GROUNDBREAKING   

                                         EARTH OBSERVATION / GIS

NATURAL RESOURCE MANAGEMENT     

 

oil block 5A - MASSIVE DISPLACEMENT IN TRADITIOnal land use

 

- The oil fields of Sudan / Southern Sudan -

 

Summary of Prins peer reviewed research results - covering latest advances and monitoring system

 

Background and Introduction

In 2006, Prins was commissioned by the European Coalition on Oil in Sudan (ECOS) to map land-use changes in Oil Block 5A, focusing on 1999–2003. Previously, he conducted similar analyses in Blocks 3 and 7, always using Landsat imagery—a continuous, multi spectral archive dating back to the mid 1980s. Landsat’s seven spectral bands (visible and infrared) reveal crop types, soil moisture, biomass, rangeland condition, and more (for a lay overview, see NASA’s Landsat summary). Over 25 years of work in Sudan and South Sudan—including uncovering the Darfur crisis—has demonstrated Landsat’s power to detect large scale human impacts with local precision.

 

 

 

Evolution of the 5A Study

 

·         2006–2010: Initial ECOS report used unsupervised (ISODATA) classification of early dry‑season Landsat images to map traditional agro‑pastoral footprints. Change detection matched ground reports of village attacks and mass displacement.

 

·      2013: Invited keynote at Swedish Space Days introduced verification using Very High Resolution (VHR) imagery from Google Earth (QuickBird‑2).

 

·      2013–2015: New humanitarian crisis generated a wealth of VHR data, enabling systematic validation and refinement of Landsat methods.

 

·      2017: Publication in the International Journal of Remote Sensing documented a robust, peer‑reviewed monitoring system.

Below is a concise overview of how the traditional agro pastoral land use informs our satellite derived change patterns. The detailed text is intended for specialists or engaged laypersons; casual readers can focus on the figures and their captions.

 

Traditional Agro Pastoral Land Use

1. Seasonal Floodplain Environment
  o Located in a flat, flood prone basin crossed by the White Nile and Sudd wetlands.
  o Populated by Nuer and Dinka agro pastoralists, where cattle outnumber people.


2. Homestead & Cropping Cycle
  o Settlements (mud hut “tukuls”) sit on natural levees just above floodwaters.
  o Wet season (May–Nov): People and cattle concentrate on homesteads; crops grow and are harvested by early dry   season.
  o Dry season (Nov–May): Cattle range up to 50 km/day seeking fresh grasses; homestead fields lie fallow.


3. Satellite Footprint
  o Post harvest, bare soil reflectance around homesteads creates a bright “footprint” in early dry season imagery.
  o Absence of that bright signature year to year indicates displacement of people and cattle.

 

Figures

Fig 1: Landsat 27 Nov 1999 (Bands 3,2,1) showing natural colors: lush vegetation (green) vs. farming footprints  (whitish). Overlay of towns, watercourses (blue), oil block border, roads, seismic grids (yellow).


Fig 2–5: On ground photos of tukuls, cattle (abigar), and grazing impacts demonstrating how degradation appears in multispectral data.


Fig 6: Series of early dry season Landsat images (1994–2004) illustrating shifting farming concentrations along sandy levees.


Fig 7–8: Examples of how transhumance movement and bushfires can obscure—but intense activity still reveals—human and livestock presence.

 

Key Takeaways

Robust Change Detection: Consistent patterns across unsupervised classification, VHR verification, and change detection align closely with mass displacement reports.


Seasonal Timing: Early dry season (November) images maximize contrast between human footprint and lush regrowth, minimizing bushfire noise.


Method Validity: Peer reviewed methods achieved high mapping accuracy, with conservative thresholds ensuring certainty.

 

Explore More

• Full reports, datasets, and publications are available via the buttons below.


• Engage with our front end Earth Observation tools documenting human rights applications.
 

