This manual is under development. If you have any questions, comments or suggestions for improvement please email them to ross.roberts@aucklandcouncil.govt.nz.

• Introduction
• Before you start
• Getting started
• Database design
• Core metadata
• Spatial features
• Quality assurance

Please read these instructions before applying for access to the database – edit access is limited to individuals with the appropriate skills and qualifications, while read-only access is more widely available.

To identify if you need direct access or API access see section 2.

1.0 Introduction

No single organisation in New Zealand has overall responsibility for landslides. As a result, landslide data has been stored in multiple discrete locations, often in widely varying formats, with no ability to compile the data to get an overall understanding of the distribution of landslide risk across the country.

The New Zealand Landslides Database is designed to enable multiple organisations to deposit factual information about every landslide into a single repository, and for those organisations to be able to access the data for their own purposes.

By sharing this data each organisation will gain significant benefits by understanding how landslides might affect their people and assets. Without this consolidated, consistent and quality assured landslide inventory any efforts to assess spatial landslide hazard or risk will be limited. Landslide hazard assessments are greatly enhanced if they include information on the magnitude, encompassing the distribution, type, density, size and impacts of landslides, and temporal frequency of past landslide events.

2.0 Before you start

2.1 Getting access to edit

To help ensure that the data entered is robust and reliable, only suitably qualified and experienced people will be given edit access. In general the level of knowledge and skill needed is that of an Engineering New Zealand Professional Engineering Geologist (PEngGeol) or equivalent. Users are expected to have a robust understanding of landslide classification schema, and to understand the importance of reliable quality data management.

A fully-functional QA system is being developed for the database. In the interim, a limited QA process is in place (see section 6) and access will be strictly limited and peer-review will be required to maintain data quality.

To be granted access to the database, please enter your details in this form: https://forms.office.com/r/VFN9kdJVgy

2.2 Getting access to view

Use the button at the top of this page to apply for view only access. This is suitable for users who want access to all the data but don’t need to edit the records, or don’t have the appropriate experience.

2.3 Getting access within other GIS systems

The data can also be viewed through external GIS systems in our partner agencies through an API. Please use the button at the top to request access, or LandslideSupport@beca.com for technical details.

2.4 Getting support

For support, please contact LandslideSupport@beca.com

3.0 Getting started

Access the database here: https://app.facilitytwin.com/

The quick-start guide explains the basics of login, navigation and data entry:

4.0 Database design

4.1 Design philosophy

Understanding the reasoning behind the database design will help the user with data entry. The basics are described below:

• Each landslide is unique – it has a unique identifier that never changes.
• Only factual data is stored. To avoid issues with confidentiality, the database only stores information about the physical attributes and publicly available consequences. Private data (such as insurance claim details, landowner contact details etc) are expected to be stored in separate databases within the relevant organisations, which should be linked to the unique landslide ID to maintain data integrity and privacy.
• Each landslide can have a parent – a small landslide may be part of a larger feature. These links can be added at any time, as new information becomes available.
• Minimal data is required to create a record. When a landslide is first reported, little may be known about it. The database is structured to allow data to be added over time.
• Landslides change. As new data is added, each change is recorded. A time-slider allows change over time to be reviewed. This includes spatial mapping, which can be edited and added to over time, and these changes will all be tracked.
• Quality assurance is essential. The history tracking described above allows a clear, documented trail of all changes (change log), records who made each change and the reason for the change (e.g. the change may be that new information became available, or the landslide reactivated and physically changed, or to correct an older error in the data).
• Data is not duplicated. Where data is more appropriately entered into a different database, it’s referenced instead (e.g. boreholes should be entered into the NZ Geotechnical Database and cross-referenced).
• Where possible, data entry is limited to drop-down options aligned with the Hungr (2014) update to the Varnes classification scheme.

4.2 Structure

The database comprises two individual elements:

• A public reporting tool, to allow anyone to submit key information about landslides for review. This is a simple form designed for use by anyone to collect the basic information required for decision-making by experts.
• The main database, containing detailed information about each landslide. This is accessible only to registered users, with the intent being that these are limited to Professional Engineering Geologists (PEngGeol) or others with similar levels of understanding about landslides.

Each landslide commences with a master record, which is given a unique ID and is associated with a single point on a map. All the core landslide schema data is recorded against this location.

All the more detailed features (such as polygons showing the landslide area) are linked to this master record.

