The Common Data Cycle

There are many scenarios where data can play an important role on a farm. While these situations can be very different, they follow a similar data-usage pattern. This page documents a common data cycle that can be applied to many areas of your farm — vineyards, orchards, row crops, and beyond. To see these concepts in practice, see the Tutorials.

Data Cycle Steps Include:

1. Collection
2. Visualization
3. Filtering and Trimming
4. Interpolation
5. Sampling Validation Points
6. Translation

#Collection

There are many ways to collect spatial data on your farm:

Proximal Sensors: Proximal sensors are devices that are used in the field to measure things like NDVI, soil electrical conductivity, harvest yield, and more. These sensors log data that can often be uploaded.

Handheld Collector Data: Every.Farm allows data to be collected directly from a mobile device in the field. Any information that can be counted, seen, etc, by a human being, can be collected using a Data Collector.

Remote Sensors: More and more, services are coming online that give growers access to remotely captured data — usually gathered by satellites in space. Every.Farm's built-in satellite plugin streams imagery directly from open-data archives (NASA, ESA, USGS, USDA) and turns it into analytical layers on your map. This data can come in many forms but is often a raster; see the Raster Analysis Plugin.

Once data is in Every.Farm, it can be organized into Folders, shared with collaborators via Collaboration Settings, and even edited by moving and deleting features within datasets.

#Visualization

Visualization is the process of representing data visually. Within Every.Farm, this usually means coloring mapped data based on a variable. Each dataset in Every.Farm (as well as the farm/farm blocks) has settings for establishing how the data is visualized (see Upload Data). Visualization is important throughout the following steps because it provides visual feedback as data is being processed.

#Filtering and Trimming

Data collected within biological systems tends to have noise and extend beyond the geographic boundaries that we are interested in learning about. Every.Farm provides simple features that allow for noise to be filtered out and data to be trimmed by copying a dataset. By filtering and trimming data, our visualization will become more distinct and we will begin to see trends emerging on our maps.

#Interpolation

Even with data filtered and trimmed, it can still be hard to see broad, useful trends across a farm. Interpolation is a form of statistical analysis that smooths geographic data and makes it much more useful for implementing management strategies. As a bonus, interpolated datasets are rendered onto common grids which make them useful for comparing regions of your farm over time.

#Sampling and Validation Points

With most datasets, it is useful to be able to validate data by collecting a relatively small number of high-accuracy samples in the field that can then be compared with the dataset to ensure a correlation exists. Every.Farm allows for sampling grids to be generated and data to be collected with Data Collectors.

#Translation

Once data has been processed and a variety of sample data collected, we can use the Translator Plugin to translate correlated datasets into useful agronomic data maps. For instance, an NDVI map might be used in conjunction with a handful of berry-count or yield sample points to generate a complete spatial map of those measurements across the farm.