Pollen you can read, thanks to Finland
Our pollen forecast runs on SILAM, the atmospheric model of the Finnish Meteorological Institute. Seven species, every hour, for the whole of Europe and well past the Caucasus. Included in the Starter plan at no cost.
The model behind the forecast
SILAM is built and run by people who have spent careers on atmospheric composition. We did not build a pollen model. We asked whether we could use theirs, and they said yes.
Finnish Meteorological Institute, Atmospheric Composition unit
SILAM stands for System for Integrated modeLling of Atmospheric coMposition. It forecasts what is actually in the air: pollen, dust, smoke and pollutants. The pollen part is the reason it matters here. It carries species that most other models leave out, and it publishes them openly for anyone to use.
In September 2026 the Atmospheric Composition unit gave us permission to use their data on two conditions. Name the source, and identify our traffic so they can warn us before anything about the service changes. Both are met on every single request we make. That second condition is not a formality: it is how a small platform like ours finds out that something moved, before a chart quietly goes blank.
Read more at silam.fmi.fi and the Finnish Meteorological Institute. If you work with air quality or pollen yourself, their data is worth your time.
Attribution: hourly pollen data by SILAM, Finnish Meteorological Institute. Shown on every chart that uses it.
Asking politely, storing carefully
Open data only stays open if the people using it are considerate. Here is what we do so that a growing number of growers never becomes a burden on a public weather service.
One question per area
The model works in cells of about 11 km. Two fields in the same cell get the same answer, so we ask for the cell and not for each field. Asking finer would cost the institute more traffic and tell you nothing new.
Six hours between questions
SILAM produces one run a day. Asking every ten minutes would return the same numbers, so we do not. A new run is on your screen within a quarter of a day, and the server is left alone in between.
We remember, so you see further
The model holds three days back. Your chart shows a full week. Every answer we receive is stored, so the history grows on its own without a single extra request.
Your clock, not theirs
The model thinks in UTC. Your chart runs on the local time of your site, so pollen hours line up with your own sensor readings instead of sitting an hour or two beside them.
One at a time
However many growers open their dashboard at once, never more than one of our requests is at their door. The rest are served from what we already have.
Honest when something is off
If their servers are down, you keep the last series we received and the chart says so. It will not tell you there is no pollen at your location when the truth is that we could not reach the model.
Three ways to read the same hours
Pollen species differ by more than a hundredfold on a quiet day. One scale cannot serve them all, so you choose the one that answers your question.
Grains per cubic metre
The plain numbers, hour by hour, on the same scale as the other pollen data in your dashboard. Best when you follow one or two species of similar size.
Small and large together
Every step on the axis is ten times the one below. A species at 0.2 grains stays readable next to one at 20, instead of lying flat against the axis.
How bad is it, really
SILAM converts each species with its own threshold into one index from 1 to 5, and names the species behind the number. Twenty grains of ragweed and twenty of birch do not mean the same thing, and this view knows that.
Set the place once
Your locations live in a single list. Add a field, a greenhouse or a second site once, and every window on your dashboard offers it: pollen, weather, air quality, history and the satellite view.
- Switch location and the whole dashboard follows
- The same place feeds the Sentinel-2 satellite view, with NDVI and SAVI over your own plot
- No second list to maintain, and no chart quietly pointing at the wrong village
Pollen is one of several sources in your dashboard. The five species from CAMS, delivered through Open-Meteo, sit alongside the seven from SILAM, so you can lay two models over each other and see where they agree. Hazel is the one species CAMS does not carry, which is where this whole story started: in the Benelux it is the first to flower, from January to March, and it was missing.
See it on your own field
The Starter plan is free and includes the pollen forecast. Create your account, add your location, and the seven species appear for the place you actually grow.
Create your free account → Our commitment →