 

The core data about each landslide is stored in the master metadata table, which creates one source of truth for the landslide. A complex landslide may require details to be recorded for each smaller landslide within the main feature (for example, where each landslide has a different style or rate of movement). This is achieved by creating each sub-landslide as an individual record, creating the over-arching landslide as an individual record, and linking all the sub-landslides to the over-arching landslide as their ‘parent’.

In addition to the core metadata, an unlimited number of spatial features can be assigned to each landslide. These are entered graphically, and are summarised in the table below. More details are presented below.

Feature class	Types	Typical content
Landslide features	Points Lines Polygons	Landslide area, source area, debris area, hummocky ground, sinkholes, springs, scarp lines, cracks, damaged structures, breaks of slope, streams, landslide dammed lakes, etc.
Corrective features	Points Lines Polygons	Drains, retaining walls
Photographs	Points	Images
Investigations	Points	Links to NZ Geotechnical Database

Each landslide can also have unlimited attachments. These can include pdf reports, photographs, spreadsheets of monitoring data, maps, URLs to link to external websites, etc.

5.0 Core metadata

In the main data entry page, the metadata is collected into groups of related information to simplify the data entry. These can be entered in any order, and will be ‘lodged’ into the database and sent for QA review when the Submit button is pressed. This button does not need to be pressed after every page, or after every field, and no data will be lost if you exit the database without pressing this button.

The core metadata is as follows:

Group	Data fields
1. Identifiers	1.1 Parent landslide 1.2 Landslide name
2. Location	2.1 Location description 2.2 Location recording method or device 2.3 Estimated error in location recording method
3. Description and setting	3.1 Physical setting 3.2 Geological Setting 3.3 Geomorphological Setting 3.4 Landslide Description
4. Movement	4.1 Primary movement type 4.2 Primary movement type subclass 4.3 Secondary movement type(s) 4.4 Complexity of mass movement
5. Date and activity	5.1 Has the landslide been active in the last 1000 years? 5.2 Landslide activity 5.3 Velocity class / rate of movement 5.4 Date of first movement 5.5 Accuracy of dating method
6. Shape and dimensions	6.1 Slope angle prior to landslide 6.2 Slope angle of displaced material 6.3 Angle of rupture surface (translational) 6.4 Aspect (direction of movement) 6.5 Length along ground zone of depletion 6.6 Length along ground zone of accumulation 6.7 Maximum height of main scarp 6.8 Maximum width of landslide perpendicular to direction of movement 6.9 Maximum depth from original surface to surface of rupture 6.10 Travel angle 6.11 Estimate of landslide area (min) 6.12 Estimate of landslide area (max) 6.13 Estimate of landslide volume (min) 6.14 Estimate of landslide volume (max) 6.15 Volume/area estimation method
7. Materials	7.1 Materials in landslide 7.2 Debris type 7.3 Vegetation cover
8. Triggers	8.1 Primary causal triggering factor 8.2 Other factors influencing instability 8.3 Estimated magnitude of earthquake causing landslide 8.4 Rainfall total in preceding 24 hours 8.5 Rainfall total in preceding 3 days 8.6 Rainfall total in preceding 7 days
9. Consequences	9.1 Damage description 9.2 Approximate number of injured people 9.3 Approximate number of fatalities 9.4 Approximate damage cost 9.5 Environmental damage 9.6 Features and assets impacted by the landslide
10. Legacy data links	10.1 GNS – Number of Slips 10.2 GNS – Size Category 10.3 GNS – Injuries 10.4 GNS – Deaths
11. Information sources	11.1 Source of landslide information (description) 11.2 Source bibliography 11.3 Project name / number 11.4 Original database (for bulk uploaded files)
12. Quality assurance	12.1 Reported by (Name) 12.2 Reported by (email) 12.3 QA Name 12.4 QA Person email 12.5 QA level 12.6 QA status 12.7 Comments

Each of these data fields has specific guidelines on what can be entered:

5.1 Identifiers Group

Field	Drop-down options (where relevant)	Expected format	Guidance
1.1 Parent landslide	Look-up of existing landslides	Choice	Where the landslide being entered is a sub-feature of a larger area of mass movement, select the parent landslide here to create a link between the two features. A thorough search should be done before a parent is assigned. There should be a ‘genetic’ link or obvious spatial enclosure for a landslide feature to have a parent.
1.2 Landslide name		Text	Common name for landslide feature. Historical published names should be favoured (e.g., ‘Casuarina Crescent’or ‘Grooms Slip’) and this may also include the year (e.g., ‘Beauty Point 1952 Landslide’).

