Claim: Climate change caused the devastating floods in part of Brazilian´s Southeast region, study says

Torrential rain made over 90,000 people homeless in Minas Gerais state, where the probability of far higher volumes of rain than expected has increased 70% owing to industrialization and global warming.

Peer-Reviewed Publication


Climate change was the main cause of the extremely heavy rainfall that led to severe flooding, deaths and massive damage in cities across Minas Gerais state, Southeast Brazil, in January 2020, according to a study published in the journal Climate Resilience and Sustainability.

Using climate modeling for the region, the study shows that the effects of industrialization and global warming increased the probability of far higher volumes of rain than expected by 70% compared to scenarios with average temperatures 1 °C -1.1 °C lower.

The authors of the study also quantified the damage caused by the extreme event, estimating that over 90,000 people were made temporarily homeless, and at least BRL 1.3 billion (USD 240 million) was lost in the public and private sectors. Most of the material damage was to public infrastructure (BRL 484 million), housing (BRL 352 million), and retail stores and other services (BRL 290 million). Human-induced climate change is blamed for 41% of the total.

The article was published a week after the Intergovernmental Panel on Climate Change (IPCC) stressed in its sixth assessment report that “it is unequivocal that human influence has warmed the atmosphere, ocean and land”, contributing to rapid intense changes in all regions of the planet.

The extreme precipitation event that occurred in Southeast Brazil on January 23-25, 2020, led to floods and landslides, causing damage to infrastructure and deaths. The event resulted from an intensified South Atlantic Convergence Zone (SACZ) combined with the emergence of the Kurumí subtropical cyclone (KSC) over the South Atlantic. Both phenomena increased moisture content throughout the region.

Belo Horizonte, the capital of Minas Gerais, had the wettest January in its history. According to the National Institute of Meteorology (INMET), rainfall totaled 935.2 millimeters (mm), or almost three times the average for the month, with 320.9 mm falling in three days. At least 56 deaths were considered due to flooding and landslides at the time.

The authors emphasize that the study offers new insights regarding the urgent need to take action on climate change, which is already having a significant impact on society in Southeast Brazil: “This calls for immediate improvements in strategic planning to focus on mitigation and adaptation,” they write, adding that public management and policies must evolve from disaster response to include prevention of future disasters. 

The study was supported by FAPESP and derived from a workshop led among others by Sarah Sparrow, a researcher at the University of Oxford (UK) and last author of the article. The workshop was sponsored by the Climate Science for Service Partnership (CSSP Brazil), a collaboration between UK institutions and organizations in Brazil, including the National Space Research Institute (INPE), the Natural Disaster Surveillance and Early Warning Center (CEMADEN), and the University of São Paulo (USP).

Held online in December 2020 in partnership with Liana Anderson, a researcher at CEMADEN and penultimate author of the article, the workshop discussed a method known as extreme event attribution that aims to quantify the potential effects of human-induced climate change on the probability of extreme weather events.

Two working groups analyzed the Minas Gerais extreme precipitation event. One focused on the influence of climate change on rainfall, while the other quantified the impacts on the population. The findings of both groups were integrated in the published article.

“Close interdisciplinary collaboration enabled us to produce a high-quality study with consistent results in only a few months,” said Ricardo Dal’Agnol, a researcher in INPE’s Earth Observation and Geoinformatics Division and first author of the article.


The global climate model used for attribution was the Hadley Centre Global Environmental Model version 3-A (HadGEM3-A), with simulations of extreme weather events. Two experiments were conducted to help design scenarios, one considering only natural factors such as variations in solar irradiance and volcanic activity, and the other also considering anthropogenic factors, such as land-use changes and greenhouse gas emissions compared with the pre-industrial level (1850).

According to the latest IPCC report, the planet’s average temperature is 1.1 °C higher than in the period 1850-1900, the reference period used to approximate pre-industrial temperatures before the sharp rise in emissions of greenhouse gases such as CO2 and methane.

To understand the spatial distribution of daily precipitation in the study area and estimate precipitation for the attribution analysis, the researchers used the CPC Global Unified Gauge-Based Analysis of Daily Precipitation, and the Climate Hazards Group InfraRed Precipitation with Station Data (CHIRPS), which integrates satellite imagery and rain gauge station data to create time series for trend analysis and monitoring.

Data on disasters, including location, type, causes, and damage, was extracted from Brazil’s S2iD Integrated Disaster Information System, which holds data recorded by Civil Defense and local government staff within ten days of an occurrence. The authors “highlight the importance of having integrated disaster information systems such as the Brazilian S2iD, which conveys valuable and timely information that allows quantification of the impacts from extreme events”.

