Project Name:

Testing Sentinel-2 vegetation indices for the assessment of the state of winter crops in Bulgaria (TS2AgroBg)

Project Manager: Assoc. Prof. Petar Dimitrov, PhD

Technical Manager: Assoc. Prof. Georgi Jelev, PhD

Project Summary

The TS2AgroBG project is motivated by the application potential of remote sensing technology in Bulgarian agriculture sector. While aerial photographs have been used in this sector for a decade for agricultural parcels identification, the full potential of satellite applications for crop growth assessment and crop mapping remains to be explored. The innovative Sentinel-2 mission will expand the possibilities by providing free imagery with high spectral and spatial resolution suitable for crop state monitoring at field level. The PROBA-V satellite acquires 100 m data at high temporal frequency enabling crop mapping based on image time series. In this context, it is opportune to raise awareness of the capability of these advanced technologies to the Bulgarian policy makers as well to other agricultural stake-holders. This project could also promote the development of applications and services for this emerging market in Bulgaria.


For effective crop management it is important to have timely information about the distribution and areas of different crops and about their growth status and probable yield. To acquire this information, satellite imagery from Sentinel-2 and PROBA-V seems to be of great benefit. In this project, data from these two satellites will be used to generate agriculture-relevant products, including: raster layers with crop bio-physical variables, crop state assessment maps, crop masks and national crop map. Because this is a pilot study, the project will focus on the most important winter crop in Bulgaria – wheat. In addition to the common bio-physical variables like Leaf Area Index (LAI), fraction of Absorbed Photosynthetically Active Radiation (fAPAR) and Canopy Chlorophyll Content (CCC), the project will investigate the possibility to retrieve total nitrogen content, fresh and dry biomass (AGBf and AGBd), and fraction vegetation cover (fCover) from Sentinel-2 Level 1C data. Vegetation indices calculated from Sentinel-2 data will be used to develop empirical regression models for computation of bio-physical variables. Dedicated field campaigns will be conducted to obtain consistent ground truth dataset for the regression models’ calibration and validation. The data collected will include the following bio-physical variables: total nitrogen content, total (a+b) chlorophyll content, LAI, fAPAR, AGBf, AGBd, and fCover. The Institute of Soil Science, Agrotechnologies and Plant Protection "Nikola Poushkarov" will provide valuable support during the field campaigns and laboratory analyses of plant samples. Based on these bio-physical variables a crop state assessment map will be generated. To facilitate the generation of crop masks in operational context an extension of the software SPIRITS is proposed to handle and process Sentinel-2 Level 1C data. This task will be carried out by the Belgian partner – VITO. In addition, sub-pixel crop map at national level will be produced through classification of POBA-V 100 m time series data. Classifications of Sentinel-2 imagery over two training/test areas will be used to calibrate the national level PROBA-V classification. Finally, the possibility for assimilating the Sentinel-2 derived bio-physical variables into a crop growth model in order to improve winter wheat yield prediction will be examined.                          


The main objectives for the project are to:

1.    Conduct series of field campaigns in a selected test site and to provide geo-database containing ground measured bio-physical variables and semi-synchronous remote sensing imagery;

2.    Calibrate and validate regression models for retrieval of different bio-physical variables using Sentinel-2 vegetation indices applicable to winter wheat grown in Bulgaria; and then use this information to generate crop state assessment map;

3.    Extend the capability of the software SPIRITS to assess the crop state using Sentinel-2 data. In absence of long Sentinel-2 time series, the capability of the software will be demonstrated by creating crop masks in the study region;

4.    Produce crop map at national level using sub-pixel classification of PROBA-V 100 m time series;

5.    Assess the possibility for assimilating Sentinel-2 derived bio-physical variables into the WOFOST crop growth model.





  • Space Research and Technology Institute – Bulgarian Academy of Sciences (SRTI-BAS)leading partner
  • Institute of Soil Science, Agrotechnologies and Plant Protection "Nikola Poushkarov" (ISSAPP "N. Poushkarov")
  • Vlaamse Instelling voor Technologisch Onderzoek NV (VITO)


The Project is funded by the Government of Bulgaria through an ESA Contract (4000117474/16/NL/NDe) under the PECS (Plan for European Cooperating States).


The view expressed herein can in no way be taken to reflect the official opinion of the European Space Agency.