Review of key datasets (version 11)

[pzwsk]

This issue is to help discuss and review list of key datasets of the OpenDRI Index, from version 10 - currently online - to next version 11.

Main known issues:

• Users do not always clearly understand what dataset is expected
• There is a lack of contextual information for each of the dataset, why is it useful, how is it used, when and where, etc.
• Being able to check core data attributes for each dataset would be of high value
• Hazard datasets are not consistent between each others, for instance: flood hazard map but seismic hazard curve and seismic hazard model
• Some datasets are still very vague (what is really flood protection measure, only infrastructure or more than that)
• Critical infrastructures is too broad, we need to breakdown into detailed infrastructures

[pzwsk]

Below is the suggested structure for version 11 of datasets with full list of datasets available here for comments and review https://docs.google.com/document/d/18kajVv_YgiDKYyn4pVpnEQykWgVMoG786evA0s35tbs/edit#

Key dataset name

Name of the key dataset.

Category

• The category the dataset belongs to.
• Base data
• Exposure
• Hazard
• Vulnerability curves
• Risk indicators

Hazard

Which hazard the dataset applies to (not suitable for all datasets).

• Coastal flooding
• Cyclone
• Earthquake
• Landslide
• River flooding
• Tsunami
• Volcano
• Water scarcity

Description

A paragraph describing the key datasets - what is expected - in a clear language and including minimum criteria for a submitted dataset to be considered valid (resolution, granularity, etc.).

Attributes

A list of core data attributes (information) the dataset is expected to contain. For instance, building dataset is expected to contain information on :

• Footprint
• Height
• Roof type
• Material
• Use type

Those attributes would be aligned with data standards if existing.

Resolution

An indication on the expected or existing resolution or granularity.

Formats

List of digital formats typically used to provide and read such dataset. This should include compression format only if specific to the dataset.

Rationale

A paragraph describing why this dataset is important to disaster risk management and climate change adaptation with examples of how it has been used in past projects.

References

Sources used to define the different elements above such as links to standards, glossary, definition, reports, etc.

[pzwsk]

List of key datasets version 10 is available here https://index.opendri.org/methodology.html#datasets

List of suggested key datasets version 11 visible here https://coggle.it/diagram/XJTRmtGbqDO9GHiL/t/-

[stufraser1]

v11 is a clear improvement on v10 I think. Still some inconsistencies exist.

Exposure

• Infrastructure has quite detailed split but Education/Health could also be considered buildings (could split to res, com, ind, edu, health, other civic buildings) but this could become a large category, where as the type/usage of building is usually defined by the attribute 'occupancy' within the dataset. Perhaps we could instead specify at the top level 'transport infrastructure (subcategories of: road, rail, bridge, tunnel, port, airport)' and 'supply infrastructure' (power, potable water, wastewater, telecomms)?
• What does 'Business' (and 'economy') dataset refer to?
• Is 'Land' a definition of landuse-landcover?
• Agriculture could comprise forestry, livestock, or crop.

Base data

• 'ortopho' should be replaced by 'aerial imagery' per the google doc, and as it is the more recognised term.
• How do we deal with seamless topo-bathym datasets e.g. more commonly derived from lidar, if topo and bathym are separate? Could we combine the two into 'elevation data'? These should be included in DEM, I think.
• Landuse-landcover can be exposure or base data. I think is would sit better under base data (i.e. where users would most likely expect it as in my experience it is most often used to develop exposure information, not used directly as the exposure data)
• 'soil' - could be broadened to geology (bedrock geology data is also important in seismic modelling)
• New category of 'hydrographic' - to include channel hydrography, watershed boundaries, water bodies?

Hazard

• Storm surge gauge data would more accurately be sea level gauge data as they record tsunami, tidal levels, storm waves too. This category could also be extended to seismic monitoring data.
• 'Hazard map' should be defined as 'return period hazard map' to distinguish from historical.
• 'Historical event' should be defined as historical event footprint/distribution if this is supposed to be a map of hazard intensity experienced/observed in a historical event.
• 'Projection' needs clarifying (not included in the Google Doc - please advise intention here).
• There are some categories that refer directly to hazard intensity, which should be aligned as options under 'return period hazard maps' and 'historical event': tsunami wave height and Strong wind speed.
• I would say 'site conditions' and 'flood protection' site better under base data
• A category of geological features could be introduced, which would nicely encompass maps of fixed features that are not themselves hazard maps: 'active volcano' and 'seismic fault'.
• 'seismic hazard curves' are again peril specific and could be included as one peril option under something that more clearly describes an exceedance probability curve - these can also be determined for tsunami wave height, surge height, windspeed etc.

Can we align the hazard categories with discussion we've been having on the risk data library and ThinkHazard. the list should include extreme heat, wildfire, other types of drought too.