 

Figure 1. Landsat natural-color image (Bands 3,2,1) of Oil Block 5A (27 Nov 1999). The flat, seasonally flooded plain of South Sudan is bisected by the White Nile. Lush vegetation appears green; exposed soil from intensive cattle farming appears whitish. Major towns are labeled, watercourses in blue, and the eastern block boundary outlined. Seismic survey grids are digitized in yellow. Image predates construction of the all-weather road from Leer via Thar Jath to Bentiu.

Files, reports, links and pdf are found in the button of this page

 

The traditional land use

The study area lies adjacent to the White Nile and the Sudd wetlands, a flat landscape characterized by seasonal inundation. Grasslands and wooded savanna dominate, managed by Nuer and Dinka agro-pastoralists. Much of southern Sudan consists of floodplains and marshes ideally suited to their transhumant lifestyle, in which cattle outnumber people.

  • Settlements: Permanent villages—clusters of circular mud huts (“tukuls”) with thatched roofs—sit on natural river levees just above flood levels. Often, cattle huts (larger than human dwellings) form part of these homesteads, which lie along streams.
  • Wet Season (May–Nov): Families and their herds remain at these elevated homesteads, growing and harvesting crops by November.
  • Dry Season (Nov–May): Herds range widely (30–100 km/day) into floodplain and riverine wetlands for fresh grasses.

Grazing around homesteads exposes soil and degrades vegetation, but most areas recover each rainy season—except where overgrazing or permanent structures impede regrowth. In this study, severe degradation was limited, confined to some settlement centers (e.g., Boaw).

 

 

Figure 2. High concentration of Nuer cattle (“abigar”) grazing near homesteads. Nuer pastoralists—also called “Abigar people”—live in conical thatch-roof huts (tukuls) visible in the background.

 

 

Figure 3. Cattle grazing footprint in multispectral imagery. Repeated grazing exposes soil and degrades vegetation, producing a bright signature easily detectable in Landsat data. (Photo credit: Peter Lawson)

 

Figure 4. Seasonal cattle movement to wetlands. During the dry season, livestock range into major floodplain areas near rivers to access fresh grasses. (Photo credit: Nakachew Minuye)

Figure 5. Typical village on natural levee. Aerial view shows tukuls clustered on slightly elevated sandy banks. Bright grayish patches are recently tilled fields; yellowish tones indicate dry grass or fallow land. (Photo credit: Sharon Hutchinson)

 

Monitoring Agro-Pastoral Land Use from Space

Data Source & Timing

  • Landsat 5 & 7 (16-day revisit) provided multispectral imagery since the mid-1980s.
  • Early dry-season images (November) offer the clearest contrast: bare-soil signatures from post-harvest cropping and prolonged cattle grazing appear as bright “footprints,” while lush regrowth still blankets surrounding grasslands.

Detection Logic

  • If a homestead’s bright footprint disappears or weakens in the subsequent year’s November image, it indicates that people and cattle were absent—evidence of displacement.
  • Cloud-free, minimally noisy images are rare; for 1998–2002, only one or two suitable November scenes per year were available for change-detection comparisons.

Spatial Dynamics

  • Figure 6 (125 × 128 km, Bands 3,2,1, 1994–2004): November images highlight green vegetation vs. bright bare soils. Close-up (h) shows homesteads on elevated sandbanks (cf. Figure 5). Vertisols appear as dark, moist soils.
  • Dry-season challenges: Transhumance and seasonal bush fires (Figure 7) can mask grazing footprints. However, intense herding activity still produces bright signatures (Figure 8) that override fire scars.

Why Landsat?

  • Multispectral sensitivity to soil reflectance detects even light grazing, unlike visible-only VHR or aerial imagery (Tueller 1989; Booth & Tueller 2003).
  • Consistent archives from 1984 onward allow reproducible, large-area time-series analysis.
  • 30 m resolution may miss small features but excels at capturing broad land-use patterns unobservable in higher-resolution yet single-band images.

