Satellite Data Shows Resilience of Tigrayan Farmers in Crop Cultivation During Civil War

TL;DR

The paper addresses how the Tigray War impacted cropland area, using a remote-sensing–driven workflow that combines LSTM-based cropland mapping (via OpenMapFlow), four-class change detection, and unbiased, sample-based area estimation anchored to ACLED conflict data. It reports cropland areas of $1{,}132 \pm 133$ kha in 2020 and $1{,}217 \pm 132$ kha in 2021, with modest gains ($70{-}160$ kha) and losses ($19{-}79$ kha), and finds a slightly higher potential cropland loss within a 5 km conflict buffer ($0{-}3\%$) than outside ($0{-}1\%$). The study demonstrates that Tigrayan smallholders largely sustained cultivation during wartime and showcases a transparent, reproducible methodology that leverages satellite time-series, ground-truth labeling, and unbiased inference to monitor food security in inaccessible regions. The approach offers timely, evidence-based insights for humanitarian planning and highlights the value of remote sensing and statistical estimation in conflict settings.

Abstract

The Tigray War was an armed conflict that took place primarily in the Tigray region of northern Ethiopia from November 3, 2020 to November 2, 2022. Given the importance of agriculture in Tigray to livelihoods and food security, determining the impact of the war on cultivated area is critical. However, quantifying this impact was difficult due to restricted movement within and into the region and conflict-driven insecurity and blockages. Using satellite imagery and statistical area estimation techniques, we assessed changes in crop cultivation area in Tigray before and during the war. Our findings show that cultivated area was largely stable between 2020-2021 despite the widespread impacts of the war. We estimated $1,132,000\pm133,000$ hectares of cultivation in pre-war 2020 compared to $1,217,000 \pm 132,000$ hectares in wartime 2021. Comparing changes inside and outside of a 5 km buffer around conflict events, we found a slightly higher upper confidence limit of cropland loss within the buffer (0-3%) compared to outside the buffer (0-1%). Our results support other reports that despite widespread war-related disruptions, Tigrayan farmers were largely able to sustain cultivation. Our study demonstrates the capability of remote sensing combined with machine learning and statistical techniques to provide timely, transparent area estimates for monitoring food security in regions inaccessible due to conflict.

Figures (20)

• Figure 1: Map and timeline of conflict events from ACLED database
• Figure 2: Process diagrams for creating annual cropland maps and computing area estimates. Map images are for illustration purposes and do not represent actual results.
• Figure 3: Left: 12-month time series of Sentinel-2 NDVI for ground-truth plots annotated as cultivated ("Crop"), fallow with weeds ("Fallow Weed"), or fallow with no vegetation at all ("Fallow None"). Solid lines depict the mean and shaded regions depict the standard deviation of the time series for all samples. Right: Visualizations of example ground-truth plots in PlanetScope September 2021 basemap. The plot labeled "Fallow with weeds or grass" has a similar appearance to fields labeled with crop types, while the field labeled "Fallow (no vegetation at all)" is easier to distinguish. Note that our map classification used Sentinel-2 inputs (left-most time series) while our response design used PlanetScope data for photointerpretation of reference sample labels (similar to those shown on the right).
• Figure 4: True positive rate (TPR) versus false positive rate (FPR) for range of thresholds used for post-processing algorithm, evaluated for the ground-truth dataset and validation dataset for Tigray in 2021.
• Figure 5: Annual cropland and change maps overlaid with zone boundaries (solid lines) and 5 km buffer around conflict events (dashed lines).