Experimental statistics: Crop yield forecast based on satellite images

As experimental statistics, the Natural Resources Institute Finland presents forecasts of 11 arable crops by region for 2026 based on satellite images: autumn wheat, spring wheat, rye, fodder barley, malting barley, oats, rapeseed, rapeseed, potato, pea and broad bean. Compared to the averages for the whole country, everyone's hectare yields differ from the average crop level of -6 to +22%. Estimates are based on satellite images taken between 1.5. -17.8.2026. 

The yield forecast for spring wheat is 3694 kg/ha and is 8% above average.

The autumn wheat yield forecast is 4332 kg/ha and is 4% above average.

The yield forecast for rye is 3286 kg/ha and it is 6% lower than average.

The yield forecast for feed barley is 3700 kg/ha and is 4% higher than average.

The yield forecast for malting barley is 4341 kg/ha and it is 22% higher than average.

The yield forecast for oats is 3423 kg/ha, i.e. it is average.

The yield forecast for turnip rape is 1315 kg/ha and is 8% above average.

The yield forecast for rape is 1646kg/ha and is 2% above average.

The yield forecast for potato is 27043 kg/ha, i.e. 6% lower than average.

The pea yield forecast is 2687 kg/ha and is 6% higher than average.

The yield forecast for broad bean is 2008 kg/ha and is 7% higher than average.

The hectare harvest forecasts of the experimental statistics have been calculated by region. A more detailed description of the methodology is given in the Methodological Statement. 

Data based on satellite data from experimental statistics are now presented for the third time in connection with crop statistics. The yield estimate of the yield statistics was published on 21 August, when the following crop estimates were published based on the yield estimates given by ProAgria experts and the crop areas reported by the farmers in connection with the grant application: autumn wheat, spring wheat, rye, barley, oats, triticale, mixed grain, turnip rape, rape, sugar beet, pea, bull bean, dry grass and silage. The next crop estimate for crop statistics is 25.9.

Subscribe to the statistic publication here.

What is an experimental statistic?

Experimental statistics depict new areas of phenomena, utilise new kind of data or are based on new calculation methods. Experimental statistics can include, for example, completely new statistics, tests of the impacts of method changes of existing statistics or achievements of international development work.

Experimental statistics have no official quality or method description or an established publication schedule. This sets them apart from the official Finnish statistics.

Luke publishes regional yield estimates that have been calculated using satellite images and machine learning methods. The results are based on work carried out in a development project funded by Eurostat.

Experimental statistics are still under development. We encourage the people utilising the data to give their feedback on experimental statistics for their further development.