EconPapers    
Economics at your fingertips  
 

Deep Learning and Remote Sensing for Restoring Abandoned Agricultural Lands in the Middle Volga (Russia)

Artur Gafurov () and Maxim Ivanov
Additional contact information
Artur Gafurov: Institute of Environmental Sciences, Kazan Federal University, Kazan 420097, Russia
Maxim Ivanov: Institute of Environmental Sciences, Kazan Federal University, Kazan 420097, Russia

Land, 2024, vol. 13, issue 12, 1-16

Abstract: Abandoned agricultural lands in the Middle Volga region of Russia, which appeared because of socio-economic transformations after the collapse of the USSR and the liquidation of collective farms, represent a significant potential for increasing agricultural production and economic development of the region. This study develops a comprehensive approach to assessing the suitability of these lands for return to agricultural turnover using machine learning methods and remote sensing data. Sentinel-2 satellite imagery and a deep neural network based on MAnet architecture with Mix Vision Transformer encoder (MiT-b5), which achieved an accuracy of 93.4% and an IoU coefficient of 0.84, were used for semantic segmentation of modern agricultural land. Land use dynamics since 1985 were analysed using Landsat 4–9 data, revealing significant areas of abandoned arable land. Land suitability was assessed, taking into account natural resource factors such as topography, soils and climatic conditions. The results showed that the total area of land suitable for reclaimed land is 2,014,845 ha, which could lead to an increase in wheat yield by 7.052 million tons. The potential cumulative net profit is estimated at 35.26 billion rubles (about US$352.6 million). The main conclusions indicate the significant economic and social potential of returning abandoned land to agricultural turnover, which requires a comprehensive approach that includes investment in infrastructure and the introduction of modern agro-technologies.

Keywords: abandoned lands; Middle Volga region; semantic segmentation; deep learning; remote sensing; vision transformer; land reclamation; agricultural turnover (search for similar items in EconPapers)
JEL-codes: Q15 Q2 Q24 Q28 Q5 R14 R52 (search for similar items in EconPapers)
Date: 2024
References: View references in EconPapers View complete reference list from CitEc
Citations:

Downloads: (external link)
https://www.mdpi.com/2073-445X/13/12/2054/pdf (application/pdf)
https://www.mdpi.com/2073-445X/13/12/2054/ (text/html)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:gam:jlands:v:13:y:2024:i:12:p:2054-:d:1533591

Access Statistics for this article

Land is currently edited by Ms. Carol Ma

More articles in Land from MDPI
Bibliographic data for series maintained by MDPI Indexing Manager ().

 
Page updated 2025-03-19
Handle: RePEc:gam:jlands:v:13:y:2024:i:12:p:2054-:d:1533591