Full Text Available
Note: Clicking the button above will open the full text document at the original institutional repository in a new window.
Multi-sensor remote sensing for slope stability monitoring in open-pit mining
Thesis (DPhil)--Stellenbosch University, 2025.
Saved in:
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Thesis |
| Published: |
Stellenbosch : Stellenbosch University
2025
|
| Subjects: | |
| Tags: |
No Tags, Be the first to tag this record!
|
| _version_ | 1867613836591235073 |
|---|---|
| access_status_str | Open Access |
| author | Haupt, Shelley |
| author2 | Engelbrecht, Jeanine |
| author_browse | Engelbrecht, Jeanine Haupt, Shelley |
| author_facet | Engelbrecht, Jeanine Haupt, Shelley |
| author_sort | Haupt, Shelley |
| collection | Thesis |
| dc_rights_str_mv | Stellenbosch University |
| description | Thesis (DPhil)--Stellenbosch University, 2025. |
| format | Thesis |
| id | oai:scholar.sun.ac.za:10019.1/132411 |
| institution | Stellenbosch University (South Africa) |
| last_indexed | 2026-06-10T12:42:28.529Z |
| license_str | Other — see source repository |
| provenance_str_mv | Harvested via OAI-PMH from SUNScholar — Stellenbosch University Repository |
| publishDate | 2025 |
| publishDateRange | 2025 |
| publishDateSort | 2025 |
| publisher | Stellenbosch : Stellenbosch University |
| publisherStr | Stellenbosch : Stellenbosch University |
| record_format | dspace |
| source_str | SUNScholar — Stellenbosch University Repository |
| spelling | oai:scholar.sun.ac.za:10019.1/132411 Multi-sensor remote sensing for slope stability monitoring in open-pit mining Haupt, Shelley Engelbrecht, Jeanine Kemp, Jaco Stellenbosch University. Faculty of Arts and Social Sciences. Dept. of Geography and Environmental Studies. Slopes (Soil mechanics) -- Stability -- Monitoring Strip mining -- Monitoring -- South Africa -- Limpopo Mines and mineral resources -- Remote sensing Open-pit mining -- South Africa -- Limpopo -- Palabora Multisensor data fusion UCTD Thesis (DPhil)--Stellenbosch University, 2025. Haupt, S. 2025. Multi-sensor remote sensing for slope stability monitoring in open-pit mining. Unpublished doctoral dissertation. Stellenbosch: Stellenbosch University [online]. Available: https://scholar.sun.ac.za/items/3e586101-6abc-4c36-8d0e-b0d336053149 ENGLISH ABSTRACT: Open-pit mines are some of the largest geotechnical structures in the world and are located in areas where the ore can be extracted near the surface. Strain and stress changes in the rock masses caused by mining activities can lead to movement of the slope and ultimately lead to slope failures. Slope failures have severe consequences such as damage to equipment, disruption to mining activities, economic implications and, in extreme cases, loss of lives. Therefore, slope monitoring programmes have been established to reduce the impact of slope failures in mining processes. Due to the risks involved, it is important to detect precursory movements before the slope failure and to make reliable time-to-failure projections. Conventional monitoring techniques, like visual inspections, instrumentation, and surveying, offer high-quality measurements but are typically limited to specific locations with predetermined risks. In contrast, remote sensing methods have recently been used for slope stability monitoring, providing continuous coverage over entire slopes with high measurement densities. The purpose of this study was to assess the contribution of multi-sensor remote sensing techniques dedicated to the detection and monitoring of slope instability in open-pit mines. The objective was to test if multi-sensor remote sensing techniques can be used to detect, measure and monitor slope instabilities with sufficient accuracy, reliability, repeat-interval, and resolution to replace or complement conventional slope stability monitoring techniques. Three remote sensing technologies were evaluated, including i) optical image cross-correlation using aerial photographs, ii) advanced time series interferometric synthetic aperture radar (InSAR) based on X-band and C-band data, and iii) satellite video data coupled with structure-from-motion algorithms for deriving a digital elevation model (DEM) for elevation model differencing. The image cross-correlation