The study area in southeastern Minas Gerais was subdivided into 12 mesoregions (official units comprising a group of cities that share geographical and societal features), with a total of 194 municipalities. The state has 853 cities all told. “The most affected mesoregions were metropolitan Belo Horizonte, Vale do Rio Doce and Zona da Mata. Together they accounted for 91% of public economic losses and 93% of private economic losses, 92% of total material damage [and] 91% of the total displaced population,” the authors write, adding that these areas also displayed “the most alarming numbers regarding vulnerability to disasters of residents and dwellings in officially mapped risk areas”.

They also note that although the event was extreme and influenced by climate change, its impacts were exacerbated by lack of urban risk management planning, mitigation and adaptation strategies, as well as underinvestment in infrastructure, and that it may have disproportionately affected the poor living in high-risk situations such as precarious hillside housing. 

“We, therefore, interpret the impacts of this event as a socially constructed climate disaster,” they write, suggesting that future studies should investigate the impact of extreme weather events on poor and vulnerable people. “Moreover, future research could also address the increasingly complicated interactions of human, economic and political aspects within ecological systems,” they add.

According to Dal’Agnol, the model developed to analyze the Minas Gerais disaster can be applied to other regions. “We used scenarios based on the model and satellite rainfall data to estimate probabilities,” he said. “The methodology can be used to analyze other events. At the time of our study, we found little research on extreme weather events in Brazil. More studies like this one are needed to identify regions that are particularly vulnerable to climate change so that government and public policy can be properly prepared to prevent future disasters.”

FAPESP supported the study by awarding scholarships to Dal’Agnol (19/21662-8), Carolina Barnez Gramcianinov (20/01416-0), and Márcia Marques (19/17304-9). 


About São Paulo Research Foundation (FAPESP)

The São Paulo Research Foundation (FAPESP) is a public institution with the mission of supporting scientific research in all fields of knowledge by awarding scholarships, fellowships and grants to investigators linked with higher education and research institutions in the State of São Paulo, Brazil. FAPESP is aware that the very best research can only be done by working with the best researchers internationally. Therefore, it has established partnerships with funding agencies, higher education, private companies, and research organizations in other countries known for the quality of their research and has been encouraging scientists funded by its grants to further develop their international collaboration. You can learn more about FAPESP at and visit FAPESP news agency at to keep updated with the latest scientific breakthroughs FAPESP helps achieve through its many programs, awards and research centers. You may also subscribe to FAPESP news agency at


Climate Resilience and Sustainability




The article “Extreme rainfall and its impacts in the Brazilian Minas Gerais state in January 2020: Can we blame climate change?”



From EurekAlert!

2 8 votes
Article Rating
Newest Most Voted
Inline Feedbacks
View all comments
October 19, 2021 2:10 pm

Climate change (aka 2mm rise in Atlantic sea surface) woke up this beast that is determined to join Europe to Americas.
One crater is normal, two craters might be a coincidence but three craters is a very bad omen.

Phil R
Reply to  Vuk
October 19, 2021 6:57 pm

And four craters signals the end of the world.

J Mac
October 19, 2021 2:14 pm

Bottom line: Climate change caused it to rain more than average!

Every hard rain is now blamed on ‘climate change’? Ugh….

Reply to  J Mac
October 19, 2021 3:00 pm

According to several of our trolls, if anything is different compared to last year, CO2 caused it.

October 19, 2021 2:20 pm

There are two proposals that need to be addressed.
a) the floods being caused by climate change
b) climate change is anthropogenic caused by increased carbon dioxide
As always the logic contained in these sort of claim is a) is absolutely true AND b) is absolutely true;
whereas the reality is more subtle than that
a) may be true (personally I don’t know, I am not a climatologist or meteorologist) or could be completely false.
b) is not absolutely true but could be absolutely false.
There needs to be more analogue thinking rather than digital thinking.
Also average implies sometimes there will be less rain and sometimes there will be more. There may be the very occasional time when there’s almost zero or there’s a very large amount. Who knows what is “normal”?

Last edited 1 year ago by JohnC
Reply to  JohnC
October 19, 2021 3:02 pm

Since CO2 and H2O have overlapping absorption bands, it is impossible for CO2 to have much, if any, impact on areas that already have lots of water in the air.

