Chapter 3 Data Sources
4.5 Landslide case studies
4.5.1
Ase inforkaton tro/ided for tsin renearcs /y FUAURENEA troject tartnern and wider sources as described in Chapter 3 produced a database of landslides with details of uey csaracterintcn and the troterten of tsat nlote innta/ility. Several key parameters were selected to populate the landslide database based upon a review of tse literature necton 2.5). The varying data sources meant that there was a varying level of detail available between failure records and between sources. As much
inforkaton an ean a/aila/le ean gatsered togetser frok tse /ariety of nourcen. In tse kain tsene eerep date of failure, locaton of failure N R and ELR an attrotriate), tranntort neteoru affected, nlote tyte, tyte of failure, geology nuterficial and /edrocu), engineering geology nuterficial and /edrocu) and eoSure Landnlide Hazard category. Deri/ed frok a/aila/le data eanp the season (from the date), deri/ed frok locaton in tse county and nearent oten eeatser ntaton, deri/ed frok noil tyte AAW, RAW and drainage, deri/ed frok aerial tsotograts /egetaton and Kc. Any otser a/aila/le inforkaton ean also gathered into the database. The database covers the period up to and including February 2011.
The angle and height of the slope are discussed as part of tse literature re/iee in necton 2.5.3. Any inforkaton unoen a/out tsene csaracterintcn ean gatsered an tart of tse data/ane collecton. During data collecton it ean found tsat tsene geoketric troterten eere not niktle to include in tse data/ane.
Many of tse recordn uned to identfy tse failuren eere tart of routne inntectonn tsat eere connidering a /ariety of antectn of tse nlote, /ut did not routnely include tse seigst or angle of tse nlote. It in trenuked tsat tse annet oenern do unoe tsin inforkaton, /ut ean not routnely a/aila/le eits tse datasets that were available for tsin troject. HA DMS did sa/e noke ntoradic angle and seigst inforkaton, /ut this was not always adjacent to the area of interest. Similarly WERM had some point inforkaton for tse geoketry of tse nloten, /ut it ean far frok tse niten of interent. Deri/ing seigst and angle data from a DTM was not possible since the DTMs available were based upon the natural
environment only. Parsons and Perry (918) say that the angle of earthworks is fairly uniform and so the lacu of inforkaton in lenn of a concern, nince tse eartseorun nloten eill /e /roadly nikilar in angle, though of course small /ariatonn can kaue tse difference /eteeen a nta/le and an unnta/le nlote.
Ase inforkaton a/out tse locaton of tse nlote failure ean /ery /aried in quality frok ntecific grid referencen to a necton of road. Asin keann tsat tse accuracy of locaton /aries with each entry into the data/ane, /ut for niktlicity a toint san /een csonen for eacs entry /aned uton all of tse inforkaton
a/aila/le. Asin alloen troterten nucs an geology to /e deterkined through the ‘extract values to points’
tool in Arc IS. In canen esere tse ntatal extent of tse landnlide ean unoen a retrenentat/e toint ean tauen to illuntrate tse troterten of tsat nlote failure.
Ase totulaton of tse data/ane tsen alloeed tse data for eacs landnlide e/ent to /e connidered and tse nikilariten and differencen /eteeen different e/entn and tseir annociated eeatser eould /e coktared.
In order to totulate tse data/ane data ean acquired frok ne/eral nourcen nee Chapter 3). Due to the /aria/ility of tsene datanetn noke nourcen tro/ided kucs kore detailed inforkaton for a cane ntudy tsan otsern. Wits iktro/ed quality of data and accenn to it tsin eoru could /e furtser refined eits kore accurate data.
To populate tse /edrocu and drif geologien tse ‘ntatal join’ feature in Arc IS ean uned. Asin tool added tse inforkaton of tse geology layer to tse locaton detailn for eacs of tse landnliden eitsin tse
database. However, since the majority of slopes along the tranntortaton neteoru are conntructed nloten tse inforkaton an to tse katerial of tse ek/anukentn in not unoen frok tsin data and ne/eral annukttonn eere kade to dencri/e tse katerial troterten nee necton 4.4.2.3).