techniques were used to assess the use of aerial photographs to detect horizontal displacements at the Palabora open-pit mine in the Limpopo Province of South Africa where a slope failure occurred in 2004. The horizontal displacements in the north-south (N/S) and east-west (E/W) directions were extracted using the COSI-Corr algorithm and aerial imagery captured in 2009, 2012, 2015 and 2018. The results showed that ongoing movement was observed towards the centre of the pit with movement rates between 2.1 and 4.1 m/year. The limitation of this technique is that only the horizontal north-south (N/S) and east-west (E/W) components of the movement could be extracted and when observing movements of slopes, vertical components of movement are also expected, consequently, the real displacement magnitude will be underestimated. Therefore, while pixel cross-correlation techniques can be used to retrieve measurements of pit wall stability, the measurements will need to be complemented with technologies sensitive to the vertical component of movement (such as InSAR) or elevation model differencing) if a true representation of total movement magnitude and direction is needed. The second technique that was assessed was advanced time series InSAR which used the small-baseline-subset (SBAS) method to derive slope movements in an open-pit iron ore mine in the Northern Cape Province of South Africa. Both X-band and C-band SAR data were used. The results derived from the X-band data indicated that, between January and December 2017, the cumulative displacement was between 21.3 mm and 23.4 mm. For the C-band InSAR data, movement rates of between 47.8 and 49.4 mm/year with a cumulative displacement of up to 52.3 mm from January 2020 to May 2021 were measured. While InSAR is sensitive to small-scale movements the study revealed several limitations. Firstly, using X-band SAR data, low measurement point densities were observed due to high levels of noise resulting from dynamic surface conditions. Another limitation is the maximum detectable limits which are dependent on the wavelength of the sensor. The maximum detectable limit is 25.7 cm/year for X-band data acquired every 11 days and 42.7 cm/year for C-band data acquired at 12-day intervals. If slope movement velocities exceeding this limit are expected, higher revisit intervals or longer wavelength data is required. Finally, InSAR data only measures the component of the movement that is in the direction of the line-of-sight (LoS) of the sensor. Therefore, if any component of the movement vector is in a direction perpendicular to the LoS, the movement cannot be measured. This implies that horizontal north-south (N/S) movement cannot be measured. Where horizontal northward or southward movement is anticipated for a particular slope geometry, alternative monitoring techniques such as optical image cross-correlation should be used in such instances. An alternative to the direct measurement of surface displacement using cross-correlation or InSAR data is by observing the change in elevation of surfaces through DEM differencing approaches. To test the ability to extract high-accuracy DEMs from satellite video data, this research investigated the use of satellite video data captured by the SkySat-C constellation over a mine site in Australia. The satellite video data consisted of 25 individual frames of images, and structure-from-motion (SfM) algorithms were used to extract a DEM. The satellite video DEM was compared to a LiDAR-derived DEM and a root mean square Error (RMSE) of 3.1 m was obtained. Such errors from successive DEMs when used in an elevation model change detection approach may exceed and overprint smaller changes related to small-scale slope instability before collapse. Therefore, depending on the scale of precursory slope movements before slope collapse (only a few centimetres), satellite DEM data may not provide the precision necessary for early detection of slope movement. The results of these investigations could be used to evaluate the operational limitations of using a multi-sensor remote sensing approach for slope stability monitoring. While remote sensing can enhance the detection, measurement and monitoring of slope stability in open-pit mines, there are several