5.2 Location Group

Field	Drop-down options (where relevant)	Expected format	Guidance
2.1 Location description		Text	Common name for landslide feature. Historical published names should be favoured (e.g., ‘Casuarina Crescent’or ‘Grooms Slip’) and this may also include the year (e.g., ‘Beauty Point 1952 Landslide’).
2.2 Location recording method or device		Text	Describe the method used for identifying the location (e.g. hand-held GPS, aerial photographs etc).
2.3 Estimated error in location recording method	<1m <10m <100m <1km <2.5km <25km	Choice	

5.3 Description and setting group

Field	Drop-down options (where relevant)	Expected format	Guidance
3.1 Physical setting	Coastal cliff Cutting (e.g. road, railway, quarry) Inland cliff Man-made fill Natural slope Other	Choice	
3.2 Geological Setting		Text	Text summary of the geological setting including general structural history and lithological types if known. Should be concise, e.g., ‘Brecciated fault contact between sandstone and mudstone’.
3.3 Geomorphological Setting		Text	Text summary of the geomorphological setting including landforms, slope, aspect, interpreted Tertiary and/or Quaternary environmental settings and related processes. Should be concise e.g., ‘Steep NE facing slope on old volcanic centre, deep weathering profile’.
3.4 Landslide Description		Text	General description of the landslide

5.4 Movement group

Field	Drop-down options (where relevant)	Expected format	Guidance
4.1 Primary movement type	Erosion Fall Flow Retaining wall failure Slide Slope deformation Spread Topple Other Not collected Unknown	Choice	As defined in Hungr 2014 (with additions)
4.2 Primary movement type subclass	Changes depending on choice in 4.1	Choice	As defined in Hungr 2014 (with additions)
4.3 Secondary movement type(s)	Same as 4.1		As defined in Hungr 2014 (with additions)
4.4 Complexity of mass movement	Multiple events, multiple styles (composite) Multiple events, one style (multiple/successive) One event, multiple styles (complex) One event, one style (single)	Choice



5.5 Date and activity group

Field	Drop-down options (where relevant)	Expected format	Guidance
5.1 Has the landslide been active in the last 1000 years?	Certainly Probably Probably not No Unknown	Choice	Allows filtering out of inactive landslides
5.2 Landslide activity	Active Reactivated Suspended Inactive – Dormant Inactive – Abandoned Inactive – Stabilised Inactive – Relict Unknown	Choice	
5.3 Velocity class / rate of movement	Extremely rapid Very rapid Rapid Moderate Slow Very Slow Extremely slow Unknown	Choice	Date that the landslide first occurred. If not known accurately, give a best estimate and clarify in the ‘accuracy of dating method’ field. If no indication is know, set date to 1 Jan 1000
5.4 Date of first movement		Date/time	Use this to give an indication of how reliable the date of first movement is.
5.5 Accuracy of dating method	Second Minute Hour Day Month Year Decade Century Unknown Not recorded	Choice



5.6 Shape and dimensions group

Field	Drop-down options (where relevant)	Expected format	Guidance
6.1 Slope angle prior to landslide		Number	As defined in figures 3.3, 3.4, 3.5, and 3.7 of Cruden & Varnes
6.2 Slope angle of displaced material		Number	As defined in figures 3.3, 3.4, 3.5, and 3.7 of Cruden & Varnes
6.3 Angle of rupture surface (translational)		Number	As defined in figures 3.3, 3.4, 3.5, and 3.7 of Cruden & Varnes
6.4 Aspect (direction of movement)	North North-East East South-East South South-West West North-West Unknown	Choice
6.5 Length along ground zone of depletion		Number
6.6 Length along ground zone of accumulation		Number
6.7 Maximum height of main scarp		Number
6.8 Maximum width of landslide perpendicular to direction of movement		Number
6.9 Maximum depth from original surface to surface of rupture		Number
6.10 Travel angle		Number
6.11 Estimate of landslide area (min)		Number	
6.12 Estimate of landslide area (max)		Number
6.13 Estimate of landslide volume (min)		Number
6.14 Estimate of landslide volume (max)		Number
6.15 Volume/area estimation method		Text	Describe how the volume was estimated and comment on likely accuracy



