Yield data and analytics in GeoPard

In this article:

  • Using yield data in precision agriculture
  • In-depth yield data analytics in GeoPard Agriculture 
  • Visualization of each attribute in Yield files
  • Correction of raw yield data 
  • 5 Practical examples of usage of yield maps
Raw and cleaned yield data in GeoPard
Raw and cleaned yield data in GeoPard

Yield data allows you to make more informed decisions and improve growing efficiency.
Field management zones constructed from multiple years of yield data are suitable for an initial assessment of potential yield and soil nutrient variability to make future crop management decisions.
Analysis of yield data can be converted to a variable rate application map and used, for example, for fertilizer application.
Yield data calibration is another topic you need to consider, we will cover it in a separate blog post.

The advanced analytics in GeoPard is that you can perform multi-layer analysis by combining multiple layers of data into one map and look for relationships between the data layers. 
Combined productivity zones can be generated based on vegetation indices from satellite imagery, topography, data from machinery such as yield, electrical conductivity, soil moisture, and others, as well as agrochemical analysis results.

Visualization of yield files is done automatically after downloading the file, its automatic processing and cleaning. Two versions of maps are shown below – the original image with data from the equipment monitor as is, and the GeoPard visualization. The raw data has been converted into a gradient continuous surface image, for easier understanding of the the field heterogeneity and creating management zones. 
Each of the attributes of the yield file is available for visualization, such as moisture, yield mass, yield volume wet and dry, downforce, fuel consumption, machine speed, and so on.

Raw data correction means that if a point in the field is unnatural, it will be smoothed (for example, working over not the full width of the combine header). When creating Zones based yield data, you can correct individual zones and polygons. 

Let’s take a look at some practical examples of using yield maps and other GeoPard data layers.

1. Management zones based on yield data. Management zones can be constructed based on either one year’s yield data or multiple years. It is important to note that you cannot directly stack yields from different years, as you will get a bias in favor of one of the years. To reduce this effect, GeoPard applies several algorithms to make the weight of each year even.  You can set the importance of a single year through the Weight tool when you create a Multi-layer map. Such field management zones can be used to build application/prescription/Rx (VRA) maps, calculating the potential yield in each zone.

Multi-year and multi-layer yield potential map
Multi-year and multi-layer yield potential map

2. Multi-layer zones with yield data and other data sources (topography, soil, sensor, satellite). Yield data can be added to multilayer analytics and set the weight it will have on the final zones. In this example, three layers of data are added to the map: Yield, Satellite imagery, and Topography. You can combine any data layers you consider relevant for analytics. The multi-layer map can be used for further analytics and creating VRA maps. 

Yield, Topography and Satellite imagery
Multi-layer zones: Yield, Topography and Satellite imagery

3. Yield calculation on zone and field level. To analyze different treatments, seed varieties and agronomic practices you probably want to compare average and total yield in each zone, strip or between fields. GeoPard automatically calculates this for you to make it easier to compare yield in absolute numbers. 

Yield in abs numbers based on Yield files
GeoPard calculates yield in abs numbers based on Yield files. Total and average for field and each zone

4. Dependency zones basedon yield data. Zones based on yield data can be overlaid on other data zones and you can search for dependencies between data layers. This example shows the overlay of high yield and average protein (1) and low yield and high protein (2) of different wheat varieties in a field.
Other examples include the relationship between the influence of topography on yield, the intersection between low yields and the lack of macro- and micronutrients in the soil, soil moisture and electrical conductivity (EC) layers.

Intersections of different yield and protein levels
Intersections of different yield and protein levels

5. Variable Rate application (VRA) maps based on yield data.  You can create prescription maps for different operations – fertilizing, seeding, spraying, irrigation and planning of soil sampling. You can edit the number and shape of the zones. 
You can also build a prescription map for a variable rate application by combining yield data with other data sources (soil, EC, satellite, topography). 

Variable rate Seeding rates per zone
Variable rate Seeding rates per zone

Even if you do not have yield data, you can use GeoPard multi-year zones (up to 33 years) based on satellite imagery or combine it with other data layers like topography to start your precision agriculture journey. These analyses often correlate with yield data, but this is another story.

Multi-Layer Analytics

Maps generated with a single data layer and several data layers.

Precision agriculture is capable of generating vast amounts of data in the form of yield data, satellite imagery, and soil fertility, among others. Lack of easy-to-use cloud precision software toolkits that assist crop producers in converting field data layers into useful knowledge and actionable recommendations limits the application of precision agricultural technologies. In precision agriculture, management zones are areas within a field that have similar yield potential based on soil type, slope position, soil chemistry, microclimate, and/or other factors that influence crop production. The producer’s knowledge of a field is a very important piece of the process. Management zones are thought of as a mechanism to optimize crop inputs and yield potential.
The big challenge is to build management zones that perfectly reflect field variability. A combination of different layers like satellite imagery, soil fertility, topography derivatives, and yield monitor data is the next logical step to generate more responsive management zones.

Multi-layer analytics (also known as integrated analysis) is becoming a part of the GeoPard geospatial analytics engine.

Classic combinations of integrated analysis parameters include one or more of yield data, NDVI map, elevation, and soil sensor physicochemical characteristics. GeoPard supports these parameters and in addition, allows the inclusion of other field data layers either already available in the system or uploaded directly by the user (soil sampling, yield datasets, etc.). As a result, you are free to operate with the complete set of parameters doing integrated analytics:

Yield data
Remote sensing data:
    –   Potential productivity map (single-year and multi-year)
    –   Stability/variation map
    –   Vegetation indices NDVI, EVI2, WDRVI, LAI, SAVI, OSAVI, GCI, GNDVI
Topography:
    –   Digital elevation
    –   Slope
    –   Curvature
    –   Wetness index
    –   Hillshades
Soil data:
    –   pH
    –   CEC (cation exchange capacity)
    –   SOM (soil organic matter)
    –   K (potassium)
    –   Thin topsoil depth, lower available water holding capacity (drought-prone soil)
    –   EC (electrical conductivity)
    –   and other chemical attributes available in the uploaded dataset

It’s important to emphasize that custom factors are configured on top of every data layer to assign the desired layer weight. You are very welcome to share your integrated analytics use cases, and build management zones maps based on your knowledge of the field while selecting data sources and their weights in GeoPard.

Pictures in this blog containes a sample field with data layers (like a productivity map covering 18 years, digital elevation model, slope, hillshade, 2019 yield data) and various combinations of integration analytics maps. You can follow steps of evolution of management zones while extending integration analytics with an additional data layer.

    Request Demo

    By clicking the button you agree our Privacy Policy

      Subscribe


      By clicking the button you agree our Privacy Policy