OUTPUTS > PILOT / USE CASE 5

AGRICULTURAL, COPERNICUS DATA-BASED ANALYSIS: ADVANCED TECHNIQUES BASED ON AN

Agricultural analysis based on Copernicus data

Earth observation is a promising field with for helping to cope with integral problems such of sustainability of food production and environmental preservation.

Copernicus datasets, basis for satellite-based agricultural analysis: data from several Sentinel-family satellites producing radar and visible light data with up to 10m resolution.

Summary

This Use Case, in sum, is expected to make a contribution in:

Overall: to enable the users to make the most of PROCESS architecture, thanks to the open, generic APIs developed.
Agricultural analysis: to refine, steer and validate the behaviour of any Earth System modelling system in much more flexible way than today, with specific advances such as:
- Improvement of models (i.e., identify deviations of the Earth System model outputs).
- Fast in-depth diagnosis (i.e., predict the relative importance of upcoming observations in a specific application like forestry or biodiversity).
- Focus on ad hoc analysis (i.e., separate the human impact from natural system’s baseline development).

Challenges

Copernicus’ huge amount of information (3PB/month from radar data and 4,5PB from visible light data) also leads to a multi-disciplinary problem requiring integrating simulation models.
It is necessary to treat land surface processes and human interventions within one model.

Objectives

Main goal of PROCESS for agricultural, Copernicus data-based analysis: an easy-to-use yet powerful solution for realistic simulation of natural processes and impacts of human interventions:

Enabling users to focus on analysis.
A trade off between few (but ) rigorous principles and maximum predictive power.

Methodology

Techniques and procedures

Based on modelling framework PROMET (Processes of Mass and Energy Transfer) as a validation tool for services interfacing Copernicus data sets.
Use of Multi-Model Simulations linked with observational data, in a combined modelling of the macroscale carbon cycle, water and energy inputs, key nutrients in the soil , plant metabolism and human behaviour.

PROCEDURE/workflow:

Configuration of input variables within the PROCESS portal (IEE).
Deployment of a PROMET run from the IEE on the LRZ infrastructure.
Stage-Out of the output data via LOBCDER for the end-user.

Specific resources for the Use Case

Ad hoc infrastructure as an evolution of PROCESS architecture

Tools: simultaneous use of processing and analysis elements:

Modelling
Simulation
Pattern recognition
Verification and dimension reduction for datasets’ management

Results

For the field of activity

A combination of solutions for final users is available:

A framework that allows the user to run its solution from the portal without exposing their source code.
A series of advanced techniques based on an agricultural analysis of Earth observation data available through a generic API and web-based interface to Copernicus data for:
- Tight coupling of the modelling applications
- Observational data
- Simulation results with associated metadata

For the entire project

A standardized multipurpose interface solution and connection to the closed sourcemake.
Stress-testing of advanced data analysis techniques to be applied to other fields.

References and outcomes

Documentation

Deliverables:

Partners:

partner7_agh_znk_wbr_pms-e1522082784742-n

PILOT / USE CASE 4

Ancillary pricing for airline revenue management

PILOT / USE CASE 1

Exascale learning on medical image data

AGRICULTURAL, COPERNICUS DATA-BASED ANALYSIS: ADVANCED TECHNIQUES BASED ON AN

Agricultural analysis based on Copernicus data

Summary

Challenges

Objectives

Methodology

Techniques and procedures

Specific resources for the Use Case

Results

For the field of activity

For the entire project

References and outcomes

Documentation

Partners:

PILOT / USE CASE 4

PILOT / USE CASE 1

Tell us about your project

Maximilian Höb

PROCESS Coordinator

Miguel Saldaña

PROCESS Communication Manager