5.7 Materials group

Field	Drop-down options (where relevant)	Expected format	Guidance
7.1 Materials in landslide	Rock Soil Ice Boulders Cobbles Gravel Sand Silt Clay Other Unknown Not collected	Choice	
7.2 Debris type	Blocky Chaotic Unknown Other Not collected	Choice	
7.3 Vegetation cover	Unknown Not collected None Pasture Soil con plantings Scrub Indigenous forest	Choice	



5.8 Triggers group

Field	Drop-down options (where relevant)	Expected format	Guidance
8.1 Primary causal triggering factor	Earthquake Rainfall Tectonic / volcanic uplift Erosion / scour / piping Devegetation Melting snow/ice Weathering / material degradation Excavation Structure failure Broken pipes Drawdown of reservoir Groundwater change Mining Artificial vibration Stress change Other Unknown Not collected	Choice	Select the primary causal triggering factor. Where the primary cause isn’t known, choose ‘unknown’ and put the likely triggers into the ‘other factors causing instability’ field.
8.2 Other factors influencing instability	As per 8.1	Multichoice	
8.3 Estimated magnitude of earthquake causing landslide		Number
8.4 Rainfall total in preceding 24 hours		Number
8.5 Rainfall total in preceding 3 days		Number
8.6 Rainfall total in preceding 7 days		Number



5.9 Consequences group

Field	Drop-down options (where relevant)	Expected format	Guidance
9.1 Damage description		Text	
9.2 Approximate number of injured people		Number	
9.3 Approximate number of fatalities		Number	
9.4 Approximate damage cost		Number
9.5 Environmental damage		Text
9.6 Features and assets impacted by the landslide	Private house (building, not including gardens) Public or commercial building Private driveway or garden Public road Public walkway or path Railway Pipes, powerlines or other infrastructure Retaining wall or sea wall Public land Other	Multichoice



5.10 Legacy data linkages to GNS database group

These fields allow the import of data from the existing GNS landslides database without over-writing the equivalent fields with different data entry options in the new database. For example, in the new database the number of deaths is an actual number, not a choice range.

Field	Drop-down options (where relevant)	Expected format	Guidance
10.1 GNS – Number of Slips	None Many Several Not collected Unknown	Choice	
10.2 GNS – Size Category	Large Moderate Small Not collected Unknown	Choice	
10.3 GNS – Injuries	None Many Several Not collected Unknown	Choice
10.4 GNS – Deaths	None Many Several Not collected Unknown	Choice	



5.11 Information sources group

Field	Drop-down options (where relevant)	Expected format	Guidance
11.1 Source of landslide information (description)		Text	
11.2 Source bibliography		Text
11.3 Project name / number		Text	
11.4 Original database (for bulk uploaded files)		Text	



5.12 Quality assurance group

Field	Drop-down options (where relevant)	Expected format	Guidance
12.1 Reported by (Name)		Text	
12.2 Reported by (email)		Text	
12.3 QA Name		Text	The name of the person who has undertaken QA checking of the most recent submission.
12.4 QA Person email		Text	The email address of the person who has undertaken QA checking of the most recent submission.
12.5 QA level	(A) Complete and Correct (B) Complete (C) Incomplete – data entered. (D) Unknown	Choice	(A) Complete and Correct – only assigned after check of data entry and field verification of the data by an engineering geologist. (B) Complete – all data easily available entered, further QA may be required. (C) Incomplete – some data missing. (D) Unknown – unknown level of QA on historical data. Add notes in the comments section explaining your decision.
12.6 QA status	Open Problem Closed	Choice	Open (fit for publication), Problem (an issue has been identified, but can remain published while being assessed), Closed (not fit for publication due to outdated, unreliable, or incomplete data). Add notes in the comments section explaining your decision.
12.7 Comments		Text	Free field to comment on QA issues

6.0 Spatial features

The database is intentionally structured to allow data entry with only a location, as the spatial extent and other physical details are often not collected when the landslide is first identified. These features can be added later (through the web mapping interface) as the information is collected, and can be edited over time as the landslide changes or as new data is obtained.

The history records all changes to the spatial features, and the time-slider can be used to visualise change over time.

6.1 Overview

Features that can be sketched for each landslide are:

	Points	Lines	Polygons
Morphology		Crown Concave break of slope Ground cracking Convex break of slope Eroded gully Flank Scarp	Landslide area Gully erosion Source area Debris area Hummocky ground
Structures	Damaged structure	Damaged structure	Damaged structure
Water	Sinkhole Spring/seepage	Stream	Landslide dammed lake Pond or lake
Corrective action	Corrective action feature	Corrective action feature	Corrective action feature
Photo	Photo
Ground investigation	Investigation point
Other	Other Unknown	Other Unknown	Other Unknown

For each of these spatial features, there is a small metadata set to allow useful description.

