# 1. Executive summary
The Portugal Almond Production Forecast is an estimate of the national almond crop for a given marketing year, produced by Demeter in partnership with Portugal Nuts. It is constructed bottom-up from every almond orchard in the country and takes into account age, weather, water availability, condition and a range of other metrics. It is calibrated against the official production record published by the Instituto Nacional de Estatística (INE, Statistics Portugal) and cross-checked against various statistics provided by Portugal Nuts. It is expressed in tonnes of in-shell almonds together with a kernel equivalent. It is published with a range that reflects the method's measured historical error.
The forecast is published as a point-in-time estimate. It reflects all the information available on the date it is issued and is updated as the season progresses. Applied retrospectively to every season since 2017, the method's average error against the final INE figure is 8.3%, measured at the last print before harvest. Further detail on historical accuracy is provided below.
# 2. Design principles
Bottom-up construction. The forecast starts from a registry of every standing almond orchard in the country, each with its own estimated planting year and production class. It does not employ a sampling methodology, grower surveys or top-down trend extrapolation. National production is the sum of the expected production of each orchard.
Tonnes of crop as the goal In a young and fast-growing industry, yield per hectare is a misleading statistic: a wave of newly planted, non-bearing orchards drags the average down while total production rises. The forecast outputs a number of tonnes, and reports yields only as a derived quantity. This is appropriate for a young and fast-growing industry where productive orchards are distributed across a wide range of ages and maturities.
Tailored to the region. The core methodology builds on Demeter's experience forecasting production in other crops and regions, but each input is validated independently for its contribution to accuracy in Portugal specifically. The level is fitted to the INE's official production series. Some inputs are unique to Portugal, including those provided by Portugal Nuts.
Point-in-time and like-for-like. The current season is compared with historical seasons over the same calendar window. A forecast issued in June compares June-to-date inputs in the most recent year with June-to-date inputs in every prior year.
Transparent on accuracy and validation approach. Every published figure carries its range. The method's track record is disclosed in this document and elsewhere. Accuracy is measured by leaving each year out in turn, re-fitting the model without it, and scoring the prediction against the actual figure.
# 3. Conceptual framework
The forecast combines three layers. The first two feed the model, while the third is a set of independent checks applied before publication.
| Layer | Input captured | Source | Update frequency |
|---|---|---|---|
| 1. Orchard base | Location and size of each orchard, its production class (such as planting frame and irrigation source) and its age; orchards removed, abandoned or otherwise degraded; historical footprint in each past year | Satellite and aerial imagery, parcel declarations | Annually, with in-season removal monitoring |
| 2. Growing conditions | The season's weather relative to history, as measured by the relevant Demeter Growing Season Index | Hourly meteorological data reanalysis | Daily |
| 3. Independent checks | Hyper-local ground stations, crop health, qualitative and ad hoc quantitative checks | Various | In-season |
# 3.1 The orchard base
The full extent of the national orchard footprint is compiled, and validated by Demeter. Each is assigned a production class, distinguishing for example recent irrigated commercial orchards from older commercial orchards under deficit irrigation or traditional rainfed orchards. These labels are derived from observable characteristics as determined by Demeter. Each orchard is also labelled with its year of planting or a range of years if a precise determination cannot be made. In Portugal these labels cover a period considerably in excess of 20 years.
The orchard base also reconstructs orchards that stood in earlier years but have since been removed, so that each historical season is modelled on the orchard base that actually existed at the time. This considerably improves the accuracy of forecasting models.
The various classes are each assigned a default yield curve derived from a range of inputs. The combination of these yield curves provides a baseline from which deviations are estimated based on conditions in the current and prior seasons.
# 3.2 Growing conditions
Growing conditions are captured through their effect on the Demeter Almond Season Index - Portugal (DASI-P). This index is available at the level of each orchard and allows the model to incorporate the influence of weather in both the current and prior season.
DASI-P is calibrated on 25 years of weather conditions in Portugal and captures the variables that consistently affect yield outcomes, quality and cost of production. DASI-P commences live publication on 28 September 2026, and further information on the index will be available on its dedicated page from that date.