Analytical Workflow (Prins 2009)

  1. Image Selection: Choose least cloudy November scenes (1987–2006).
  2. Classification: Apply unsupervised ISODATA (60 classes) in ERDAS Imagine Pro; merge the two brightest classes to delineate the traditional land-use footprint.
  3. Change Detection: Generate Boolean maps to visualize spatial and temporal patterns of expansion or disappearance of the footprint, corroborating displacement events reported on the ground.
  4. Verification: Exclude cloud-contaminated pixels (minimal impact) and confirm patterns with independent ground reports (Christian Aid, HRW, ECOS).

This approach revealed a 72–99% reduction in traditional land use across focus areas between 1999 and 2003, fully consistent with documented mass displacements. No other plausible explanation emerges from the Landsat archive.

 

Figure 6. Time series of early-dry-season Landsat images (1994–2004 - higher resolution image here ). Natural-color frames (Bands 3,2,1) over a 125 × 128 km area. Close-up (h) highlights homesteads on elevated sand banks—the same pattern seen in Figure 5.

 

Figure 7a. Kuac settlement area at the onset of dry season. Bush-fire scars (dark brown) partially mask light grazing footprints, but concentrated cattle tracks (bright) still reveal livestock presence. A dried stream bed with herding tracks is visible north of Kuac.

Figure 7b: Example of how farming activity will overrule the effect of bush fires. This case show an area beteen Touc and Wumlit where active farming activity neutralize the effectie of an bushfire scare within 8 days. This means, if recent farming activity is presence it will overrule the bushfire effect and will e.g., be captured by a ISO-classifier.

Figure 8. Dry-season change (1995–2002). Four Landsat scenes show consistent bright farming footprints in 1995 (a) and 2000 (b). By Mar 2002 (c), central footprints shrink while northeastern and southeastern areas expand. By Dec 2002 (d), the farming signature has shifted south. (higher resolution image here).

Figure 9. VHR vs. Landsat comparison. Left: QuickBird-2 (0.5 m) natural-color image showing huts and fields. Right: Landsat false-color composite (Bands 7,5,3) with vegetation in green, bare fields in whitish, and wet areas dark bluish. The red overlay on the Landsat panel indicates the VHR coverage.

Figure 10. Tasseled Cap transformation of Figure 9 area. Derived brightness, greenness, and wetness layers highlight biophysical contrasts—bare fields appear dark and dry compared to surrounding vegetation in the scenario. The transect line shows spectral response across the farm field.

Figure 11. Unsupervised ISO classification (60 classes) for late 1994–early 1995. Two classes capture the seasonal land-use footprint, illustrating consistent high concentrations of farming activity before oil development. Settlement patterns match 1980s Russian maps.

 

Figure 12. Sequence of ISODATA classifications (late 1999–early 2003). A) Traditional land use concentrated in Nhialdiu. B) Early dry-season expansion after reported attacks along the oil road. C) Post-attack shift southwest toward Wicok and Chotchar. D) Peak farming footprint in south-southwest, matching reported displacement. E–F) Continued footprint in the southwest amid seasonal bush fires.

 

Figure 13. Questionnaire-based displacement maps (Feb–Mar 2002). Local interviews mapped village clearances, escape routes, and refugee destinations.  The maps also included escape routes the places where they ended up. In summery, it showed a displacement from Nhialdiu area and most people were first driven south west and then south.

 

 

Figure 14. Focus areas for change detection. Three zones within Oil Block 5A—northern oil-road buffer (2.5 km), Nhialdiu vicinity, and southern displacement region—digitized from multiple ground-report sources (e.g., HRW 2003).

Figure 15. Combined Landsat change and ground-report events (1999–2004). Yellow explosions mark attacked settlements; green triangles indicate refugee camps. Change-map statistics confirm reported desertions by end 2002 and increased land-use activity in southern refugee zones.It can further be seen that at the south end of the oil road - the area around settlement Bieh show depopulation, which also have been reported (ECOS 2002; HRW 2003; Sudanreevs;  Relifeweb 2002) to be attacked several times during early 2002.

