Agro GIS

Application of "Agro GIS" system 

Automated assessment system of crops condition and yields prediction, according to satellite imagery (Agro GIS) created to:

  • Control the use of agricultural land.
  • Automatically determine the status of crops in the phases of growth and development.
  • Identify fields with deviations from the norms, their evaluation, systematization.
  • Identification of necessity of nitrogen fertilizing and adjustment of their doses.
  • Determination of the dynamics of terrestrial biomass for major crops.
  • Prediction of crop yields.

Download the presentation - geographic information system for the agricultural business or automated assessment system of crops condition and yields prediction.

Other possible features that are not included in this system can be found here.

The functioning of the automated evaluation system of crops condition is based on the use of multispectral satellite images*. All researches using materials of satellite imagery base on the principle of absorption and reflection of electromagnetic spectrum emission.

For each phase of the crop growing and their condition there are strict relationships between absorbed and reflected emission energy of the electromagnetic spectrum. Using special software and mathematical algorithms, we interpret the calculated data as a general assessment of each individual field, form a map of nitrogen fertilization and yield prediction about 1.5 - 2 months before harvesting.

 *Multispectral images - images that allow various spectral calculations and convert the spectra to obtain thematic products.

The necessary input data for the system formation:

The contours of fields - information about the boundaries of each field that is being tracked in the system. Typically, this information provides by customer in the form of geospatial data (eg vector files in the form of fields contours; kml / kmz files formed using Google Earth; plans, diagrams, maps and so on.). This information is required to determine the boundaries of space imaging and accordance formation of received calculations according to each individual field.

Information about culture - every cultivated culture has its own characteristics. Assessments scales and mathematical algorithms of yield prediction for each individual culture. This is related with the difference in period of sowing, harvesting and different vegetation phases of crops.

Actual last year's harvesting - the information necessary to create the statistical matrix for yields calculation - putting into the system corrective parameters for the "current" monitoring season. Data provided by "last year" fields that were planted with "predictive" culture. The main requirement - to "last year" and "current" fields must be located in a same climatic zone within cultivated areas.

Space images - a necessary element within the system. Depending on the crop that will be monitored, space imaging schedule is planning. It is caused by the different phases of growing for different crops.

You can select the following types of satellite images:

Observation monitoring - images of this class are used to monitor large areas (over 10,000 km2). Spatial resolution* - 22-30 m / pixel.

Basic monitoring - spatial resolution for such images - 10 m/pixel. The minimum possible area of monitoring using data of this type - 400 km2

Advanced monitoring - spatial resolution for these images - 5 meters/pixel. The minimum possible area of monitoring using data of this type - 400 km2.

Detailed monitoring - using satellite images with super high and medium resolution (50 cm/pixel - 2.5 m/pixel). The use of this type of monitoring is discussed individually.

Combinations of different types of data - the customer determines for himself detalization level used in the system of satellite images or their combinations. The lower spatial resolution - the cheaper image and detalization of all derived and calculated data.
For qualitative assessment of fields condition, mapping for nitrogen fertilizing and yields prediction should be at least 4 images during the whole period of monitoring.

* Spatial resolution - one of the main characteristics of satellite images. Specifies the minimum size of an object on the ground, which can be seen on the satellite image as a single pixel.

 

The system is based on a "client-server" principle. All calculation results, analytical descriptions, tables and reports, satellite images and processing results stored on the server. The server can be installed as a service on the contractor organization side or directly at the customer. The best way is to install a server at the provider and organize protected channels for information exchange with the server. The customer gets access to this information via web links without the need for installation any software or applications. All work takes place in the browser via the Internet. The system provides the ability of authorization - for issuing usernames and passwords to different users. Depending on the access level of a user (determined by individual login / password), the user receives the appropriate functionality, up to inability to view the "closed" objects as part of the general work area.

The system provides the ability of adding your own personal data (documents, comments, geospatial data) to the server using a special web interface via the Internet.

Analytics and types of information provided by the system.

Satellite images - the user can visualize all satellite images, which are used in the system, spectral indices** calculated for each of them, switching between the dates of satellite imagery for visual analysis of the development dynamics (harvesting control). All provided by customer geospatial files (contours of fields) are displayed in the system with images simultaneously.

Map of nitrogen fertilizing*** - formed in a geospatial file from satellite image and displayed in the system. Period of next map formation depend on the scheme of fertilizing. One gets reporting document with recommended norms for fertilizing.

  • The end of March - early April (frozen-melted soil);
  • End of April - beginning of May (end of tillering - early exit in the tube);
  • Middle of May - exit in the tube.

The contours of the fields - are displayed in the system as the active table. Besides the field code this table also contain area, automatic assessment of the field condition on a five point scale, the result of yield calculating for a particular field. It is possible to sort columns and switching them off. By clicking on the field in the table, the system automatically visualizes it right on the center of the screen on satellite image (and vice versa - when you click the field contour on image system highlights the field in the table), there is an analytical brief field description for selected date of the image.

A brief analytical description - the system provides the ability to display text information about the field in a separate window. Analytical description formed by analyst operator as result of satellite images decryption and mathematical calculations.

General reporting - all analytical text information can be displayed by the user in the form of common spreadsheet document (Excel file) just by clicking a key. It is possible to generate a common report for all dates and all cultures, and generate report on a specific date for the chosen culture (or vice versa).

 ** Calculated spectral index - geospatial file (thematic material) obtained through the use of processing algorithms for spectral image.

 *** Map of nitrogen fertilizing - means fertilizing map creation for winter grain crops.

As a result of commercial implementations of the system in 2012, the harvesting prediction accuracy was 96% (for winter wheat).

This information is documented by customers of the system. The system was calibrating for 5 consecutive seasons on specialized research stations. The system is calibrated for the following crops:

  • winter wheat;
  • corn;
  • sunflower;
  • oilseed rape;
  • spring wheat;
  • spring barley;
  • soybean.