limitations. In general, the long repeat intervals only allow weekly measurements in contrast to continuous measurements by some instruments. Nevertheless, one of the main advantages of using remote sensing techniques for open-pit mine slope stability monitoring is that they provide a synoptic view of the entire open-pit. This offers the ability to detect slope instabilities in uninstrumented areas or in areas where ground-based monitoring systems may not have been deemed to be a priority. The integration of various remote sensing techniques could provide valuable information to ensure a more robust slope stability monitoring programme. AFRIKAANSE OPSOMMING: Oopgroefmyne is van die grootste geotegniese strukture in die wêreld en is geleë in gebiede waar die erts naby die oppervlak ontgin kan word. Verhogings in spanning en druk in die rotsmassas wat deur mynaktiwiteite veroorsaak word, kan tot beweging van die helling van die put muur lei en uiteindelik in hellingstortings ontaard. Hellingstortings het ernstige gevolge soos skade aan toerusting, onderbreking van mynaktiwiteite, ekonomiese implikasies en, in uiterste gevalle, verlies aan lewens. Daarom is hellingmoniteringsprogramme ingestel om die impak van hellingstortings in mynbouprosesse te verminder. Vanweë die risikos betrokke is dit belangrik om voorafgaande bewegings op te spoor voordat die helling instort en om betroubare tyd-tot- ineenstorting projeksies te maak. Konvensionele moniteringstegnieke, soos visuele inspeksies, instrumentasie en opmeting, bied hoë kwaliteit metings, maar is gewoonlik beperk tot spesifieke plekke met voorafbepaalde risiko's. In teenstelling hiermee is afstandwaarnemingstegnieke onlangs gebruik vir hellingstabiliteitsmonitering, wat deurlopende dekking oor hele hellings met hoë metingsdigthede bied. Die doel van hierdie studie was om die bydrae van multi-sensor afstandwaarnemingstegnieke tot moniteringsprogramme wat toegewy is aan die opsporing en monitering van helling-onstabiliteit in oopgroefmyne te evalueer. Die doelwit was om te toets of multi-sensor afstandwaarnemingstegnieke gebruik kan word om hellingonstabiliteite op te spoor, te meet en te monitor met voldoende akkuraatheid, betroubaarheid, herhaalinterval en resolusie om konvensionele hellingstabiliteitmoniteringstegnieke te vervang of aan te vul. Drie afstandwaarnemingstegnologieë is geëvalueer: i) optiese beeldkruiskorrelasie met behulp van lugfotos, ii) gevorderde tydreeks-interferometriese sintetiese apertuur radar (InSAR) gebaseer op X-band en C-band data, en iii) satellietvideo-data gekoppel aan Struktuur-uit-Beweging algoritmes om 'n DEM (Digitale Elevasie Model) af te lei vir DEM verskil analise. Die beeldkruiskorrelasietegnieke is gebruik op lugfotos om te evalueer of horisontale bewegings by die Palabora oopgroefmyn in die Limpopo-provinsie van Suid-Afrika, waar 'n hellingstorting in 2004 plaasgevind het, gemeet kan word. Die horisontale beweging in die Noord-Suid (N/S) en Oos-Wes (O/W) rigtings is onttrek met behulp van die Cosi-Corr algoritme en lugfotos wat in 2009, 2012, 2015 en 2018 geneem is. Die resultate het getoon dat deurlopende beweging na die middel van die put waargeneem is met bewegings van tussen 2.1 en 4.1 m/jaar. Die grootste beperking van die gebruik van beeldkruiskorrelasietegnieke vir hellingstabiliteitmonitering was dat slegs die horisontale noord-suid en oos-wes komponente van die beweging onttrek kon word. Wanneer beweging van hellings waargeneem word, word daar ook vertikale komponente van beweging verwag en gevolglik sal die werklike verplasingsgrootte onderskat word wanneer beeldkruiskorrelasietegnieke gebruik word. Hierdie beperking dui daarop dat, alhoewel beeldkruiskorrelasietegnieke gebruik kan word om metings van putmuurstabiliteit te bekom, die metings aangevul moet word met tegnologieë wat sensitief is vir die vertikale komponent van beweging (InSAR of elevasie-model verskil analise) as 'n ware voorstelling van die totale beweging en rigting benodig word. Die tweede tegniek wat geëvalueer is, was gevorderde tydreeks InSAR, wat die ‘Small Baseline Subset’ (SBAS) metode gebruik het om hellingbewegings in 'n oopgroef-ysterertsmyn in die Noord-Kaap-provinsie van Suid-Afrika af te lei. Beide X-band en C-band SAR data is gebruik. Die resultate wat van die X-band data afgelei is, het aangedui dat, tussen Januarie en Desember 2017, die kumulatiewe verplasing tussen 21.3 mm en 23.4 mm was. Vir die C-band InSAR-data is bewegingskoerse van tussen 47.8 en 49.4 mm/jaar met 'n kumulatiewe verplasing van tot 52.3 mm vanaf Januarie 2020 tot Mei 2021 gemeet. Alhoewel InSAR sensitief is vir kleinskaalse bewegings, het die studie verskeie beperkings onthul. Eerstens, met die gebruik van X-band SAR data, is lae meetpunte-digthede as gevolg van hoë vlakke van geraas wat veroorsaak is deur dinamiese oppervlaktoestande waargeneem. 