 

Figure 16. Oil infrastructure development (Oct 2006). Digitized roads, drilling sites, and seismic lines over Landsat Band 7 (wetness-sensitive) reveal hydrological disruption by elevated all-weather roads near Thar Jath.More detailed image that show hydrological disturbance close to Thar Jath be viewed in the here and in the Prins 2009 report.

 

 

Figure 17. Integrated scenario of displacement and oil development. Brown areas were active farmland in 2000 but abandoned by late 2002; bright green areas show new farming in late 2002. Overlaid are reported attack locations and 2006 oil infrastructure.

 

Validation with Very-High-Resolution Imagery (Post-2010)

With the advent of freely available Google Earth VHR images (~0.5 m resolution), the Landsat-based classifications could be evaluated at much finer scales:

  1. Training & Verification:
    • Identified tukuls, farm fields, and degraded patches in VHR scenes (e.g., 2003, 2006).
    • Used these ground-truth points to verify November Landsat classifications at 0.5–2 ha Minimum Mapping Units.
  2. Accuracy Assessment (Table 1):
    • ISODATA (unsupervised): Overall accuracy > 97 %, Kappa 0.81–0.86.
    • CTA (Classification Tree Analysis): Overall accuracy > 99 %, Kappa 0.96–0.99.
    • Commission errors in ISODATA reflect conservative 10 % pixel thresholds rather than misclassification of large areas.
  3. Error Matrix Metrics:
    • Omission and commission errors quantified against VHR reference.
    • Kappa statistic employed per Fagan & De Fries (2009) to gauge classification agreement.

 

Figure 18. VHR farming area in late dry season. Circular tukuls and adjacent degraded land (digitized for training) appear as very bright patches. Dark gray zones indicate post-rainy-season bush-fire scars.

 

Figure 19. VHR-based accuracy assessment (Stockholm 2013 presentation). Google Earth VHR images and settlement databases (yellow dots) support verification of Landsat classifications at 0.5–2 ha scales. Insets: white = correct detections; green = commission errors; red = omission errors.

 

Table 1. Accuracy assessment of classification methods using QuickBird-2 VHR as reference. ISODATA (unsupervised) and CTA (supervised tree classifier with indices) both achieve high overall accuracy and Kappa statistics (per Fagan & De Fries 2009). CTA yields very fine-scale mapping at settlement level given the chosen MMU (Minimum Mapping Unit).

 

Advanced Time-Series Methods

Beyond categorical maps, narrow inter-calibration of Landsat bands enables continuous change detection:

  • Blue-band differencing (Band 1, 450 nm):
    • Calibrated across years to isolate subtle soil-reflectance changes.
    • Figure 20 illustrates 1999 vs. 2002 blue-light anomalies, highlighting emerging bare-soil areas.
  • Monitoring System Development:
    • Continuous-value time series support automated detection of gradual degradation or sudden displacement.
    • Potential for real-time alerts in humanitarian crises and rangeland management.

 

 

Further advances and details can be viewed in the Prins (2018) and the peer-reviewed summery. 

Figure 20. Blue-band change detection (1999 vs. 2002). Landsat Band 1 (450 nm) is narrowly calibrated across years to isolate bare-soil reflectance changes, clearly delineating shifts in farming extents.(see more in scientific summery).

 

Conclusion

Overall the results of the satellite mapping (Prins 2009; Stockholm 2013 presentation) fully collaborate and detail the ground reports of mass displacement of people in oil block 5a during 1999-2003.

  • Strong concordance between Landsat-derived land-use change and independent ground reports of mass displacement (1999–2003).
  • Early dry-season imagery (November) consistently captures maximum human-activity footprints; dry-season dynamics (transhumance, fires) introduce interpretive challenges that intensive grazing can still overcome.
  • Unsupervised ISODATA classification proved highly effective for large-scale pattern mapping; CTA further refines results at finer resolutions.
  • VHR validation confirms Landsat’s capacity to detect even light grazing signatures that may elude aerial or single-band analysis.
  • Multi-decadal Landsat archives—combined with modern VHR cross-checks—provide objective, reproducible evidence of human displacement and ecosystem change.
  • Future applications include population-level displacement estimates, rangeland health monitoring, and rapid-response mapping in conflict or disaster zones.

