Crop Map

Crop mapping is an integral part of field monitoring. Crop maps are important to track agricultural land use and estimate crop production at any scale – be it the whole world, a particular country, or a single field.

Often available in open access, such information is interesting both to farmers and non-farmers. Crop diversity maps are helpful to anyone eager to know what plants were cultivated all over the globe, where exactly, and when.

What Is A Crop Map, And Why Is It A Useful Source Of Information?

The crop map specifies the types of agricultural species that grow in a selected region, the size of the territory they cover, and their stage of development at a specified time.

Crop maps can tell farmers pretty much about their lands, apart from field boundaries and crop types. A custom crop map can contain data on soil moisture, seeding and harvesting dates, vegetation indices, and more. Such advanced features go beyond the standard functionality of crop maps and are added upon request.

Crop growth mapping provides necessary insights into plant health, which allows farmers to:

  • detect changes in vegetation development;
  • timely identify drought, cold, or water stress;
  • assess vegetation damage;
  • create crop plan maps and choose species for rotation;
  • track and schedule agricultural activities;
  • conclude on soil fertility and field productivity;
  • predict yields;
  • estimate profits.

However, mapping crop areas is useful not only for agriculturalists. Agriculture crop mapping helps advisors, investors, insurance agencies, educational establishments, statistical bodies, national agricultural departments, trading companies, food production enterprises, and humanitarian aid organizations. Analyzing crop mapping data, these institutions can generate statistical reports and anticipate commodity supplies in the future.

Crop Type Mapping With Remote Sensing To Track Agricultural Lands

In the past, crop mapping data was collected with traditional methods like statistical reports, inventory records, or field inspection in person. Now, remote sensing essentially simplifies this labor- and time-consuming task. The great thing about satellite monitoring in mapping crop areas is that it allows farmers to check their field areas remotely. Besides, satellite imagery analytics provide more details on farmlands than just information on cultivated crop types and field acreage.

Remote sensing examines spectral reflectance, reporting on vegetation health, e.g., plant structure, moisture, and chlorophyll content. Optical remote sensors operate within visible, NIR, and SWIR spectral bands. Additionally, radar sensors can penetrate through clouds or haze, delivering necessary data almost irrespective of weather conditions. Further, GIS systems reference satellite imagery to definite geographical locations, creating GIS maps.

Crop Map By EOSDA: What Does It Imply?

EOSDA is a global provider of satellite-driven analytics for dozens of industries, with a significant focus on agriculture. By default, crop mapping software by EOS Data Analytics is empowered with two main technologies: field boundaries detection and crop classification, distinguishing field contours and plant types. Feel free to contact our team for additional data, and our experts will include your preferences in custom solutions.

Field Boundaries Detection

The field boundaries detection technology outlines all the fields in the selected area, either in the current or past seasons, and allows calculating field total acreage with the received vector result. Users can get the obtained data extracted in the .shp or .geotiff formats.

The boundary detection technology recognizes the contours of all objects, not only fields. So, field contours for crop mapping are to be distinguished from other objects with specific spatial analysis techniques – either with classification or manual filtering.

The technology relies on cloud-free and contrast satellite imagery from Sentinel-2. Vegetation intensity affects data precision, so the choice of images depends on the plant development stage. It is why the images are typically selected in the middle of the growing season when fields are best identified as separate objects.

The field boundaries in satellite imagery are outlined through several spatial analysis steps:

  1. Segmentation to recognize separate objects in the image.
  2. Data vectorization since the initial result is acquired in the raster format.
  3. Additional processing (simplification) with in-house EOSDA algorithms.

The obtained result can serve as a separate product or be used to get other classification results in the vector format.

The minimum field size for accurate crop satellite mapping or field boundaries detection should exceed 3-5 ha, which is determined by the resolution specifics of Sentinel-2 imagery.

A Complete Crop Classification Map By Region

For crop type classification and mapping, EOSDA analysts use several data sources, both satellite and ground-based. Their combination improves outcome data accuracy (up to 90% precision):

  • Time series of satellite imagery (in particular, from Sentinel-1 and Sentinel-2).
  • Current and historical ground data on the plant types and seeding dates in the area in a specified season.

The output data quality essentially depends on input data quality and quantity, which is ensured in the following ways:

  1. The use of both optical and radar images increases the variability of results, and the final result combines several variants.
  2. The technology of cloud detection and cloud masking improves the quality of optical imagery containing segments with cloud cover, haze, or shadows.
  3. A set of tools and methods to filter invalid data helps in ground data validation and control.

Thanks to the technology application, users can get:

  • the total area of each agricultural species;
  • plant classification for current and past seasons;
  • ready 10-m resolution cropland masks in .geotiff or .shp formats.
Cropland mask doesn’t only help identify croplands vs. non-croplands and cultivated plant types within a given region. It also allows detecting rotation deviations based on this data. This way, cropland masks can facilitate more efficient crop rotation mapping.

EOSDA Crop Production Map Of Ukraine

Apart from providing many actionable insights, modern crop field maps are fun and easy to use, thanks to visualization and interactive data presentation. The EOSDA crop map of Ukraine is a vivid example of satellite technology implementation. Here is what this map of cropland use can tell you:

  • areas of registered (cadastral) and unregistered fields (if the cadastral division layer is available and relevant);
  • agricultural species produced in the selected year (from 2016 to 2019);
  • total area of each agricultural species in a selected region, district, or the whole Ukraine;
  • customized rotation data upon request.

Each plant type is assigned its specific color; for example, maize is highlighted in yellow, and cereals are shown in pink. This way, it is convenient to understand species distribution, with cereals, sunflower, maize, and soybeans ranking top.

In particular, this crop type map shows that cereals prevailed in Ukraine in 2019, produced on 12.2M ha of the country’s farmlands. Zooming in and toggling between the regions, you can see the top species and the total acreage by each of them in the selected area. For example, in 2019, the Kyiv region accounted for 279.3 ha of cereal fields, but the main type there was maize covering 546.5K ha.

If you are interested in a customized crop type mapping for a particular region, just contact us at, and our experts will advise you how to realize it best.