'n Ander beperking is die maksimum waarneembare limiete, wat van die golflengte van die sensor afhanklik is. Die maksimum waarneembare limiet is 25.7 cm/jaar vir X-band data wat elke 11 dae verkry is, en 42.7 cm/jaar vir C-band data wat elke 12 dae verkry is. Indien hellingbewegingsnelhede wat hierdie limiet oorskry verwag word, is hoër herhalingsintervalle of langer golflengte data nodig. Ten slotte meet InSAR data slegs die komponent van die beweging wat in die rigting van die sensor se siglyn is. Daarom kan enige komponent van die bewegingsvektor wat in 'n rigting loodreg op die siglyn is, nie gemeet word nie. Dit impliseer dat horisontale noord-suid beweging nie gemeet kan word nie. Waar horisontale noordwaartse of suidwaartse beweging vir 'n spesifieke hellinggeometrie verwag word, moet alternatiewe moniteringstegnieke soos optiese beeldkruiskorrelasie dus gebruik word. 'n Alternatief vir die direkte meting van oppervlakverplasing deur kruiskorrelasie of InSAR data is om die verandering in hoogte van oppervlaktes waar te neem deur DEM verskil analises uit te voer. Om hierdie benadering te toets is die gebruik van satellietvideodata, vasgelê deur die SkySat-C konstellasie oor 'n mynterrein in Australië, ondersoek. Die satellietvideodata het uit 25 individuele rame van beelde bestaan, met voldoende oorvleueling om 'n DEM met behulp van Struktuur-uit-Beweging (SfM) algoritmes te ontrek. Die satellietvideo DEM is vergelyk met 'n LiDAR-afgeleide DEM en 'n wortel van gemiddeldekwadraatfout van 3.1 m is verkry. Sulke hoë foute van opeenvolgende DEMe wanneer dit gebruik word in 'n hoogte-model veranderingsopsporingsbenadering die kleiner veranderinge wat verband hou met kleinskaalse hellingonstabiliteit voor 'n ineenstorting, oorskry en uitwis. Daarom mag satelliet DEM data, afhangende van die skaal van voorafgaande hellingbewegings voor ineenstorting (slegs enkele sentimeters) nie die akkuraatheid bied wat nodig is vir vroeë opsporing van hellingbeweging nie, Die resultate van hierdie ondersoeke kan gebruik word om die operasionele vermoë van die gebruik van 'n multi-sensor afstandwaarnemingsbenadering te bepaal. Doctoral 2025-06-06T08:09:43Z 2025-06-06T08:09:43Z 2025-03 Thesis https://scholar.sun.ac.za/handle/10019.1/132411 Stellenbosch University 148 pages : illustrations, maps application/pdf Stellenbosch : Stellenbosch University |
| spellingShingle | Slopes (Soil mechanics) -- Stability -- Monitoring Strip mining -- Monitoring -- South Africa -- Limpopo Mines and mineral resources -- Remote sensing Open-pit mining -- South Africa -- Limpopo -- Palabora Multisensor data fusion UCTD Haupt, Shelley Multi-sensor remote sensing for slope stability monitoring in open-pit mining |
| title | Multi-sensor remote sensing for slope stability monitoring in open-pit mining |
| title_full | Multi-sensor remote sensing for slope stability monitoring in open-pit mining |
| title_fullStr | Multi-sensor remote sensing for slope stability monitoring in open-pit mining |
| title_full_unstemmed | Multi-sensor remote sensing for slope stability monitoring in open-pit mining |
| title_short | Multi-sensor remote sensing for slope stability monitoring in open-pit mining |
| title_sort | multi sensor remote sensing for slope stability monitoring in open pit mining |
| topic | Slopes (Soil mechanics) -- Stability -- Monitoring Strip mining -- Monitoring -- South Africa -- Limpopo Mines and mineral resources -- Remote sensing Open-pit mining -- South Africa -- Limpopo -- Palabora Multisensor data fusion UCTD |
| url | https://scholar.sun.ac.za/handle/10019.1/132411 |
| work_keys_str_mv | AT hauptshelley multisensorremotesensingforslopestabilitymonitoringinopenpitmining |