reporting with numbers

Climate

Constructing stories

1) Avoid using multiple figures and statistics in quick succession.

(A good rule of thumb is one per sentence — whether in a written story or a broadcast script).

DO THIS

A story on climate change in Texas includes the following paragraph: “2022 was one of the driest years on record for Texas, and about 49% of the state was still in drought conditions at the end of December. The drought resulted in failed crops, low yields for farmers and diminished grazing, which forced ranchers to cull their cattle and led to the highest amount of livestock sold — nearly 2.7 million — in more than a decade.”’

DON’T DO THIS

A story on COP27 includes the following sentence: “A recent study released by the World Bank said that climate-related events will squeeze more than 132 million people into poverty worldwide with African countries losing between 10 and 15 percent of their GDP by 2050.”

DISCUSSION

In climate reporting, numbers are often presented as if they are the story. This assumption frequently leads to jam-packed, quantitatively dense paragraphs that reference many different concepts. The second example points to two different economic indicators (poverty and GDP), and places figures for these in the same sentence even though they are not comparable. This is simply too dense for a single sentence. A better approach would be to divide this into two sentences – one focusing on the global poverty projection, and a second focusing on GDP losses in Africa.

2) Always explain where numbers or statistics came from

DO THIS

An article explains that heat pumps can cut carbon dioxide emissions “around 38 to 53 percent from home heating,” and cites the University of California study on which this estimate is based.

DON’T DO THIS

Another article discussing the impact of heat pumps states that a local organization’s distribution of heat pumps has “[kept] almost 8,000 tons of carbon dioxide out of our atmosphere,” but provides no indication as to where this data comes from.

DISCUSSION

Phrases like “according to a study by…” or “a new paper published in…” are helpful here, as they identify the source of a given data point and allow audiences to independently verify it.

3) When discussing research, don’t just summarize key findings and claims. Share evidence, and describe the methods researchers used.

DO THIS

An article discussing a study on carbon dioxide emissions devotes a paragraph to exploring the researchers’ methodology: “To do the study, first Callahan looked at how much carbon each nation emitted and what it means for global temperatures, using large climate models and simulating a world with that country’s carbon emissions, a version of the scientifically accepted attribution technique used for extreme weather events. He then connected that to economic studies that looked at the relationship between temperature rise and damage in each country.”

DON’T DO THIS

An article highlights the results of a study on ocean temperatures, but says nothing about how researchers determined that “oceans hit record temperatures in 2022.”

DISCUSSION

It may be tempting to write off discussion of research methodologies as being too boring or obscure for a news story. These insights, however, can help audiences understand how climate science works, and will give them opportunities to independently evaluate scientific claims and findings. The second example says nothing about how the researchers reached their conclusions, and leaves readers in the dark as to how one measures ocean temperatures. By contrast, the first example includes an effective discussion of the cited study’s methodology, and helps readers see how the researcher determined the economic impact of carbon dioxide emissions.

4) Include enough context to help your audience understand why new numbers or statistics are important.

DO THIS

An article citing a study finding that droughts in 2022 were “made 20 times more likely by climate change” includes the following explanation: “Researchers from World Weather Attribution, a group of scientists from around the world who study the link between extreme weather and climate change, say this type of drought would only happen once every 400 years across the Northern Hemisphere if not for human-caused climate change. Now they expect these conditions to repeat every 20 years, given how much the climate has warmed.”

DON’T DO THIS

An article exploring the worst weather disasters of 2022 cites the same study to say only that the year’s “widespread Northern Hemisphere drought … [became] 20 times more likely because of human-caused climate change.”

DISCUSSION

Numbers may draw some attention, but offering them without context does not allow the audience to understand what exactly that number says about a given climate-related development, or whether there has been a significant change. The second example above correctly summarizes the conclusion of the cited study, but does not really help audiences understand what “20 times more likely” means. By contrast, the first example (which discusses the same study) provides a baseline figure, noting how an event only expected to happen once every 400 years is now expected to happen “every 20 years.”

5) Present concrete, real-world examples of the effects of climate change.

DO THIS

A story discussing a report from the World Meteorological Organization states that “weather-related disasters have increased fivefold over the last 50 years and are killing 115 people per day on average – and the fallout is poised to worsen.”

DON’T DO THIS

A story discussing a report from the World Health Organization states that “currently…12.6 million people die globally due to pollution, extreme weather and climate-related disease. Climate change between 2030 and 2050 is expected to cause 250,000 additional global deaths, according to the WHO.”

DISCUSSION

The second example above highlights impacts on a scale so immense (“250,000 additional global deaths”) that it’s hard to actually grasp. By contrast, by reducing the scale to a daily average, the effects of climate change conveyed in the first example are much more visceral.