Reply to  MarkW
October 19, 2021 11:53 pm

I think you may have misunderstood what I was getting at, which is that the reasons as described in the article or pretty much every similar article is that two proposals are incorrectly conflated as a binary AND truth table. When the reality is a couple of separate probabilities that may overlap indicating a possible cause and effect. If that makes sense.

Reply to  JohnC
October 19, 2021 6:28 pm

JohnC:There may be the very occasional time when there’s almost zero or there’s a very large amount. Who knows what is “normal”?”

You seem to know what normal is because you just gave an accurate description of it

Normal is sometimes near zero rain and sometimes a deluge, with all points points in between represented. The better terms are ‘usual’ and ‘unusual.’

Pick anywhere you like and plot a histogram of the annual rainfall. Let’s say the usual total for the place you pick is between 25 and 65 inches of rainfall per year.

But then there will be unusual years towards each end of the histogram. There will be values that some people will never see in their lifetime, say 110 inches, but it happens. And if the records are long enough, that 110 inches will crop up every now and then. It may be years and years, but it will show up again.

I’m OK with ‘normal’ being used for the values in the broad central usual range. If it’s 25 inches of rain one year and 65 inches the next, then that’s ‘normal’ or a usual value. But 110 inches is unusual... but normally you’ll see that every century or so.

Average does not equal normal. Normal is the usual range.

Reply to  H.R.
October 19, 2021 11:58 pm

The difficulty is that people have forgotten or not realised the meaning of normal has a statistical meaning. Assuming a Gaussian distribution rather than a Fermi-Dirac or other distribution. Doesn’t normal mean the range defined by the mean or maybe median +/- a number of standard deviations, it’s a long time since I did any statistics!?

Reply to  JohnC
October 20, 2021 4:43 am

You’re not missing a thing, JohnC.

TV and radio weather segments have been using ‘normal’ for quite some time now. It’s a better word to use in place of ‘average’ if you’re selling a narrative.

WUWT has a lot of meteorologists reading and commenting here and they have pointed out that in days of yore, many media meteorologists used ‘average’ for temperature and rainfall. It hasn’t come up recently, but some of them get really wound up about the use of ‘normal’.

So you’d hear, “The high was 75(F) today, which is a little above the average of 71(F) for this day. The record high was 98(F) back on such-and-such date and the record low was 47(F) back on some other date.”

Hmmm… that doesn’t sound very alarming and it gives people a good idea of just what that 75(F) means. So the message is that there was nothing unusual about that high for the day

Now, most reports will use ‘normal’ and 75(F) is “hotter than normal” implying that we’re all gonna fry and we’re all gonna die. Using ‘normal’ instead of average makes any temperature that doesn’t hit right on the average seem to be ‘abnormal’.

Using ‘normal’ instead of ‘average’ and giving a range helps sell the narrative that the weather we’re seeing is ‘abnormal’ because… Climate Change and humans are responsible.

I’m a retired engineer and what I’ve been discussing with you is what I’ve learned here on WUWT from our participating meteorologists. I’m no expert on the ‘when’ and ‘why’ the change to using ‘normal’ instead of average occurred.

Keep your eyes peeled for the next time meteorologists here weigh in on the topic of how to report on weather. That’s always a “hotter than normal” discussion 😜

very old white guy
Reply to  JohnC
October 20, 2021 6:43 am

What is normal is what yesterday was like, what today is like and what tomorrow will be like.

Michael E McHenry
October 19, 2021 2:21 pm

I did a search of the NY Times archives 1929-1941 (a warming period) for BRAZIL FLOODS. 105 hits with lots of devastating floods and loss of life.

Rud Istvan
Reply to  Michael E McHenry
October 19, 2021 2:31 pm

In that period, the warming was natural, even according to IPCC. Just not enough change in CO2. QED.

Michael E McHenry
Reply to  Rud Istvan
October 19, 2021 2:48 pm

The point is heavy rains in Brazil is not new. It’s not new nor anything to do with asserted CO2 induce climate change. So the study is bogus if not fraudulent as they failed to do historical research

Rud Istvan
Reply to  Michael E McHenry
October 19, 2021 3:15 pm

Well, they have a downscaled GCM, so don’t need any historical facts. /s Besides, playing computer games is much easier than doing actual historical research.

Alan the Brit
Reply to  Rud Istvan
October 20, 2021 12:24 am

I gave up reading after the sentence;

“Using climate modeling for the region, the study shows that the effects of industrialization and global warming increased the probability of far higher volumes of rain than expected by 70% compared to scenarios with average temperatures 1 °C -1.1 °C lower.”

