Climate +
data journalism
Using data to recognise evidence, question it, and sharpen climate stories.
Areena Arora
Who am I?
Editorial coder and data journalist working across reporting, code, information design, and visual storytelling.
SESSION ONE
Data is not a software skill first. It is a reporting skill.
Journalistic judgement and inquiry matter more than software know-how.
Start with what the number means.
- What is being counted?
- Who is missing?
- Who benefits from the official definition?
Climate data is not just temperature.
Climate journalism is about physical change, human exposure, public decisions, money, institutions, and unequal consequences.
- Hazards
- Exposure
- Vulnerability
- Official response
A temperature reading does not automatically tell us who was at risk.
A heat story needs an inventory.
Temperature records
Spatial and satellite data
Administrative records
Budgets and procurement
Regulatory records
Health, labour, and livelihoods
Community evidence matters.
Local rain gauges, fisher logs, farmer diaries, panchayat registers, photographs, oral histories, school attendance records, and other crowd sourced data can aid further verification.
What does the official system fail to record?
Count
→Compare
→Locate
Test
→Hold accountable
Count
How many heatwave days occurred? How many claims were approved? How many monitoring stations reported usable data?
Compare
How does this district compare with its past, its neighbours, or a relevant baseline?
Locate
Where is the change concentrated? A map or district table can identify places for field reporting.
Test
Does the evidence support an official claim?
Hold accountable
Who had responsibility, money, authority, or prior warning?
From weather observation
to public-interest story
The story moves from “heat happened” to “protection was unequal.”
“The Decoding Heat series explores shifts in broad and local weather systems; cascading patterns within ecosystems and urban areas; how India is forecasting and planning for its heatwaves, and more. Through our storytelling, we attempt to answer: what drives and is driven by a warming country?”
Minimum temperatures are rising across the country with many states and UTs seeing warmer nights.
Many parts of India are witnessing unusually high minimum temperatures, with several locations recording minimum temperatures close to 30°C. In the last week of May, IMD observations indicated that minimum night temperatures were warmer than normal by 3.1°C to 5.0°C.
- What was the central journalistic question?
- What datasets or records were combined?
- What reporting was still required beyond the data?
Common pitfalls
- Weather is not automatically climate.
- Averages can hide extremes.
- Geographies may not match.
- Raw totals can mislead.
- Official records can undercount.
More pitfalls
- Missing data is often patterned.
- Correlation does not establish cause.
- Short periods invite overclaiming.
- Modelled and observed data are different.
- False precision weakens trust.
Exercise
Choose one theme: heat, floods, water, health, agriculture, energy, law, or climate finance.
- Write one broad statement common in coverage.
- Turn it into a data-led question about unequal impact.
- Turn it into a data-led question that tests accountability.
Data does not replace reporting.
It can tell us what to count, where to look, whose claim to test, and who may be missing.
SESSION TWO
Finding and inspecting
climate data
Before analysis, establish what the dataset is, where it came from, what each row represents, and what it leaves out.
Inspection workflow
Reliable does not mean
unquestionable.
It means the source is identifiable, the method is discoverable, and the records can be traced. Official data is essential, but the administrative purpose of a dataset may differ from the journalist's question.
Starting points for
Indian climate reporting
Demonstration dataset
State-wise location based installed capacity of Renewable Power as on 30 June 2026.
First, source and scope.
- The source is the Ministry of New and Renewable Energy.
- The table is a snapshot as of 30 June 2026, not a time series.
- One row represents a state or Union Territory, plus an Others row.
- The title says location based: capacity is assigned where the project is physically located.
Second, definitions.
- Installed capacity is the maximum output equipment is designed to produce.
- Capacity is measured in megawatts. It is not electricity actually generated.
- Rooftop solar includes PM Surya Ghar Yojana capacity.
- Total renewable energy depends on which technologies the ministry includes.
Third, structure and quality.
The table is wide: each geography is a row and each technology is a column. Several component columns feed into subtotal columns, so do not sum both a subtotal and its components. Blank cells and zeroes require care.
Fourth, comparison.
Fifth, claim.
The table can show how state renewable-energy rankings change depending on whether leadership is measured by total megawatts, technology mix, rooftop adoption or the inclusion of large hydro. It cannot show generation, consumption, coal displacement, reliability or household benefit.
Group exercise
Debrief
- What does one row represent?
- Which definition most affected your result?
- What is the strongest pattern you found?
- What is one claim the dataset clearly supports?
- What additional dataset, document, expert or interview would be needed before publication?
SESSION THREE
Communicating
climate data clearly
Correct analysis can still fail if the writing buries the finding, leaves a technical term undefined, or turns installed capacity into a claim about electricity produced.
Communication test
Explain your Session 2 finding without using the words dataset, spreadsheet, analysis, correlation or significant.
- What did your listener think was measured?
- Did they confuse capacity with generation or consumption?
- Did they understand whether large hydro was included?
Write without
overwhelming readers.
Lead with the finding.
An analysis of state-wise renewable-energy data found substantial differences across India.
Gujarat and Rajasthan lead India in installed renewable capacity, but the list changes sharply when large hydropower is excluded and rooftop solar is considered separately.
Use the exact measure.
- The table reports installed capacity in megawatts.
- Do not write that a state produced, supplied or consumed that amount of renewable electricity.
- Separate observation, interpretation and cause.
- Each step requires different evidence.
Five-sentence structure
Individual writing exercise
Write a short story of no more than 300 words using your group's evidence from Session 2. Imagine it is the opening of a reported climate or energy story, not a classroom answer.
- Headline that does not overclaim.
- First sentence that leads with the finding.
- One meaningful comparison, one definition, one caveat and one next reporting step.
Peer editing checklist
- Mark one sentence that is immediately clear.
- Mark one number that needs more context.
- Underline where capacity is confused with generation or consumption.
- Circle any technical term that needs a definition.
- Mark one claim that may overreach.
Model opening
India's renewable-energy leaders look very different depending on what is counted. Gujarat and Rajasthan have the country's largest location-based renewable portfolios in the Ministry of New and Renewable Energy's June 2026 table, driven mainly by solar and wind. But Himalayan states rise sharply when large hydropower is included, while states with smaller overall totals can perform more strongly on rooftop solar as a share of their solar capacity.
Choosing a visual form
The form follows the relationship in the data.
A visual should answer the reporting question faster or more clearly than prose. Do not choose a map, line chart or complex form just because it looks more impressive.
Which chart, when?
For this dataset
- A ranked bar chart can compare total renewable-energy capacity.
- A slope chart can show how rankings change when large hydro is excluded.
- A stacked bar chart can compare technology mix.
- A dot plot can compare rooftop solar as a share of total solar capacity.
Visual pitfalls
- Capacity is not generation.
- Totals can hide definitions.
- Subtotals can be double-counted.
- Raw totals can reward large systems.
- Missing is not zero.
More visual pitfalls
- A choropleth can reward large areas.
- Colour can manufacture drama.
- One-date snapshots are not trends.
- Location is not consumption.
- Mobile readers need hierarchy.
Group visual exercise
Choose one visual form for your Session 2 finding and sketch it on paper.
- Label the title, axes or columns, unit and source.
- State whether large hydro is included.
- Add one annotation.
- Write one sentence explaining why this form is better than the alternatives.
Close
Definitions determine the story.
Ask what is being measured, how the source defines it, what comparison gives it meaning, what the data cannot establish, and who is responsible. Then show your work clearly enough that a reader, editor or affected community can challenge it.