 

 

Supplements:

 

PRINS KEY-NOTE SPEAKER AT THE SWEDISH SPACE DAYS 2013: Global reporting interpretation of speach

 

  pdf  PRESENTAION OF THE 5A CASE AND ANALYSIS - INVITED TALK REMOTE SENSING DAYS STOCKHOLM 8 APRIL 2013

pdf  REPORT TO ECOS 2009: Satellite mapping of land cover and use in relation to Oil exploitation in concession 5A

 

Client of the 2009 report: The European Coalition on Oil in Sudan (ECOS)

 

Hit Counter

 

[1] In general, an AUC of 0.5 suggests no discrimination (i.e., ability to predict farming activity), 0.7 to 0.8 is considered acceptable, 0.8 to 0.9 is considered excellent, and more than 0.9 is considered outstanding.

 

References:

Aljazeera. 2015. Accessed 26 September 2016. http://www.aljazeera.com/indepth/features/2015/ 05/south-sudan-man-catastrophe-150525120843494.html

Bastin, G. N., G. Pickup, V. H. Chewings, and G. Pearce. 1993. “Land Degradation Assessment in Central Australia Using a Grazing Gradient Method.” The Rangeland Journal 15 (2): 190–216. doi:10.1071/RJ9930190.

Booth, D.T., and P.T. Tueller. 2003. Rangeland monitoring using remote sensing. Arid Land Research and Management 17 (4): 455–467.

Christian Aid. 2002. Hiding between the Streams: The War on Civilians in the Oil Region of Southern Sudan. http://www.christian-aid.org.uk/indepth/0205suda/sudanrpt.pdf

de Guzman, D., 2002. Depopulating Sudan’s Oil Regions. European Coalition on Oil in Sudan (ECOS). http://www.ecosonline.org/back/pdf_reports/2002/depopulating%20sudans%20oil%20regions.pdf

Fagan, Matthew, and Ruth DeFries. "Measurement and monitoring of the world’s forests: A review and summary of remote sensing technical capability, 2009–2015." UMBC Geography and Environmental Systems Department Collection (2009).

Google Open Street Maps. 2015. “The Humanitarian Openstreetmap Team.” https://www.hotosm. Org

Holling, C. S. 1973. Resilience and stability of ecological systems. Annual Review of Ecology and Systematics 4:1-23

HOT. 2016. Project - South Sudan, Unity State, Bentiu. The Humanitarian OpenStreetMap Team. http:// tasks.hotosm.org/project/396

HRW. 2003. “Human Rights Watch (November 2003). Sudan, Oil and Human Rights. 754 p. Accessed 26 September 2016. http://www.hrw.org/reports/2003/sudan1103/9.htm#_ftn88#_ftn88

HRW. 2015. “’They Burned It All’ Destruction of Villages, Killings, and Sexual Violence in Unity State, South Sudan: Report.” 42 p. Accessed 26 September 2016. https://www.hrw.org/sites/default/ files/report_pdf/southsudan0715_web_0.pdf

James, C. D., J. Landsberg, and S. R. Morton. 1999. “Provision of Watering Points in the Australian Arid Zone: A Review of Effects on Biota.” Journal of Arid Environments 41 (1): 87–121. doi:10.1006/jare.1998.0467.

Lemenkova, Polina. "ISO Cluster classifier by ArcGIS for unsupervised classification of the Landsat TM image of Reykjavík." Bulletin of Natural Sciences Research 11.1 (2021): 29-37.

Maynard, C. L., R. L. Lawrence, G. A. Nielsen, and G. Decker. 2007a. “Ecological Site Descriptions and Remotely Sensed Imagery as a Tool for Rangeland Evaluation.” Canadian Journal of Remote Sensing 33 (2): 109–115. doi:10.5589/m07-014.

