Blog Post

Field work for remote-sensing-based crop type mapping

Jul 11, 2022

TRANSECT collected field samples on crop types in Tajikistan and Kazakhstan

Dried soil after irrigation, Kazakhstan
All photos © HNEE 2022 / Christoph Raab


In May 2022, TRANSECT researchers and colleagues from Tajikistan conducted an extensive field data collection expedition in Jaloliddin Balkhi, a district in the south-west of Khatlon Region, Tajikistan. Data on crop types were gathered in order to derive a remote-sensing-based crop type map for the area of interest using these reference locations. For this, a classification algorithm will be applied to multi-temporal Sentinel-2 images in the future. As statistical data on agricultural land use are spatially aggregated, this will provide detailed information on the distribution and share of different crop types for our study region.

The agriculture of the district is mainly characterised by cotton cultivation, which is irrigated using a network of large and small channels. Secondary crops include wheat, alfalfa (also called lucerne), fruit trees and other minor crops. Salinisation phenomena sometimes occur, often related to difficulties with the drainage of the fields.



A similar field work expedition was conducted in June 2022 in Panfilov district, Kazakhstan. The field work was supported by our partner university, the Kazakh National University, named after Al-Farabi.


Analogous to our study region in Tajikistan, the agricultural land is irrigated by a branched network of large and small irrigation channels. Center-pivot irrigation systems are not widespread, but were found next to the town of Koktal, in the western part of the district. Agricultural land use is dominated by maize cultivation, often in crop rotation with alfalfa. Other crops, such as wheat or fruit trees play a subordinate role in terms of area shares. In some areas of the study region agricultural land use is challenged by reed, which can be described as tall, grass-like plants found in wetlands.


Share by: