Seasonal forecast
Temperature and precipitation trends for the coming months. Explore probabilities, the spread of possible outcomes and differences between models.
Explore sectionsMaps, charts and probabilitiesExplore
Seasonal maps of Portugal
Temperature and precipitation anomalies and probabilities, with interactive navigation through the coming months.
Maps will load as you approach this section.
Click the map to view the value and its source grid cell.
About these maps
The maps use the forecast and hindcast ensemble grid, with a 1993–2016 reference period and equal weight for each model in the multi-model ensemble. Bilinear interpolation smooths the display without increasing meteorological resolution. Over the islands, grid cells of around 1° may include ocean influence; zooming does not provide a forecast for individual locations.
Probabilities refer to terciles of each model’s climatology. The P10–P90 range describes the spread of possible outcomes, rather than a calibrated measure of certainty.
Source: Copernicus C3S / CDS and the forecast centres listed · Coastline: CAOP · Processing and maps: Luso Meteo.
Loading the seasonal forecast…
Regional outlook
Outlook for the coming months
Temperature
Mean temperature anomaly · °C
Line: median. Shaded bands: the 25th–75th and 10th–90th percentiles of the forecast outcomes.
Precipitation
Accumulated precipitation departure from the model climatology · mm
Bars: forecast mean. Ranges: the 10th–90th percentiles of the forecast outcomes.
Probabilities of forecast outcomes
Below, near and above normal correspond to the three thirds of the historical distribution. The climatological reference probability is approximately 33% for each category.
Model comparison
Agreement between models is not the same as the probability of a correct forecast. Historical forecast skill has not yet been assessed by Luso Meteo.
Global seasonal maps
A global view of monthly and three-month trends, using official maps from the C3S multi-model ensemble.
Maps will load as you approach this section.
Official C3S products use their own combination of models and reference climatology. The global ensemble may include more centres than the Luso Meteo regional summary. Available periods are those published by the source.
In the tercile summary, colours highlight enhanced probabilities of below-normal or above-normal values; white does not necessarily mean “normal”. Refer to the original legend, which is preserved. Read precipitation in the unit shown on the map.
Euro-Atlantic regimes · coming weeks
Probabilities of circulation regimes in the ECMWF sub-seasonal ensemble, covering around six weeks. This chart is independent of the monthly seasonal forecast.
Waiting for the official chart.
Positive North Atlantic Oscillation.
Negative North Atlantic Oscillation.
Blocking over the Scandinavian region.
A ridge of high pressure over the Atlantic.
Outcomes without a sufficiently clear regime classification.
The bars show the proportion of outcomes assigned to each regime. A regime alone does not determine rainfall or temperature in Portugal: its effects depend on the season and the specific circulation pattern. The grey category is part of the uncertainty.
The official chart is preserved; percentages are not estimated from the image. © ECMWF · CC BY 4.0 · open original product
How to interpret
What does an anomaly mean?
It is the departure from the model’s own hindcast climatology for the same issue month and lead time. The reference period used is stated in the data. It does not correspond directly to the CHELSA climatological normal.
Probability, spread and confidence
Probabilities describe the distribution of forecast outcomes. The bands show ensemble spread, rather than calibrated confidence intervals. Forecast skill depends on the region, variable, season and lead time.
Scale and limitations
This forecast describes monthly and three-month regional trends. A drier month can still include heavy rainfall; a warmer month can still include cold days. Model resolution does not support detailed local forecasts, particularly over islands and mountains.
Source: Copernicus Climate Change Service / Climate Data Store and the forecast centres listed. Processing and presentation: Luso Meteo.
📖 Glossary
- Anomaly
- The difference between the forecast value and the model’s reference climatology. A positive temperature anomaly indicates warmer average conditions; a positive precipitation anomaly indicates an accumulated total above the reference.
- Model climatology
- The distribution derived from hindcasts for the same issue month and lead time. The Luso Meteo product on this page uses 1993–2016 as its reference period. This is neither the CHELSA climatological normal nor an observational time series.
- Ensemble and members
- A set of possible outcomes produced by a model. Each member represents one possible evolution. The spread between members helps explore forecast uncertainty.
- Multi-model ensemble
- A combination of several forecast centres. In Luso Meteo regional maps and charts, each model has equal weight, regardless of its number of members. Official global maps use the combination specified by their source.
- Terciles
- Three categories of the climatological distribution: below, near and above normal. In the historical reference, each category accounts for approximately one third of cases.
- Probability
- The proportion of forecast outcomes assigned to a category. The probability of above-normal precipitation describes the accumulated total over the period; it is not the probability of rain on a particular day.
- Median, P10 and P90
- The median is the central value of the forecast outcomes. P10 and P90 enclose approximately the central 80% of the forecast distribution. This range is not a calibrated confidence interval.
- P10–P90 spread
- The range between the 10th and 90th percentiles. A wider range indicates more diverse outcomes. A narrow range alone does not demonstrate greater forecast skill.
- Hindcast or retrospective forecast
- A forecast for past years, produced with the model system used for the current forecast. It provides a reference for calculating comparable anomalies and categories.
- Issue and lead time
- The issue identifies the forecast initialisation. Lead time is the time until the forecast period. Valid months are shown in the selectors and charts.
- Monthly and three-month products
- Monthly products summarise each month. In Luso Meteo three-month products, precipitation is summed and temperature is weighted by the number of days in each month. Successive three-month periods may overlap.
- Resolution and interpolation
- The seasonal grid has cells of around 1°. Bilinear interpolation creates smooth visual transitions between cells without adding meteorological resolution. Zooming does not turn the product into a forecast for individual locations.
- Model agreement
- Similarity in the signal predicted by different models. Several models can agree and still be wrong; agreement is not a calibrated probability of a correct forecast.
- Forecast skill
- Performance demonstrated by comparing forecasts with an independent historical reference. This page does not yet provide a forecast skill score calculated by Luso Meteo.
- Euro-Atlantic regimes
- Large-scale circulation patterns: NAO+, NAO−, blocking and the Atlantic ridge. The ECMWF chart for the coming weeks is sub-seasonal and independent of the monthly seasonal forecast.
- NAO+ and NAO−
- Positive and negative phases of the North Atlantic Oscillation. They describe different atmospheric circulation patterns; their effects on Portugal depend on the season and the specific situation.
- Pressure and geopotential at 500 hPa
- Mean sea-level pressure helps identify surface circulation. Geopotential at 500 hPa describes circulation aloft. Their anomalies alone do not determine the weather at an individual location.