seasonal forecasts is the skill, how sure we are the projection makes sense or the trust we have in the data”. The trustworthiness of a forecast is dependent on the location and the season, the predictability of the weather and our knowledge of for example soil moisture and snow cover. “The skill of the European seasonal forecast is often better in Scandinavia than for example The Netherlands”  says Dr. Gruell. In Europe the weather is very chaotic and unpredictable, especially in comparison to regions affected by large weather patterns like El-Niño and the monsoon. These large weather patterns contribute to a higher skill of the seasonal forecasts in for example South-East Asia. However, there will always be a level of uncertainty in the projections.

Dr. Hutjes sees dealing with this uncertainty as one of the main challenges in using seasonal forecast data. “For a big multinational or investor it is easier to invest based on the forecasts, they can invest five times and if then one time it does not work out the loss is manageable. For them, there will still be a net benefit in the long run.” For a single farmer this can be different, they do not have the big buffers of the multinationals. When a single farmer decides to invest in a season which is projected to have high precipitation and it turns out to be a dry season, there is a higher risk of losing a substantial part of their assets. “There is a great potential in combining seasonal forecasts with micro-credits or insurance policies to deal with this risk and uncertainty”.

Current seasonal forecasts and knowledge of the skill of these data can already provide useful input in farming decision making. And there are promising developments in store for us.

Some examples of organizations publishing global scientific seasonal forecast maps are (among others): IRI Columbia University, ECMWF, Hydrometcenter of Russia, The Deutsche Wetter Dienst New Zealand’s National Institute of Water and Atmospheric Research even has a video with the interpretation of the seasonal outlook for New Zealand. 

Note: When using these data, it is always good to look at the skill maps provided by the organization and be aware that they present a pobability and not a 100% certain outcome.

References:

Klemm, T., & McPherson, R. A. (2017). The development of seasonal climate forecasting for agricultural producers. Agricultural and Forest Meteorology, 232, 384–399. https://doi.org/10.1016/j.agrformet.2016.09.005