Maynard, C. L., R. L. Lawrence, G. A. Nielsen, and G. Decker. 2007b. “Modeling Vegetation Amount Using Bandwise Regression and Ecological Site Descriptions as an Alternative to Vegetation Indices.” GIScience & Remote Sensing 44 (1): 68–81. doi:10.2747/1548-1603.44.1.68.

Olofsson, P., G. M. Foody, M. Herold, S. V. Stehman, C. E. Woodcock, and M. A. Wulder. 2014. “Good Practices for Estimating Area and Assessing Accuracy of Land Change.” Remote Sensing of Environment 148: 42–57. doi: 10.1016/j.rse.2014.02.015.

Phillips, S. J., R. P. Anderson, and R. E. Schapire. 2006. “Maximum Entropy Modeling of Species Geographic Distributions.” Ecological Modelling 190 (3): 231–259. doi:10.1016/j. ecolmodel.2005.03.026.

Pontius, R. G. Jr, and M. Millones. 2011. “Death to Kappa: Birth of Quantity Disagreement and Allocation Disagreement for Accuracy Assessment.” International Journal of Remote Sensing 32 (15): 4407–4429. doi:10.1080/01431161.2011.552923

Erik Prins (2018) Landsat approaches to map agro-pastoral farming in the wetlands of southern Sudan, International Journal of Remote Sensing, 39:3, 854-878, DOI: 10.1080/01431161.2017.1392634

Samain, O., L. Kergoat, P. Hiernaux, F. Guichard, E. Mougin, F. Timouk, and F. Lavenu. 2008. “Analysis of the in Situ and MODIS Albedo Variability at Multiple Timescales in the Sahel.” Journal of Geophysical Research: Atmospheres 113: D14. doi:10.1029/2007JD009174

Stehman, S. V. 2004. “A Critical Evaluation of the Normalized Error Matrix in Map Accuracy Assessment.” Photogrammetric Engineering & Remote Sensing 70 (6): 743–751. doi:10.14358/ PERS.70.6.743

Tueller, P. T. 1989. “Technology for Rangeland Management.” Invited Synthesis Paper 42 (6): 442.

UNHRC. 2016. “UN Human Rights Council.” Assessment mission by the Office of the United ations High Commissioner for Human Rights to improve human rights, accountability, reconciliation and capacity in South Sudan: detailed findings, 10 March 2016, A/HRC/31/CRP.6. Accessed 26 April 2016. http://www.refworld.org/docid/56e2ee954.html

UNMISS 2015: UN Mission to South Sudan. 2015. “2015, Flash Human Rights Report on the Escalation of Fighting in Greater Upper Nile.” April/May, June 29. Accessed 26 April 2016. http://www.refworld.org/docid/5592995c4.html

Unosat. 2016. “UNOSAT LIVE Map for South Sudan Complex Emergency (CE20131218SSD).” Accessed 26 September 2016. https://unosat.maps.arcgis.com/apps/webappviewer/index.html id=cac78381476c43ebb98a7e642211809d

UNOSAT. 2015. “Rapid Damage Assessment of the Area of Rapid Damage Assessment of the Area of Ngop Ngop, Unity State, South Sudan.” Accessed 26 September 2016. http://unosat-maps. web.cern.ch/unosat-maps/SS/CE20131218SSD/Ngop_UNOSAT_20150518.pdf

Washington-Allen, R. A., N. E. West, R. Douglas Ramsey, and R. A. Efroymson. 2006. “A Protocol for Retrospective Remote Sensing–Based Ecological Monitoring of Rangelands.” Rangeland Ecology & Management 59 (1): 19–29. doi:10.2111/04-116R2.1.

 

 

 

privacy@prins engineering.com

About Us    |    References    |    Services    |    CV's    |    Contacts    |    Press     Copyright © PRINS, Eng. 2026