6) Use comparisons and analogies to help people understand what large numbers mean

DO THIS

A TV report preceding the UN COP 26 climate conference begins with four quick vignettes of people suffering through floods in Europe, monsoon rains in the Philippines, a heat wave in the Pacific Northwest, and destructive fires in California. The narration states: “Four lives among millions more distorted and lost this year alone from the impacts of climate change. A warming atmosphere isn't the sole cause of these disasters, but the evidence grows clearer every day that fossil fuel emissions make these calamities more frequent, more severe, more deadly.”

DON’T DO THIS

An article offers some “climate quick facts,” including: 1) “Human activities have raised the atmosphere’s carbon dioxide content by 50 percent in less than 200 years, says the National Aeronautics and Space Administration”; 2) “Without adaptive measures, the number of people who lack sufficient water for at least one month per year will soar from 3.6 billion today to more than 5 billion by 2050, says the UN.”

DISCUSSION

It is both easy and tempting to think of climate change as something that will only be of concern to future generations. But climate change’s effects are already evident all around us. To help audiences understand this, look for ways to draw attention to current, present-day evidence of climate change and its impacts. The first example does this rather effectively, using data from the Climate Shift Index (a tool that demonstrates how climate change is “already part of our daily lives”) to highlight ways that climate change has already altered daily temperatures. By contrast, while the information in the second example is accurate, its singular fixation on long-term trends (“50 percent in less than 200 years”) and future impacts (“5 billion by 2050”) might lead audiences to think that climate change is not a matter of direct importance in their day-to-day lives.

7) Always define key terms, and break these down into concepts that connect more easily to a general audience

DO THIS

A video segment after the 2018 Houston floods explains “A 100-year floodplain, seen here in blue, means that each year there is a one in 100 chance your home will be flooded. A 500-year floodplain, the area in green, means one in 500, or 0.2 percent chance for the area to be flooded in any year. If you live within these zones, you must purchase flood insurance, though it's at a government-subsidized rate."

DON’T DO THIS

An article recounting the history of California’s “megaflood” of 1862 defines “100-year flood protection” as a “standard” that the Federal Emergency Management Agency (FEMA) uses to make floodplain maps and flood insurance requirements, but says nothing about the assumptions that went into the construction of this standard.

DISCUSSION

Viewers may make incorrect assumptions about the meaning of terms that are often used by those measuring climate change and weather patterns. For example, they may think a 100 year flood can only happen once in 100 years. The first example gives the audience a better understanding of what this metric measures and why.

8) When reporting on global temperature rise for U.S. audiences, be sure to communicate data in both Celsius and Fahrenheit

DO THIS

A report discusses what is needed in order to “keep global temperature rise below 2 degrees Celsius (3.6 Fahrenheit) and ideally no more than 1.5 C (2.7 F) by the end of the century.”

DON’T DO THIS

An article reports that current global temperatures are “around 1.2 degrees Celsius warmer than temperatures before industrialization,” and notes that “the world must try to cap global warming to 1.5C to stave off some irreversible impacts of climate change.” Neither of these data points are converted to Fahrenheit.

DISCUSSION

Audiences in the U.S. tend not to have an intuitive feel for the Celsius scale, and for those outside of the U.S., the Fahrenheit scale may be hard to understand. Given this, the best practice is to present temperature data (especially the big 1.5 degrees Celsius / 2.7 degrees Fahrenheit figure that is our current threshold for limiting) in both Celsius and Fahrenheit.

9) Help audiences understand the assumptions that go into the production of climate science data and concepts. When these are problematic, point it out.

DO THIS

An article tracking increases in “the amount of rain falling in the biggest storms” between 1958 and 2014 contends that there are “issues with the 100-year flood standard,” and poses questions (for example, “How do you project a once-in-a-century calamity from a few decades of data?”) that prompt audiences to ponder the validity of this concept.

DON’T DO THIS

An article recounting the history of California’s “megaflood” of 1862 defines “100-year flood protection” as a “standard” that the Federal Emergency Management Agency (FEMA) uses to make floodplain maps and flood insurance requirements, but says nothing about the assumptions that went into the construction of this standard.

DISCUSSION

Scientific concepts might appear to be objective and real, but often, there’s a difference between the phenomena they seek to describe and nature itself. By calling attention to the tentativeness of scientific knowledge, you can help audiences realize the role that human values often play in the construction of this knowledge–and promote critical thinking about what “good” measurements or “good” data consist of. The first example’s questioning of the concept of the “100-year flood” is an excellent example of this. By contrast, the second example misses an opportunity to get readers thinking about how standards are set, and promotes a kind of uncritical acceptance of established terms and concepts.

