Some fruits can be stored, but stone fruits like cherries and plums cannot; they must be sold when they are ripe. To predict when they will be ready for the market, the Norwegian Computing Center (Norsk Regnesentral) and Gartnerhallen have created a model that calculates when the season will start.
Combines Packing Data with Weather Forecasts
Gartnerhallen is Norway’s largest supplier of fruit, berries, vegetables, and potatoes. They use forecasts to estimate when stone fruits will be ready for the market. These forecasts are important for pricing and calculating how much stone fruit needs to be imported from abroad. The goal is to prevent over-importing fruit that competes with Norwegian fruit. If you import from another country while the fruit in Norway is ripe, the supply will be greater than the demand, and there is a risk of having to discard unsold fruit or sell it at a low price.
Currently, Gartnerhallen creates forecasts based on historical data from packing facilities and dialogue with producers during the season. In collaboration with the Norwegian Computing Center (NR), they are now developing a model that combines historical data with weather forecasts and weather reports.
Researcher Nora Røhnebæk Aasen at NR explains how researchers use data from the Norwegian Meteorological Institute to train the model. Additionally, the new 21-day forecast from Yr is being used.
– To be able to make forecasts early in the spring, we run various simulations of how the weather might develop during the season. This year, we started running the model in March. Initially, we run it with a frequency of 2-3 weeks, while in May/June, we do it weekly. We don’t yet know how accurate this year’s forecasts will be, as the harvest is still ongoing. One thing we are very pleased with is that we managed to detect that it would be an early season already at the beginning of May, so it seems like we got that right, she says.
Norwegian Fruit
Norway has a shorter fruit season than many other countries. Being able to calculate when stone fruits are ripe and ready for the market to regulate imports can provide more profitability for producers and a more sustainable production.
NR and Gartnerhallen have regular meetings to further develop the model. The researchers have also visited Blaaflat Farm in Lærdal, run by Harald Blaaflat Mundal, who has been an active contributor to the project. On the farm, they gained insight into how production is carried out and what considerations are made by the producer.
– By transferring insights from weather forecasts into a model, you can more systematically use the forecasts in production planning, and it becomes a good supplement to the dialogue with the producers, Røhnebæk Aasen explains.
