¿Cómo podemos ayudarte?
What is it
The productivity map reflects the variability of the field by analyzing the vegetation index NDVI.
It allows dividing the field in classes from very low to very high productivity, thus creating variability polygons that will later help the agronomic advisor to make the most convenient decisions for managing the field.
How is it generated?
The productivity maps are generated based on quantile or cluster analysis of images selected over time from the Sentinel and Landsat sensors. The selection of the images correspond to key dates, where the crops reflect a high level of activity, related to the phenological and maturation stages of the vegetation.
The selected images are normalized to be able to carry out the classification. Cluster analysis consists of grouping pixels by similar values, seeking maximum homogeneity within each class and the maximum differences between classes. On the other hand, the classification by quantiles distributes equally in classses based on the values of the pixels.
Both analysis methods lead us to obtain a synthesis of information represented in the productivity map, which allows comparing responses of the soil at different times in a single thematic map.
To request a productivity map for your batch, send an email to email@example.com requesting the service.
Layer data structure
The productivity map has information associated with the layer and each class within the map.
Layer data, contains data from the creation and origin of the productivity map, such as:
Number of classes
Date of the images and sensor, the acronym S2 and L8 are used for Sentinel2 and Landsat8.
Classification method used: Cluster or Quantile.
Class data, stores relative information for each of the classes:
|Number that refers to the productivity class
|Average value of the NDVI index for the selected class.
|Minimum value of the class index.
|Maximum value of the class index.
|25th percentile of the NDVI index for the class, it is a statistical measure of dispersion. Below that value are 25% of the data.
|75th percentile of the NDVI index for the class, it is a statistical measure of dispersion. Below that value is 75% of the data.
|Indicates the standard deviation associated with the mean value of each class.
What is it for?
The Productivity Map is the ideal starting point to know the variability of the field and define the different productive environments of each management unit. Through them we can:
- Relate other types of available relevant information (yield maps, soil variables, laboratory results, etc.) which will help us make different management decisions.
- Guide field visits to key locations, and validate the different environments
- Delimit the productive zones.
- Adjust the boundaries of each productive area according to our own criteria (performance, climatic, economic, etc).
- Establish georeferenced sampling stations or sampling grids by environment.
- Apply variable rates of inputs in macro and micro environments.
All these options that are incorporated into the daily routine in the management by environments through the Productivity Map result in a deeper knowledge of the different areas of the field and thus, improve management of existing limitations.
Farm360 includes analyzes that can be carried out from a productivity map:
- Index report by area: allows you to select a date and make the intersection with each productive zone.
- Yield by area: by selecting a yield map and the productivity map, this report serves to know the average yield in each productive zone and its variability.
- Prescription assistant: from a productivity map it is possible to make variable prescriptions by selecting the dose, input and channel to apply.
The productivity map is not complete until it is validated in the field.
Each productivity environment must be compared with the real situation in the field, verifying the condition of low or high productivity.
Farm 360 has an app that allows targeted and efficient field visits. Positioning yourself at the points you want to travel, capturing photos and making annotations.