6.2 Morphology, structures & water (polygons)

Field	Drop-down options (where relevant)	Expected format	Guidance
Type	Landslide area Gully erosion Source area Debris area Hummocky ground Damaged structure Landslide dammed lake Pond or lake Other Unknown	Choice
Description		Text	Free field for description

6.3 Morphology, structures & water (lines)

Field	Drop-down options (where relevant)	Expected format	Guidance
Type	Crown Concave break of slope Ground cracking Convex break of slope Eroded gully Flank Scarp Damaged structure Stream Other Unknown	Choice
Description		Text	Free field for description

6.4 Morphology, structures & water (points)

Field	Drop-down options (where relevant)	Expected format	Guidance
Type	Damaged structure Sinkhole Spring/seepage Other Unknown	Choice
Description		Text	Free field for description

6.5 Corrective action feature (polygon)

Field	Drop-down options (where relevant)	Expected format	Guidance
Type	Ground improvement Toe weight Crest excavation Soil redistribution Surface treatment Mesh Soil nails Dewatering zone Planting Other Unknown	Choice
Description		Text	Free field for description
Integrity	Fully functional Partly functional Failed Decommissioned (replaced) Decommissioned (not replaced) Other Unknown	Choice
Date		Date/time	The date the corrective measure was put in place (not the date the entry was made in the database – this is automatically captured)

6.6 Corrective action feature (line)

Field	Drop-down options (where relevant)	Expected format	Guidance
Type	Retaining wall Gabions / mass Line of anchors Catch barrier / dam Guide barrier / dam Drainage – pipe Drainage – counterfort Drainage – other Other Unknown	Choice
Description		Text	Free field for description
Integrity	Fully functional Partly functional Failed Decommissioned (replaced) Decommissioned (not replaced) Other Unknown	Choice
Date		Date/time	The date the corrective measure was put in place (not the date the entry was made in the database – this is automatically captured)

6.7 Corrective action feature (point)

Field	Drop-down options (where relevant)	Expected format	Guidance
Type	Pile Anchor Drainage – manhole Drainage – exit point Other Unknown	Choice
Description		Text	Free field for description
Integrity	Fully functional Partly functional Failed Decommissioned (replaced) Decommissioned (not replaced) Other Unknown	Choice
Date		Date/time	The date the corrective measure was put in place (not the date the entry was made in the database – this is automatically captured)

6.8 Photo point

Field	Drop-down options (where relevant)	Expected format	Guidance
Type	Overview Detail Sketch	Choice	Allows sketches as well as photographs to be georeferenced
Orientation		Degrees	Describes the direction in which the photo was taken relative to north.
Date		Date/time	The date and time the photo was taken

6.9 Investigation point

Field	Drop-down options (where relevant)	Expected format	Guidance
Investigation point type	Inclinometer Piezometer Borehole Extensometer Survey point Other Unknown	Choice
Feature ID		Number	Unique ID from New Zealand Geotechnical Database to allow cross-referencing

7.0 Quality assurance

The primary control on quality assurance is limiting access to suitably qualified and experienced individuals. In addition, there is a QA process built into the database. The process is as follows:

1. The landslide data is entered by a suitably qualified and experienced person. They do not fill in fields 12.3 to 12.7 (in the QA section), and do not tick the “Is this landslide valid?” box. Once they have entered all available data, they press the “submit” button. At this stage the landslide appears in the “Validate Landslides” table – but is not shown on the main map.
2. The nominated QA person from the same organisation as the data entry person (or external if no suitable individual) opens the landslide from the “Validate Landslides” table and checks all data against their records
3. If the nominated QA person is satisfied that the data is correct, they:
   • Set the QA Level to “A – complete and correct”
   • Set the QA Status to “Open”
   • Tick the check-box labelled “Is this landslide valid?” – this makes the landslide visible to other users on the main map

The date, time, and user name of the individual making these changes is recorded in the change log as part of the QA process.

If more changes are made to the landslide after the QA process is completed, these will show in the QA log. It is expected that the QA person will then re-check the data and reconfirm the validity. This is not yet built-in to the system as an automated process, so it is expected that good communication will be required.