The number you see on the risk score does not come from a single sensor or from simply multiplying pollen concentration by some factor. Behind it stands a process in several steps that combines:
- pollen measurements by plant type,
- clinical thresholds by which concentrations are judged as to how "dangerous" each one is,
- weather conditions that amplify or dampen symptoms,
- air quality (European AQI index) which additionally irritates the airways,
- and — optionally — your personal sensitivity to individual allergens.
This article explains each step in detail. If you've ever wondered why 8 grains/m³ of birch is a "low level" for us while 8 grains/m³ of ragweed is "high," the answer lies precisely in the fact that different plants have different thresholds.
1. Pollen measurements by type
The starting point is daily measurements of pollen in the air for Croatian cities, expressed in pollen grains per cubic meter of air (grains/m³). The main source is the Institute for Public Health "Dr. Andrija Štampar". The page updates the measurements several times daily. For cities in Slovenia, which do not have their own measuring station on this service, the starting point is modeled concentrations from the European CAMS model (via Open-Meteo) — a useful reference point, but not a measured value.
For each allergen (birch, grasses, ragweed, olive, …) we display the numerical value (if available) and the level (low, moderate, high, very high).
2. Thresholds by plant type — DWD and EAACI
This is the most important thing to understand: the same 8 grains/m³ does not mean the same for all plants.
Clinical guidelines from the German DWD (Deutscher Wetterdienst), aligned with recommendations from the European Academy of Allergy and Clinical Immunology (EAACI), define separate thresholds for each type. Examples (values in grains/m³, daily average):
| Plant | Low | Moderate | High |
|---|---|---|---|
| Hazel, alder | 1–10 | 11–100 | >100 |
| Birch | 1–10 | 11–50 | >50 |
| Ash | 1–50 | 51–75 | >75 |
| Olive | 1–10 | 11–100 | >100 |
| Grasses | 1–5 | 6–30 | >30 |
| Rye | 1–2 | 3–6 | >6 |
| Ragweed, mugwort | 1–5 | 6–10 | >10 |
The reason: allergenic potency per grain differs from plant to plant. Ragweed is exceptionally aggressive — already 10 grains/m³ in sensitive individuals causes strong symptoms. Birch is "milder" — similar symptoms appear only around 50 grains/m³.
This means that when you see a "low" or "high" label on the site for an individual plant, it uses a threshold specific to that plant, not a universal scale.
3. From numbers to score — saturation curve
Each allergen enters the calculation with a normalized value from 0 to 1, where 1.0 corresponds to the concentration at the "high" level threshold for that plant.
We could scale it linearly: double the pollen = double the score. But that is not what we actually observe in allergy sufferers. Research on symptom severity (EAACI Task Force, EPOCHAL study) shows that:
- Symptoms rise sharply through the low and moderate levels.
- After the high threshold, the score increase slows down — doubling the pollen above "high" does not double the symptom severity. A person who already reacts will react the same way at both 30 and 60 grains/m³ of grass.
Therefore, we use a concave (saturation) curve — a formula that rises steeply at low concentrations but flattens near 1.0:
score = (1 − e^(−2.5 × x)) / (1 − e^(−2.5))
where x is the normalized value (0–1). The result is again a number between 0 and 1.
In practice: for grasses (high threshold = 31 grains/m³)
- 5 grains/m³ → score ~30 (LOW)
- 13 grains/m³ → ~53 (MODERATE) — the middle of the DWD "moderate" band ends up in our "moderate" risk
- 30 grains/m³ → ~74 (HIGH)
4. Multiple allergens in the air simultaneously — "soft-max"
What if both birch and grasses are at moderate level at the same time? We don't take just the highest value (that would "throw away" the contribution of the secondary allergen), nor do we add them in full (that would unfairly penalize pollen-allergic individuals with multiple sensitivities).
We use soft-max:
aggregate = highest_normalized_value + 0.5 × second_highest
capped at 1.0. So:
- one allergen at 0.5 → remains 0.5
- two allergens at 0.5 each → 0.5 + 0.25 = 0.75
- third and further allergens do not enter the calculation (we address them through personal sensitivity — see step 7)
The idea is clinically supported: simultaneous exposure to two allergens worsens symptoms, but not strictly additively.
5. From aggregate to pollen score
The aggregated value (0–1) passes through the saturation curve from step 3 and is multiplied by 75. Why 75 and not 100? To leave room for weather conditions in the next step.
