Research & development

Applied research for practical climate resilience.

Green Resilience develops climate-risk models, indicators, and decision-support tools that connect scientific insight to operational action.

Explore collaboration Contact us

Research focus

Our R&D work focuses on turning weather, satellite, agronomic, operational, and climate data into practical tools for anticipation, adaptation, and resilience.

Research areas

Multi-hazard agricultural risk

Understanding how drought, heat, excess water, disease pressure, and crop development stages interact to affect production risk.

Crop disease forecasting

Developing weather-dependent indicators to anticipate conditions favourable to crop diseases.

Satellite and weather-data fusion

Combining forecasts, observations, satellite indicators, and field context to monitor evolving climate risks.

AI-supported recommendations

Using AI and structured rules to translate complex risk signals into clear recommendations, while maintaining expert oversight and guardrails.

Operational impact modelling

Connecting climate hazards to impacts on logistics, safety, productivity, maintenance, and continuity.

Strategic and financial risk analysis

Supporting organisations in understanding how climate risk affects planning, investment, insurance, and resilience decisions.

From research to product

Each step is designed so that scientific insight ends as something a decision-maker can act on.

1Research insight
2Indicator design
3Platform logic
4Alerts & recommendations
5Reports & decisions

R&D pathway

From model logic to operational use

Green Resilience R&D connects scientific indicators, sector-specific thresholds, and operational constraints to practical decision-support outputs.

01
Define the decision
Spray, delay, irrigate, monitor, escalate, close access, schedule maintenance, stress-test assets.
02
Select relevant indicators
Rainfall, heat, wind, humidity, soil moisture, drought, flood, disease-conducive conditions, visibility, or compound risk.
03
Configure thresholds
Thresholds reflect agronomic logic, operational constraints, safety rules, SOPs, or partner-defined tolerances.
04
Generate risk logic
Indicators are converted into risk scores, severity bands, operational windows, and advisory outputs.
05
Deliver and validate
Outputs are delivered through dashboards, alerts, APIs, WhatsApp, SMS, reports, or partner workflows, then refined through feedback and local validation.

Risk logic

How risk scores are built

Risk scores combine hazards, thresholds, persistence, compound conditions, and operational context.

1Risk signal inputs

Threshold exceedance

Examples: wind above a defined limit, rainfall above a defined amount, heat index above a safety threshold.

Rolling windows and persistence

Examples: rainfall accumulation over 6, 12, 24, or 72 hours; multi-day heat stress; dry spell duration.

Compound triggers

Examples: wind plus low humidity for drift or dust risk; rain plus antecedent wetness for access risk; temperature plus humidity for disease pressure.

2Severity classification

Severity bands and escalation

Low

Monitor or proceed

Moderate

Proceed with caution or prepare

High

Delay, adapt, or escalate

Critical

Stop exposed activity or escalate immediately

3Operational timing check

Operational windows

Examples: spray windows, access windows, safe work windows, maintenance windows, irrigation windows, logistics windows.

Best execution window

Caution window

Avoid window

Time-to-impact

4Decision-grade output

Decision-grade outputs

Go / caution / no-go guidance with short rationale, traceability, and lead-time framing.

Go Caution No-go

The final output should help users understand what to do, why, when, and whether escalation or human review is needed.

Responsible decision intelligence

Green Resilience should not provide overconfident recommendations where data is incomplete, ambiguous, outside supported scope, or insufficient. The system should distinguish between weather alerts, probable operational or agronomic implications, and confirmed field-level diagnoses.

Clear recommendation

Where confidence is high, the system provides concise and actionable guidance.

Cautious guidance

Where uncertainty exists, the system explains the main limiting factor.

Request for more information

Where context is missing, the system asks rather than guesses.

Human escalation

Where local observation or professional judgement is required, the system recommends escalation.

Collaboration and partnerships

Green Resilience connects climate-risk research, geospatial data, Earth observation, and operational decision-support through collaborations with research institutions, data ecosystems, and innovation partners.

Research collaborations

CIHEAM-IAMM

Agricultural systems, adaptation, and rural development research.

University of Pretoria

South African agricultural systems and climate vulnerability research.

TETIS

Remote sensing, spatial information, and environmental modelling.

Earth observation and climate data

Copernicus & Sentinel Programme

Earth observation and environmental monitoring data.

NASA & Landsat Program

Open satellite imagery for land and environmental monitoring.

Open Meteo

Weather forecast and environmental data APIs.

Technology infrastructure

Google Cloud

Cloud infrastructure for data processing and platform deployment.

Innovation ecosystem

ESA BIC Sud France

Space technology and Earth observation incubation support.