# 3.3 Independent checks
The output of the model is validated against a range of instruments that do not feed the model directly. These are intended as a sense-check against conditions on the ground and serve primarily to move the headline figure within its error range.
Examples include an ongoing health monitor on all orchards in the productive base and a number of sentiment channels. Independent checks include a range of inputs from Portugal Nuts.
Not all independent checks are relevant at all stages in the season, and their impact is most important in "outlier" years. The list varies over time and is not disclosed in its entirety.
# 4. Forecast construction
Orchard-level forecasts are summed into regions. Regional totals are used for internal diagnostic purposes but are not published due to a historical record that is insufficiently long at present.
Regions are summed to a national figure, which is the source of the headline publication.
The specific class definitions, yield curve parameters, channel weights, calibration values and full list of independent checks are proprietary to Demeter and Portugal Nuts and are not disclosed.
# 5. Track record
The forecasting method has been applied retrospectively to every season since 2017. The history of the orchard base indicates that this is the first year in which the modern irrigated sector becomes accurately observable.
Each year's estimate is produced out of sample: the model is calibrated without that year, then asked to predict it.
| Year | Modern irrigated orchards at bearing age (kha) | Other classes or orchard at bearing age (kha) | Actual production (kt in-shell, INE) | Actual production (kt kernel, INE × 30%) | Estimate (kt in-shell) | Estimate (kt kernel) | Error |
|---|---|---|---|---|---|---|---|
| 2017 | 0.0 | 20.0 | 23.1 | 6.9 | 23.1 | 6.9 | +0.2% |
| 2018 | 0.7 | 20.7 | 17.1 | 5.1 | 20.6 | 6.2 | +20.4% |
| 2019 | 2.8 | 22.1 | 32.3 | 9.7 | 32.4 | 9.7 | +0.2% |
| 2020 | 4.2 | 23.5 | 31.6 | 9.5 | 36.9 | 11.1 | +16.9% |
| 2021 | 9.1 | 26.9 | 41.5 | 12.5 | 46.2 | 13.9 | +11.5% |
| 2022 | 12.5 | 30.0 | 46.2 | 13.9 | 45.8 | 13.7 | −1.0% |
| 2023 | 21.4 | 32.9 | 69.5 | 20.9 | 60.6 | 18.2 | −12.8% |
| 2024 | 24.3 | 33.8 | 90.8 | 27.2 | 88.2 | 26.5 | −2.8% |
| 2025 | 30.3 | 35.8 | 82.7 | 24.8 | 90.4 | 27.1 | +9.3% |
| 2026 | 30.0 | 35.3 | 114.9 | 34.5 |
INE in-shell figures are converted to kernel using a flat crack-out ratio of 30%.
Mean absolute error: 3.7 thousand tonnes, or 8.3%. The 2018 outlier was a year of extreme bloom-rain. As the only event of such magnitude in the long-term record, its effect was underestimated by the model. The official figure for 2025 production remains provisional.
The error above is measured at the last print before harvest, with the full growing season observed. Prints issued earlier in the season carry wider uncertainty. At the start of the crop year, for example, they rest on the orchard base alone. Uncertainty narrows as the season progresses.
# 6. Definitions
In-shell The headline in-shell figure is published on the same basis as the official INE total to facilitate comparison.
Kernel The in-shell tonnage multiplied by a flat crack-out ratio of 30%. This is consistent with the kernel-basis reporting of the International Nut and Dried Fruit Council (INC).
Bearing age Refers to orchards in their third leaf and older.
# 7. Governance, updates and scoring
The forecast is maintained by Demeter's crop forecasting team. Its model specification is versioned; the 2026/27 marketing year is served under version v2, the specification frozen for it, and the 2027/28 marketing year and the track record above under version v3. Coefficients are re-fitted whenever INE publishes a new annual figure. The specification changes only between marketing years, with advance notice and with the historical series restated for comparability.
Prints are generated on an ongoing basis as new information arrives. The most recent forecast is always available
# 8. Further information
For methodology enquiries, please contact Demeter at [email protected]. For enquiries to Portugal Nuts, please email [email protected].
Information and current values for all of Demeter's forecast and index products are available on the