Hadi Farahani, PhD

Independent environmental data consultant

Environmental monitoring data you can defend.

I help environmental, compliance and operations teams draw reliable conclusions from measurement data: QA/QC, detection limits, measurement uncertainty, compliance analysis and risk maps. Analytics, GIS and reporting built on analytical chemistry rigor, delivered as fixed-scope, reproducible projects.

Taking 1–2 projects at a time · any industry · remote

Send me your problem View services

Try it: is this result compliant?

Decision explorer: measured value, expanded uncertainty and a limit A measured value with an uncertainty interval is plotted against a limit of 100 mg/L. Sliders change the value and the uncertainty, and the verdict updates. Limit 100 mg/L 406080100120140160 Concentration (mg/L)

Inconclusive: the interval straddles the limit. A decision rule is needed.

SYNTHETIC Probability the true value is below the limit: 84%. Simplified: normal distribution, no prior information. Decision rules follow JCGM 106 / ILAC-G8.

Services

Fixed scope, written deliverables, reproducible code. Pricing is quoted per project after a short conversation.

Which fits you?

Pick one and I’ll highlight the best starting point.

Scoping sprint

A few days, fixed fee. A data-readiness check and a written plan with scope, timeline and price for the full project.

2–4 page memo · fee credited to the project

Data quality & uncertainty audit

Review of a monitoring dataset: QA/QC, non-detects, detection-limit changes, and a defensible uncertainty statement.

Typically 1–2 weeks

Compliance-ready analysis & report

Trend, exceedance and decision-rule analysis delivered as a reproducible report with code.

Typically 2–4 weeks

Geospatial risk map

QGIS and Python mapping of exposure, sensitivity and risk, with documented sources and assumptions.

Typically 2–3 weeks

Case studies

Each study uses synthetic data with known truth, so methods can be scored against the right answer. A study is marked Published only when its data, code and report are in the repository.

Flagship · in preparation

OGE-06: Groundwater quality trends

Censored data, changing detection limits that create false trends, and a coverage check showing whether stated uncertainty holds against the known truth.

EnvIntel method catalogue (25 planned studies)
  • AData quality & metrology (5)
  • BIndustrial operations (5)
  • CGeospatial & Earth observation (5)
  • DRisk & forecasting (5)
  • ECompliance & governance (5)

Planned. Individual studies are published as they are completed, roughly one per month.

How I work

Validate

Establish reliability first. Missing values, anomalies, drift, reporting limits and provenance are checked before any analysis.

Quantify uncertainty

Say how sure you are. Measurement uncertainty and censored-data methods, following the GUM (JCGM 100/101).

Decide

Turn evidence into action. Explicit decision rules (JCGM 106, ILAC-G8) and findings written for the people who must act.

Confidentiality: client data stays under NDA. I do not put client data into external AI tools without written consent, and every AI-assisted step is checked by me.

Background

PhD in analytical chemistry, with experience in research and industrial environmental projects. Peer-reviewed work is listed on Google Scholar. I also share a short insight every weekday on LinkedIn.

Publications LinkedIn GitHub
Python (pandas, SciPy, statsmodels)QGIS & rasterioSQLiteQuartoPowerShellExcelGitGUM / JCGM 100, 101, 106

Tell me about your data problem.

A short description is enough. I reply within two working days with questions or a proposed next step.

Email me