UC2/US1 Find and access earth observation satellite data from neighbour countries for SOC model

[pvgenuchten]

As a developer/researcher,

I want to find and access new EO datasets in neighboring countries fitting the requirements for my SOC machine learning model,

So I can make a SOC map for W-Europe region based on my machine learning model, that can support policy advisors of that region

Extended user story

I am researcher, who works at ILVO, which focuses on data engineering and science in Belgium, in the role of data engineer, my educational level is higher education. My main target group are land managers, farmers and policy makers and government institutes, who work in land use type agriculture . I don’t have a soil science background. I do have a data science background. My digital expertise encompasses coding, data engineering and science technologies, cloud technology, ML/AI technologies. I usually use portals Google, Inspire, datavindplaats by the Flemish government, geopunt to find information or data.

I have a machine learning model that uses earth observation satellite data to predict the soil organic content (SOC) in the topsoil of agricultural fields in Flanders. I want to find new datasets in neighboring countries fitting the requirements for my SOC machine learning model,

So I can make a SOC map for W-Europe region based on my machine learning model, that can support policy advisors of that region

I aim to find data on:

• Soil samples with an OC value expressed as %
  • The soil samples were taken in the years 2020 – 2023
  • The soil samples were taken at a specific location and are not a mixed sample for a whole parcel/field
  • The soil samples have a location expressed as POINT(x,y)
  • The soil sample was analyzed in the laboratory for OC using ISO 10694 standard method
  • The soil samples occur in the neighboring countries (NL, DE, LUX, FR, UK)
• Covariate data which support the machine learning model
  • Digital elevation model
  • Land use
  • Water content from satellite; thermal-optical trapezoid model
  • Other satellite data, such as leaf area index

Acceptance criteria

• [ ] The search mechanism allows me to set relevant filters so I don't need to go through 1000's of potential hits

[KathiSchleidt]

@pvgenuchten think this UC may need to be renamed one more time - something like "UC2/US1 Find and access recent terrestrial OC data from neighbour countries for SOC model"

[KathiSchleidt]

almost identical with #4 (same requirements)

[pvgenuchten]

4 seems to refer to datasets with a predicted organic carbon distribution, this issue seems to refer to datasets which enable the prediction of a organic carbon distribution?

[chlebas]

The User story is ambiguous as the user story is refering to EO datasets, that means remote sensing datasets but in the background it refers to measured SOC. In that sense, a part of the background is similar with the issue [https://github.com/soilwise-he/Soilwise-userstory-epics/issues/4] as mentioned by Kathi. To focus more on covariates?

[KathiSchleidt]

Extracted Requirements:

• 01.1 - Data Discovery - Query Catalogue: fine-grained search criteria including co-located measurements, OMS measurement metadata
• 01.1 - Data Discovery - Query Catalogue: Covariate data, e.g., Digital elevation model, Land use, satellite products (e.g. Water content, leaf area index)
• 01.2 - Data Discovery - map preview : View discovered data
• 04 - Download data: Download data for ML/AI
• 08 - Vocabulary: Vocabs on SOC