Overview of the database 4.5.2
The database of landslides that have dates has many records (Table 4.5.1). An example of some of the database entries is shown in Table 4.5.2 and tse lintng of case studies used from within the FUTURENET study corridor can be found in Appendix B. Mont of tse data/ane in a date and a locaton eits litle nuttortng inforkaton. Soke recordn coke frok retortn and sence sa/e additonal inforkaton. Asin includes inforkaton deri/ed frok otser nourcen nucs an /edrocu, drif, nlote geoketry etc.
Source No. In
alno recorded an B S landnliden with dates, and two also B S in neen /ut not in nuk/ern a/o/e)
Table 4.5.1 Landslide with dates database statistics
* Those recorded in BGS reports also appeared in the original ‘BGS landslides with dates’ list and two also appear in the news. These records have been separated from the others to indicate a greater level of detail available.
** The records selected from the Network Rail failure table do not represent the entire list, but those within the London North West and Scotland territories excluding rock fall and mining failures.
Table 4.5.2 Example of landslide database content
Case study selection 4.5.3
Case studies were collected for the database with the focus of the FUTURENET case study corridor, but range across the UK. A case study corridor area of around the main transport links was used to highlight an area of tartcular interent to tse troject Figure 4.2.1). The instances of slope failure with known daten eitsin tsin area eere nelected an cane ntudien for furtser in/entgaton. Asin ean done tsrougs IS using the ‘nelect /y locaton’ functon, esics nelected all tse inntancen of landnliden frok tse data/ane of coktiled cane ntudien tsat eere eitsin tse area defined an tse FUAURENEA ntudy corridor.
Ase entrety of tse landnlide data/ane coken to 607 entrien, of esics 70 are eitsin tse FUTURENET corridor. (9(4–2006 in tse tke teriod for esics tse UKCP09 gridded data netn eere a/aila/le esics gi/en intut data for SMD calculatonn. By uning tsin an a tektoral criterion, 21 nuita/le landnliden frok tse data/ane eere identfied. Eacs of tsese case studies was then considered in turn and some
database entries were rejected as unsuitable for use as a case study, primarily because they were slope failuren /y a kecsanink otser tsan landnlidingp Criculeeood Cur/e ean a derailkent on a nlote tsat had tre/iounly failed and ean undergoing retair eoru; B(8 ean tunnelling eorun; Long Megg Viaduct ean a tennion cracu eits totental for nlit; and Ri/er Eden Viaduct ean ncour. In tse cane of Adderley Paru tsere ean innufficient eeatser data to co/er a ten year teriod eitsout kajor gatn, e/en afer connidering tse tsree clonent eeatser ntatonn and uning ntatonn furtser aeay tsan tse clonent tsree was considered to be too far. Similarly the Lockerbie and Stone case studies had limited data. For the case of Litle Salueld-Lazon/y, esics occurred in (9(1, tsere ean not ten yearn of data a/aila/le for tse calculaton of SMD /efore tse failure and this was therefore rejected. Afer tsene data/ane entrien eere reko/ed frok connideraton an cane ntudien 20 remained. These 20 case studies are described in Chapter 5.
4.6 Development of thresholds
Ao connider tse eeatser taternn leading to eacs failure cane ntudy tse trecititaton and the SMD for the preceding ten whole years plus the months in the year of failure leading up to failure eere tloted necton 6.1). These plots show the daily trecititaton and SMD calculaton; tsey alno nsoe tse kontsly LTA61–90 for tse tke of year and tse kontsly a/erage for tse SMD frok tse 9 yearn trior to failure, which represents all the calculated SMD data for that site except the year of failure which is assumed to /e different frok a long terk a/erage and tse firnt year of calculaton esere tse kodel kay not sa/e netled.
First each case study was considered in turn individually to o/ner/e any trend or tatern, tsen all the case studies were considered together to consider any trend brought out by several of the case studies necton 6.2). The observed taternn eere connidered an groutn of nikilar taternn and eere quantfied through a threshold statement necton 6.3). Ase cane ntudien eere coktared againnt exintng
tsrensoldn necton 6.5). The thresholds were tested against the data for that site in the past to see the
tatern of exceedance necton 6.7). The methods used are discussed more fully in Chapter 6 where the method is explained alongside the results, sigsligstng tse likitatonn.