Sitian Xiong

Dr. Sitian Xiong


Research Focus & Metrics

Academic Career Length

5Years (2022-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
30 January 2025 in Remote Sensing
Verified
Journal Article
Open access
PDF file
SKSam Khallaghi
Sam Khallaghi
RARahebeh Abedi
Rahebeh Abedi
HAHanan Abou Ali
Hanan Abou Ali
HAHamed Alemohammad
Hamed Alemohammad
MAMary Dziedzorm Asipunu
Mary Dziedzorm Asipunu
LELyndon Estes
Lyndon Estes
Mapping agricultural fields using high-resolution satellite imagery and deep learning (DL) models has advanced significantly, even in regions with small, irregularly shaped fields. However, effective DL models often require large, expensive labeled datasets, which are typically limited to specific years or regions. This restricts the ability to create annual maps needed for agricultural monitoring, as changes in farming practices and environmental conditions cause domain shifts between years and locations. To address this, we focused on improving model generalization without relying on yearly labels through a holistic approach that integrates several techniques, including an area-based loss function, Tversky-focal loss (TFL), data augmentation, and the use of regularization techniques like dropout. Photometric augmentations helped encode invariance to brightness changes but also increased the incidence of false positives. The best results were achieved by combining photometric augmentation, TFL, and Monte Carlo dropout, although dropout alone led to more false negatives. Input normalization also played a key role, with the best results obtained when normalization statistics were calculated locally (per chip) across all bands. Our U-Net-based workflow successfully generated multi-year crop maps over large areas, outperforming the base model without photometric augmentation or MC-dropout by 17 IoU points.

30 September 2022 in Remote Sensing
Verified
Journal Article
Open access
PDF file
SXSitian Xiong
Sitian Xiong
PBPriscilla Baltezar
Priscilla Baltezar
MCMorgan Crowley
Morgan Crowley
MCMichael Cecil
Michael Cecil
SCStefano Crema
Stefano Crema
LELyndon Estes
Lyndon Estes
Cropland expansion is expected to increase across sub-Saharan African (SSA) countries in the next thirty years to meet growing food needs across the continent. These land transformations will have cascading social and ecological impacts that can be monitored using novel Earth observation techniques that produce datasets complementary to national cropland surveys. In this study, we present a flexible Bayesian data synthesis workflow on Google Earth Engine (GEE) that can be used to fuse optical and synthetic aperture radar data and demonstrate its ability to track agricultural change at national scales. We adapted the previously developed Bayesian Updating of Land Cover (Unsupervised) algorithm (BULC-U) by integrating a shapelet and slope thresholding algorithm to identify the locations and dates of cropland expansion and implemented a tiling scheme to allow the processing of large volumes of imagery. We apply this approach to map annual cropland change from 2000 to 2015 for Zambia (750,000 km2), a country that is experiencing rapid growth in agricultural land. We applied our cropland mapping approach to a time series of unsupervised classifications developed from Landsat 5, 7, 8, Sentinel-1, and ALOS PALSAR within 1476 tiles covering Zambia. The annual cropland changes maps reveal active cropland expansion between 2000 to 2015 in Zambia, especially in the Southern, Central, and Eastern provinces. Our accuracy assessment estimates that we have identified 27.5% to 69.6% of the total cropland expansion from 2000 to 2015 in Zambia (commission errors between 6.1% to 37.6%), depending on the slope threshold. Our results demonstrate the usefulness of Bayesian data fusion and shapelet, slope-based thresholding to synthesize optical and synthetic aperture radar for monitoring agricultural changes in situations where training data are scarce. In addition, the annual cropland maps provide one of the first spatially continuous, annually incremented accounts of cropland growth in this region. Our flexible, cloud-based workflow using GEE enables multi-sensor, national-scale agricultural change monitoring at low cost for users.

1 March 2026 in Environmental Research: Food Systems
Verified
Journal Article
Open access
SXSitian Xiong
Sitian Xiong
PHProtensia Hadunka
Protensia Hadunka
BMBrian P Mulenga
Brian P Mulenga
KBKathy Baylis
Kathy Baylis
PRPhilippe Rufin
Philippe Rufin
LELyndon Estes
Lyndon Estes
Farm size is a key characteristic of agricultural systems, closely linked to farming and land management practices. However, data on farm size are often difficult to obtain, especially in smallholder-dominated regions, and where available, they are rarely spatially explicit. This study examines the use of field size as a proxy for farm size, aiming to identify field size thresholds that can distinguish farms of different sizes. Using household-level data from Zambia’s Crop Forecast Survey, we applied both continuous and categorical methods to quantify the field-to-farm size relationship. Non-parametric Theil–Sen regression revealed a generally linear relationship between log-transformed field and farm sizes, particularly at the survey enumeration area scale, where farm size is proportional to field size (slope ≈ 1) and field size alone predicts farm size with a median bias of 0.017 ha. A Gaussian Naive Bayes classifier was also developed to assign fields to three categories: very small ( A : < 2 ha), small ( B : 2–5 ha), and medium ( C : 5–20 ha), and a coarser two-class scheme ( S : < 5 ha; M : 5–20 ha) that has been applied in previous studies. The Gaussian model identified field size thresholds at 0.58–0.6 ha ( A–B ) and 1.18–1.69 ha ( B–C ), depending on the prior farm size distribution. For the S–M scheme, thresholds ranged from 0.96–1.4 ha. Using these thresholds, national classification achieved F 1 scores of 0.78, 0.59, and 0.77 for A , B , and C , respectively, with B more prone to misclassification. The two-class model achieved 0.87 and 0.88 for S and M , indicating more robust classification performance under a coarser scheme. Kernel density analysis shows that field-size distributions within farm categories and the separability of these distributions, vary across regions, with the underlying factors responsible for these differences remaining an open question for future research. These results demonstrate the potential of field size as a spatial proxy for farm size to support remote sensing based farming system classification in data-scarce regions.


1 March 2026 in Environmental Research: Food Systems
Verified
Journal Article
Open access
JWJunren Wang
Junren Wang
MKMegan Konar
Megan Konar
SXSitian Xiong
Sitian Xiong
Ensuring the reliability of domestic food supplies is increasingly critical given the compounding shocks of production variability, transportation constraints, and demand surges. This study develops a spatially explicit framework for Zambia, introducing the food reliability index to quantify the probability that food requirements exceed supply under stochastic conditions. By integrating probabilistic production estimates with a gravity model of domestic trade, we evaluate reliability dynamics across rural and urban landscapes. We apply this framework to establish a 2020 baseline and project future vulnerabilities to 2050. To address projected deficits, we simulate intervention scenarios, comparing the efficacy of cropland expansion, yield intensification, and transportation cost reductions. The resulting framework provides decision-makers with a tool to identify spatial imbalances and prioritize investments for enhancing food supply resilience.