Submitted by Iryna Dronova on
Resilience of Ukrainian agriculture is critical both domestically and globally, as agriculture serves as the country's largest export sector, positioning Ukraine among the top 10 agricultural producers and exporters worldwide, with far-reaching implications for global food security. Amidst the ongoing Russian war on Ukraine, much uncertainty prevails regarding the war’s lasting impact on Ukrainian agricultural production, including the abandonment of agricultural lands. This comprehensive study explores the impact of armed conflicts on Ukrainian agriculture by interlinking two vital research components.
The first component focuses on identifying indicators of agricultural abandonment during wartime in Ukraine. We tackle the complexities of this spatially and temporally diverse phenomenon, which is challenging to detect and map due to various drivers influencing abandonment during conflicts. Leveraging unsupervised classification of Landsat imagery, we construct phenological profiles using Normalized Difference Vegetation Index (NDVI) time series for agricultural and non-agricultural land cover types. Through extensive analysis, we aim to detect and map distinct temporal forms of abandonment during the 2014 conflict in Donbas and the present time.
In the second component, we investigate the factors influencing agricultural resilience in Ukraine amidst armed conflicts. Utilizing historical and near-real-time data, we analyze the impact of war, socioeconomics, climate, and agricultural intensification on critical resilience indicators, including agricultural yield, productivity, and abandonment. Employing fixed effects linear regression models, we assess different time periods: pre-Russian invasion (before 2014), the initial wave of violence and invasion (2014-2015), and the most recent invasion (2022 onwards).
This study aims to deepen our understanding of Ukrainian agriculture's resilience amidst armed conflicts. By incorporating historical data and employing remote sensing techniques to understand agricultural abandonment in both temporal and spatial dimensions, we gain valuable insights into the intricacies surrounding agricultural resilience. Our research will increase undertanding of Ukrainian agricultural resilience, and also inform the development of effective post-conflict recovery strategies in the agricultural sector. By ensuring the stability and sustainability of Ukrainian agriculture, our findings contribute to addressing the critical global concern of food security amidst the challenges posed by conflicts.
- The undergraduate will primarily assist graduate and undergraduate labmates with data collection and analysis needs that include but are not limited to (1) analyzing satellite imagery and auxiliary data for agricultural fields with high likelihood of abandonment; (2) identify potential drivers of abandonment of fields from the imagery and literature review; (3) data cleaning and organization for a geospatial analysis of agricultural resilience; and (4) prepare training and testing polygons for machine learning applications.
- The student will be expected to prepare updates for the research team and a written report of findings by the end of the semester.
- The student is also expected to have a check-in with Sarah Hartman or the entire research team at least every two weeks throughout the semester.
- Dr. Iryna Dronova and PhD Candidate Sarah Hartman will serve as the primary mentors for the student.
- This lab uses a hybrid work environment. Some meetings and work can be done remotely while others will be in-person.
- Previous experience in ArcGIS, Python, Google Earth Engine, and/or JavaScript is desirable
- Willingness to learn new software and improve GIS and coding skills
- Familiarity with Microsoft Office and Google Drive products
- Ability to work independently
- Communication skills
- Desire to use and advocate for open-access tools to inform the decision-making of current events
- Note: You do not need to be familiar with Google Earth Engine prior to starting this project. Previous experience with JavaScript, Python or another coding language is desirable, but we can also help interested candidates get started with this if it is their first time. The project scope will be determined based on the student’s interest and coding experience.