One feature worth knowing: an individual measurement at the "high" DWD threshold level by itself yields a full pollen score (75). This is a design decision — once you are in the DWD "high" band, symptoms are already significant, and weather can push the score into HIGH or VERY HIGH.
6. Weather and air quality — modifier
Weather significantly affects how much pollen remains in the air and how irritating it is to the mucous membrane, and polluted air independently worsens airway reaction. We add (or subtract) points based on:
| Condition | Change |
|---|---|
| Wind > 25 km/h | +15 |
| Wind 15–25 km/h | +8 |
| Wind < 5 km/h | −5 |
| Humidity < 40 % | +10 |
| Humidity > 75 % | −8 |
| Rain > 5 mm/24h | −20 |
| Rain 1–5 mm/24h | −10 |
| Temperature ≥ 30 °C | +8 |
| Temperature ≥ 25 °C | +5 |
| Air quality "moderate" (AQI 41–60) | +3 |
| Air quality "poor" (AQI 61–80) | +6 |
| Air quality "very/exceptionally poor" (AQI > 80) | +10 |
The logic is well-documented in meteorology and pulmonology:
- Wind disperses pollen farther from the source; strong wind = more pollen per m³.
- Dry air keeps pollen suspended; humid air brings it down to the ground.
- Rain washes pollen out of the atmosphere — this is the strongest negative modifier.
- Heat accelerates pollen emission and itself strains the mucous membrane.
- Polluted air (suspended particles PM2.5/PM10, NO₂, ozone) itself irritates the airways and amplifies the allergic response. This is why worse air quality only raises the score — clean air does not lower it below the baseline. We use the European air quality index (EAQI) from the European Environment Agency, obtained from the Open-Meteo Air Quality service.
The total weather and air quality modifier ranges from −33 to +43 (air quality contributes only to the positive side). The final sum is still capped at 0–100 (see step 8).
7. Personal sensitivity (optional)
By clicking on an allergen in the "Personalize forecast" panel you can mark:
- Normal (default) — the allergen enters the calculation normally
- Sensitive — your contribution from that allergen is doubled, reflecting a stronger personal reaction
- Excluded — the allergen does not enter the calculation at all
Settings are remembered in your browser and apply to all cities. No registration or account is needed. If you use multiple devices, you need to set the preferences separately on each.
8. Final score — levels and outdoor activity recommendation
The sum of pollen points and weather and air quality modifiers is capped to the range 0–100 and that number is the risk score. From it we derive the level:
| Points | Level | Outdoor activity recommendation |
|---|---|---|
| < 30 | SAFE | Unlimited |
| 30–49 | LOW | Up to 6 hours |
| 50–69 | MODERATE | 2–3 hours |
| 70–84 | HIGH | At most 1 hour |
| ≥ 85 | VERY HIGH | Avoid outdoor activity |
The outdoor activity recommendation is orientative — based on general guidelines for an average allergy sufferer with rhinitis. A person with asthma or severe allergy should spend less time outdoors at all levels, while a person without allergies can freely be outside even at "very high".
Sources and limitations
- Thresholds by type are taken from the DWD guideline "Pollenflug-Gefahrenindex", aligned with EAACI definition of the pollen season (Pfaar et al., 2018; 2024).
- Symptom saturation curve is supported by dose-response studies, particularly the EPOCHAL study (2024) and earlier work (Krzych-Fałta et al., 2008).
- Weather modifiers reflect standard meteorological relationships documented in Sofiev et al. (SILAM model) and the DWD ICON-ART model.
- Air quality uses the European index (EAQI) from the European Environment Agency (EEA), obtained via the Open-Meteo Air Quality API (CAMS source). The link between air pollution and worsening respiratory/allergic symptoms is documented in numerous epidemiological studies (e.g., WHO Global Air Quality Guidelines, 2021).
What this score is not
Although the numbers behind it are scientifically supported, this is an orientative model:
- It does not account for your personal allergic profile (except for optional sensitivity).
- It does not distinguish severe asthmatics from those with rhinitis only.
- It does not account for medications you take.
- It does not predict lag effects — one extremely high pollen day can cause symptoms 3 days later.
For individual assessment and therapy, consult your family physician or allergist. This page is informative, not medical advice.
Note
A detailed discussion of limitations can be found on the About pelud.net page. We welcome feedback, corrections, and suggestions for algorithm improvements.