Crop Aggregation

The aggregation process for crop products involves additional steps specific to this term type.

Aggregation stages

Aggregation of crops proceeds in multiple stages:

1. All public Cycles on the platform corresponding to a given time period, crop (where the crop is the primary product), and country are gathered.
2. Cycles are grouped into four categories: conventional and rainfed; conventional and irrigated; organic and irrigated; organic and rainfed. Data are aggregated within each group, creating up to four sub-aggregations. These groupings allow data to be weighted to account for substantial differences in farming practices. Further groupings (e.g., greenhouse vs field grown) will be included in future.
3. The sub-aggregations are combined into one aggregation, with each sub-aggregation's Cycle and Site information weighted by the area share of its group, using data from FiBL for organic farming and AQUASTAT for irrigation. These weightings are currently country-specific but not crop-specific.
4. The impact assessments of the four sub-aggregations and the final country aggregation are then recalculated.

Example

Creating an aggregation for Quinoa, grain - Peru - 2010-2025.

All Cycles with primaryProduct = Quinoa, grain, country = Peru, and endDate between 2010-2025 are gathered. Four sub-aggregations can be created:

• Quinoa, grain - Peru - 2010-2025 - organic - irrigated
• Quinoa, grain - Peru - 2010-2025 - organic - rainfed
• Quinoa, grain - Peru - 2010-2025 - conventional - irrigated
• Quinoa, grain - Peru - 2010-2025 - conventional - rainfed

The final aggregation Quinoa, grain - Peru - 2010-2025 is generated by assigning different weights to each sub-aggregation. According to FiBL, the average share of organic agriculture in Peru between 2010 and 2025 was 2%, and according to AQUASTAT the average share of irrigated land was 30%. The following weights are therefore assigned to calculate the weighted mean:

Information on the weightings used for the sub-aggregations can be found in the Cycle and Term descriptions in the weighted aggregations.

Each sub-aggregation's weight is the product of its organic share and its irrigation share:

$$w_s = O \times I$$

Where:

• $w_s$ = the weight of the sub-aggregation
• $O$ = $p_{org}$ for an organic sub-aggregation, $1 - p_{org}$ for a conventional one
• $I$ = $p_{irr}$ for an irrigated sub-aggregation, $1 - p_{irr}$ for a rainfed one

The organic share comes from the region-standardsLabels-isOrganic lookup (FiBL), averaged over the aggregation period:

$$p_{org} = \min\left(1, \frac{V_{org}}{100}\right)$$

Where:

• $p_{org}$ = the organic share of the country's area
• $V_{org}$ = the lookup value as a percentage, averaged over the period

The irrigation share comes from the region-irrigated lookup (AQUASTAT):

$$p_{irr} = \frac{A_{irr}}{A_{total}}$$

Where:

• $p_{irr}$ = the irrigated share of the country's area
• $A_{irr}$ = the irrigated area, averaged over the period
• $A_{total}$ = the total area for the same site type, averaged over the period

Both shares fall back to the default region when the country has no lookup value, and to 0 when neither does — so an unknown organic share means the country is treated as entirely conventional. The irrigation share is resolved by site type in order: cropland, agriculture, all, then land area.

Plantation weighting

Plantation crops, which are defined by the lookup isPlantation, are weighted based on their plantation phase. There are three phases:

• Preparation phase: The first year of plantation, which can include activities like tree clearing or soil liming, that only occur once in a plantation's life.
• Non-productive phase: The period where a plantation is not producing a commercially viable harvest, this does not include years where the yield is simply lower than that of a fully mature plantation.
• Productive phase: The period where a plantation is producing a commercially viable harvest. The yield of this harvest may still vary during the productive phase (increasing up to maturity and then decreasing at end of life).

The practices Preparation phase (permanent crops) and Productive phase (permanent crops) are used to sort the cycles into the categories following the logic below:

• Productive phase: the practice Productive phase (permanent crops) = TRUE.
• Preparation phase: the practice Productive phase (permanent crops) is not TRUE, and the practice Preparation phase (permanent crops) = TRUE.
• Non-productive phase: neither practice is TRUE.

The phases are tested in that order, so a Cycle marked as both productive and preparation is counted as productive.

The weight for each category is then calculated from the default glossary values for plantation_lifespan and plantation_non_productive_lifespan:

$$w_p = \begin{cases} \frac{L - N}{L} & \text{productive phase} \\ \frac{365}{L} & \text{preparation phase} \\ \frac{N - 365}{L} & \text{non-productive phase} \end{cases}$$

Where:

• $w_p$ = the weight of a Cycle, by the phase it describes, which is not stored
• $L$ = the plantation lifespan in days
• $N$ = the non-productive lifespan in days
• $365$ = the preparation phase, fixed at one year

When either lookup value is missing, every Cycle is weighted 1 and the phases are not distinguished.

Information on the weightings used for the plantation phase can be found in the Cycle and Term descriptions of the sub-aggregations.

Primary product

When aggregating the primaryProduct, values are adjusted to equivalent Dry matter before averaging. The Dry matter value to which they are adjusted is the default value in the glossary, dryMatter.

Water completeness

When aggregating water completeness, if any of the sub-aggregations are irrigated, the rainfed sub-aggregations are excluded from the water completeness calculation. This ensures the weighted aggregation does not include irrigated Practices with no water Inputs and water completeness set to TRUE.

World aggregations

World aggregations are created by combining verified country aggregations. The country aggregations are weighted by their production volume, based on data from FAOSTAT.

$$w_c = \min\left(1, \frac{P_c}{P_{world}}\right)$$

Where:

• $w_c$ = the weight of country $c$
• $P_c$ = the production quantity of the product in country $c$, averaged over the aggregation period
• $P_{world}$ = the world production quantity of the product over the same period

A missing production quantity falls back to 1 on each side independently, so a country with no production data is weighted $\min(1, 1 / P_{world})$ rather than dropped. When the product has no production-quantity lookup at all, every country is weighted equally. If no country ends up with a weight above zero, the World aggregation is not produced.

The countries combined, and the weight each was given, are listed in the description of the World aggregation.