Cluster Call: Soil Data Landscape - 8 April 2026 1200 US-Eastern

[ktoddbrown]

Registration for this (Soil Ontology and Informatics) meeting is in the ESIP Community Calendar

Opening - 5 min [0:00]

• Review and consent to agenda
• Roll call ESIP SOI - RollCall
• Confirm notetaker
• Confirm next meeting Calendar

Check ins - 5 min [0:05]

Networking with other soil data nerds is important! In randomly generated break out rooms (2-3 people per room) touch base with someone. No notes.

• What should we call you and what is your affiliation?
• How are you showing up? excited, sick, tired, distracted,... No notes here please
• What did you find that was exciting or interesting since our last call?

Information - 5 min [0:10]

Anything new that you’ve heard related to this group that you would like to share? (Celebrations, advertisements, publications, or other announcements)

Decision - Scoping soil data landscape - 10 min [0:15]

Soil_data_landscape.md was initially created in 2021 and has been updated inconsistently over the years. The original goal here was to create a list of all current soil informatics resources and data products to increase awareness within the cluster and beyond. We were initially looking for vocabularies, code lists and other semantic resources. This was then extended to data products themselves.

We would like to move this to a more machine readable format [JSON/YAML/BibTex] to support a dynamic web interface similar to https://esipfed.github.io/disasters-trusted-data-now/ (also see #9). This would also need a process set up to continue to maintain the list (flag stale links and update new resources) and a publicity plan for stakeholder audiences.

Round: Is there anything missing? Do you have any clarifying questions?

Brainstorm - How could we do this? - 10 min [0:25]

1. Harvest tags from the current list (categories and control vocabularies)
2. Weight the options of machine readable formats
3. Transcribe the document into 'cards'
4. Weight options for site generator
5. Scope criteria for inclusion
6. Generate vetting process/schedule
7. Update and extend resource list

Round: What else could be done to move this forward?

Decision - Who is doing what when? - 10 min [0:35]

INSERT SUMMARY OF LAST TOPIC

Round: Who is doing what when?

Brainstorm - What format should we use for the 'cards'? - 10 min [0:45]

• One or multiple resources per file?
• What format makes the most sense?
• INSERT OTHER DISCUSSION TOPICS

Round: What is your first reaction here?

Close out - 5 minutes [0:55]

[ktoddbrown]

Notes

Decision - Scoping soil data landscape - 10 min [0:15]

Soil_data_landscape.md was initially created in 2021 and has been updated inconsistently over the years. The original goal here was to create a list of all current soil informatics resources and data products to increase awareness within the cluster and beyond. We were initially looking for vocabularies, code lists and other semantic resources. This was then extended to data products themselves.

We would like to move this to a more machine readable format [JSON/YAML/BibTex] to support a dynamic web interface similar to https://esipfed.github.io/disasters-trusted-data-now/ (also see #9). This would also need a process set up to continue to maintain the list (flag stale links and update new resources) and a publicity plan for stakeholder audiences.

Round: Is there anything missing? Do you have any clarifying questions?

• How well do we signal our understanding of a class of assets?

• How do we identify resources and when they get incorporated? What is the difference between citizen vs expert driven

• How do you represent assets in multiple repositories, provenance?

• Details will matter, what does active mean and how is it determined?

• TODO: add something on maitainance and updates ie governance.

Brainstorm - How could we do this? - 10 min [0:25]

1. Harvest tags from the current list (categories and control vocabularies)
2. Weight the options of machine readable formats
3. Transcribe the document into 'cards'
4. Weight options for site generator
5. Scope criteria for inclusion
6. Generate vetting process/schedule
7. Update and extend resource list

Round: What else could be done to move this forward?

• ? Pro - GitHub but would google sheets be easier to work with? As a two step process. Version control - not on google sheets

• Is the focus semantic resources or data products? Data is huge!

• What is semantics?

  • Definitions and terms
  • Methodologies - method codes
  • Ontologies
  • API data exchange or data models

• Shared repositories of methodologies that can be instantiated in different ways

• Think carefully about maintenance and how to drive value to come back to over time. https://developmentseed.org/blog/2024-08-26-eoapi-humanitarian/

• Soil moisture Mesonet data could have relevant interoperability needs. Where do networks and monttoring efforts fit in here?

• TODO: Is there urgent needs to preserve data given FS and USGS turnovers

• Ask Dylan at NRCS - how are individual pedons and lab data getting saved?

• TODO: add in more explicit definition of semantics to scoping

• TODO: add in governance discussions earlier

Decision - Who is doing what when? - 10 min [0:35]

INSERT SUMMARY OF LAST TOPIC
Round: Who is doing what when?

• CORY: pulling together a subset of the entries to a google sheet for us to start pulling in tags

Brainstorm - What format should we use for the 'cards'? - 10 min [0:45]

• One or multiple resources per file?
• What format makes the most sense?
• INSERT OTHER DISCUSSION TOPICS
  Round: What is your first reaction here?

• YAML is a happy place for many of us.
• JSON-LD is another option.
• XML seems to be sunsetting.
• LinkML can translate between YAML and JSON

[ktoddbrown]

Soil data landscape scope

We want to create a list of all current soil informatics resources to increase awareness within the cluster and beyond. Soil_data_landscape.md was initially created in 2021 and has been updated inconsistently over the years. We were initially looking for vocabularies, dictionaries (term-definition pairs), code lists (particularly method descriptions), and other semantic resources. This was then extended to data products themselves. There is also interest in monitoring networks and organizations.

We would like

• to move from the current free text formation to a more machine readable format,
• connect the machine readable formats to a dynamic/static web interface to increase discoverability of these resources,
• set up a governance process to scope, maintain, and extend this collection including flagging stale links and adding new resources,
• publicize this resource and engage feedback from stakeholders.

[ktoddbrown]

Next steps

• Cory will take a subset of the markdown list and pull it into a google sheet for us to start working on tags and categories

1. Harvest tags from the current list (categories and control vocabularies)
   • Resource type:
   • vocabulary - terms only
   • dictionary - terms and definition
   • ontologies - inter-related terms
   • organizations - who is stewarding what resource or project
   • data products - ?do we want to go here?
   • project - ?do we want to go here?
2. Decide how to decide
3. Weight the options of machine readable formats
   • XML many of the US agencies use XML but there is generally a migration away from this
   • YAML is generally felt to be very approachable
   • JSON(LD?) is a current common standard but might be harder to work with as a flat text file
   • BibTex is not common
4. Transcribe the document into 'cards'
   • Generally one card per resource, remember that resources will be interconnected
5. Weight options for site generator
6. Scope criteria for inclusion
7. Generate vetting process/schedule
8. Update and extend resource list