10) When discussing projections, make sure to include baseline figures

DO THIS

An article on a rise in greenhouse gas emissions states: “Carbon dioxide levels rose by more than two parts per million (ppm) for the 11th consecutive year: the highest sustained rate of CO2 increases since monitoring began 65 years ago.”

DON’T DO THIS

An article states that new drilling in the Gulf of Mexico and Alaska will result in “more planet-warming carbon dioxide — up to 110 million tons (100 metric tons) annually — from U.S.-produced oil and gas by 2030.”

DISCUSSION

Without a benchmark or baseline, audiences won’t be able to understand how expected developments in the future compare to present-day realities. The second example omits any reference to contemporary data, which makes it hard to know exactly how big “110 million tons” is. To make sense of this, readers would need to know how this figure compares to a typical year, or how much future drilling might reverse progress on emissions reduction.

11) When discussing models, give audiences the information they need to understand how they work

DO THIS

An article citing a study finding that droughts in 2022 were “made 20 times more likely by climate change” includes the following explanation: “Researchers from World Weather Attribution, a group of scientists from around the world who study the link between extreme weather and climate change, say this type of drought would only happen once every 400 years across the Northern Hemisphere if not for human-caused climate change. Now they expect these conditions to repeat every 20 years, given how much the climate has warmed.

DON’T DO THIS

An article discussing the phenomenon of “weather weirding” states that “climate models make clear more is coming,” but says nothing about these models or how they make projections.

DISCUSSION

Models are an incredibly important part of climate science, which means referencing them is an unavoidable part of climate-based reporting. In the first example above, the article explains how the model was created, what the changes are and how that compares to the past. The second example, on the other hand, is vague about what model is changing and why. When discussing models, it’s important to help audiences understand how they translate past or present data into estimates for the future. Without this information, audiences will be unable to understand where predictions come from and what they mean in practical terms. When models are too complex to explain, consider providing links or sidebars (for example, https://www.climate.gov/maps-data/climate-data-primer/predicting-climate/climate-models) that give audiences the necessary information.

12) Don’t overinterpret — highlight uncertainties and things we don’t yet know about.

DO THIS

An article observing that New York City has “gone 321 consecutive days without snowfall” quotes a scientist who explains that this phenomenon “is not solely attributable to global heating.”

DON’T DO THIS

An article observes that “if we make it to 333 days with no measurable NYC snow that’s the longest streak ever,” and quotes a climatologist who states that “we’re starting to see in the last 20 years a consistent warming that tells us that there is a climate issue here."

DISCUSSION

Given how many and varied the impacts of climate change are, it can be tempting to attribute almost any weather-related event to it – especially record-breaking events like New York City’s consecutive snowless days streak. But much can influence meteorological phenomena, and it’s important to let audiences know that multiple factors can produce weather patterns. The first example above explains how a combination of rising temperatures and La Niña wind patterns are working together to keep snow out of New York City. By contrast, the second example lends the impression that climate change is single-handedly responsible for the city’s incredible run of snowless days.

13) Remind audiences that climate and weather are different things

DO THIS

An article on climate change in Kentucky makes a distinction between long-term impacts like increasing annual rainfall amounts and event-specific impacts, clarifying that when it comes to floods and other kinds of natural disasters, “determining whether or not they are being affected by climate change is more difficult.

DON’T DO THIS

Media coverage of California’s weather in early 2023 was “quick to link a series of powerful storms to climate change,” as were some officials. But researchers told the media outlet afterwards “they had yet to see evidence of that connection.”

DISCUSSION

The terms “weather” and “climate” are often used interchangeably, but in fact, they are very different things. Weather has to do with short-term phenomena–like what you see when you look out your window. It refers to day-to-day changes in temperature, precipitation, and the like. By contrast, climate has to do with long-term trends and averages – like, for example, average rainfall amounts in a single area over the course of multiple decades. There is no way to establish clear cause-and-effect relationships between climate change and individual weather events but we can attribute general, long-term meteorological trends to climate change.

Failure to distinguish between weather and climate can make audiences confused about what climate change actually means, and also has the potential to fuel climate change denialism. But when experts have made that connection, be sure to report their findings and why they made that call.

In this dispatch from California during a 2021 heatwave, for instance, the reporter describes the concept of "attribution science," and mentions how a European meteorological agency had linked a recent deadly heat-wave in the Pacific NW specifically to climate change. TV news coverage of historic rainstorms in California repeatedly failed to link these kinds of events to climate change, according to an analysis conducted by Media Matters.

Provide feedback.