Go straight to content
<
<
The unused value of long range forecasts during the drought summer of 2018

What if we had seen it coming?

The unused value of long range forecasts during the drought summer of 2018

COLOURBOX28448590

Photo: Lilian Willemsen/Colourbox

Insight

Published: 26.01.2026
Oppdatert: 27.01.2026

Katrine Jaklin

Norway’s agricultural sector is becoming increasingly vulnerable to weather and climate risks, and the summer of 2018 stands out as an especially difficult year. With extreme temperatures, record low precipitation, and an overall crop failure of around 40 percent, the crisis had major consequences for both farmers and authorities. But what if farmers and key stakeholders in agriculture had had access to more accurate and earlier forecast information?

In a new study, researchers Manuel Hempel and Erik Kolstad have examined the retrospective potential of long‑range forecasts (extended‑range forecasts) for managing the 2018 drought. Long‑range forecasts bridge the critical gap between standard 10‑day weather forecasts and seasonal forecasts that look months ahead. By looking several weeks into the future, these forecasts can give agricultural decision‑makers more time to prepare for extreme events.

Forecasts that hit the mark

The study shows that modern meteorological models actually detected the drought signal in 2018 several weeks in advance. A forecast issued on 18 June 2018 already showed a strong probability of a shift toward warmer and drier conditions in late June and early July. Updated forecasts a week later significantly strengthened this signal, with 90–100 percent probability of unusually high temperatures in southern Norway.

Manuel Hempel fig2
These probabilistic forecast maps, showing temperature (top row) and precipitation (bottom row) for the period starting on 18 June 2018, display forecasts for lead times from one week (leftmost column) to four weeks (rightmost column). In the temperature maps, red indicates probabilities of above-normal (above median) conditions, whereas blue represents probabilities of below-normal (below median) conditions. For the precipitation maps, blue signifies probabilities of above-normal precipitation, and red denotes probabilities of below-normal precipitation. White areas indicate probabilities between 40% and 60%, where the forecast uncertainty is high or near-average conditions are expected.

This means the information technically existed, but it never reached those who needed it. In 2018, there was no active dissemination of such long‑range forecasts to the Norwegian agricultural sector.

Barriers to action: From forecast to decision

For a forecast to have value, it must lead to actions that reduce vulnerability. The researchers used James Hansen’s theoretical framework to evaluate five conditions that need to be in place. In 2018, only two were fulfilled:

  • high user demand
  • and that the forecasts were relevant and accurate


What was missing was:

  • Clear communication and dissemination of the forecasts
  • Support for interpretation and training
  • Real capacity for farmers to act — such as irrigation systems, flexible subsidy schemes, and clear regulations
  • Institutional anchoring ensuring that long‑range forecasts are systematically included in planning and preparedness


Both farmers, advisors, and agricultural stakeholders later described that they would have greatly benefited from forecasts extending beyond ten days. Weather conditions are crucial for everything from fertilization and sowing to harvest strategies and resource use, and uncertainty makes planning difficult. Nevertheless, extended‑range forecasts were neither known nor used in agriculture in 2018.

Those who later received training through the study’s user panel interpreted the forecast signals more accurately and understood earlier how serious the situation would become. Experience turned out to be essential for making use of the forecasts.

The gap between knowledge and action

One of the most striking findings in the study is that knowledge of the drought alone would not have been enough for individual farmers. For grain farmers, the biggest limitation was the lack of irrigation systems. In 2018, fewer than 10 percent of grain producers had access to such technology. Without physical infrastructure to counteract water shortages, even the most accurate forecast becomes “theoretically useful but practically unusable.”

Still, the potential at the organizational level was significant. If agricultural organizations and authorities had seen the forecasts in mid‑June, they could have:

  • Mobilized crisis groups earlier to coordinate measures
  • Clarified subsidy regulations faster, enabling farmers to harvest grain as fodder before its quality was damaged by the drought
  • Secured imports of fodder from abroad earlier, before market pressure increased


The road ahead: A proactive model

The study concludes that Norway needs a more proactive approach to climate risk. This requires not only better models but also strengthening of the “physical and institutional infrastructure.”

The researchers point to the Norwegian Agricultural Advisory Service (NLR) as an ideal bridge‑builder. NLR has contact with over 24,000 members and can act as a skilled interpreter, translating complex meteorological data into practical, farm‑level advice.

To truly benefit from long‑range forecasts in the future, we must shift from being reactive — spending billions on compensation after damage has already occurred — to being proactive through investments in irrigation systems and flexible regulations. Forecasting is a powerful tool, but it requires the capacity to act when the alarm goes off.

Contact

Manuel Hempel
Manuel Hempel

Forsker III