Any time I read “modelling”, “our models show”, “the modelling shows”, etc. I roll my eyes heavenwards in despair. Modelling will ALWAYS show what the model was programmed to show, based upon the fact that the programmers need to show a given amount of warming for a given amount of CO2 based on an assumed climate sensitivity to it in the atmosphere!!! There is no way a computer model can think for itself & question its output!!! I have always told graduates I have worked with, who just loved the wizardry of there puter programmes for structural analysis, “do a rough hand calc first to get a feel for the output from the printer, so that you can see when “hey, that beam/column is way too big/small for that location, etc!” These climate modellers have NO feel for what is right or wrong in their outputs, they simply swallow that output hook, line, & sinker, because it must be right we programmed the model to show that!!!

PS Griffy baby, still awaiting a response from my request for you to justify why a few million years ago, when there was around 19 times as much CO2 in the atmosphere than today, yet the Earth was smack bang in the middle of an Ice-Age!!! Why is it that modern climatificky scientists ignore the paleo-climatic record in favour of their models???

Rud Istvan
October 19, 2021 2:29 pm

There is just one problem. It has been know for at least a decade that neither of the GCM regional downscaling methods works well, if at all. Technical Explanations in essay Last Cup of Coffee in ebook Blowing Smoke. More alarmist nonsense in runup to COP26 failure.

John Harrison
October 19, 2021 2:41 pm

To determine the percentage change in energy content of the atmosphere would require, in my understanding, the percentage change in temperature on the absolute scale. Taking the ball park figure of 300K in 1850 and an increase of 1K in that time gives a percentage increase in energy content of approximately 0.33% in a century and a half. Now that is, just recently, supposed to be responsible for massive disturbances in precipitation, temperature distribution, wind speeds and directions and, in fact, in all things weather related. Does this seem likely or even possible? Is there some fault in my logic?

Coeur de Lion
October 19, 2021 2:45 pm

Wasn’t there unprecedented snow is Brazil the other day?

Rud Istvan
Reply to  Coeur de Lion
October 19, 2021 3:16 pm

Yes, plus killer frost in the coffee belt.

Alan the Brit
Reply to  Coeur de Lion
October 20, 2021 12:28 am

Snow in Brazil is not news, it has happened before a few years back. However, we must simply remember that CO2 warms & cools everything whenever it is required to do so – FACT!!! I wonder, during COP26, will they de-fizz the bubbly before consumption, or will they allow there bodies to diffuse it naturally???

Reply to  Coeur de Lion
October 20, 2021 5:59 am

Yep, and it has been raining heavily for the last 2 weeks with more flooding (we live right next to Minas on the coast).
the problem i have with the study is this,

using a global average increase to study a local effect is not science.
So until someone can explain to me exactly how a warmer arctic causes more rain in the subtropics no the other hemisphere(!) i’ll ignore this ‘study’, but it will no doubt end up being counted in the 99% consensus.

Brazil has many issues but increasing the energy price will not solve any of them. Rather the reverse actually.


October 19, 2021 2:45 pm

Dismal attribution “science”. My theory is that coins are being biased towards heads, this makes long sequences of heads much more likely, hence this particular long sequence of heads was caused by the bias. Works just as well for the alternate theory of a bias towards tails.

October 19, 2021 2:56 pm

The areas affected seem to be rapidly growing industrial areas, creating a significant UHI effect. Even so, for Belo Horizonte (a) the record is like many for this area truncated and (b) the unadjusted data shows little evidence of dramatic warming.

I am unpersuaded by the argument that these floods reflect climate change as opposed to some periodic shift in rainfall patterns.

Reply to  Bernie1815
October 19, 2021 3:05 pm

Not just UHI, but lots and lots of roofs and paved areas. I’d be surprised if there are any retention ponds in the region.
As a result of the development, run-off is much faster and of greater volume, since in the past some of the water would have been absorbed by the soil and clung to the leaves of trees.

October 19, 2021 2:57 pm

Climate change causes rain in a rain forest.


Alan the Brit
Reply to  MarkW
October 20, 2021 12:29 am

How peculiar!!!

October 19, 2021 2:59 pm

CO2 overlaps the absorption bands of H2O. It is impossible for CO2 to have any impact on temperatures in places that have lots of water in the air.

October 19, 2021 3:00 pm

Isn’t Barazil the country with rain forests ?
And burned areas of that forest will be a reason for landslides f.e. when it rains stronger.

Last edited 1 year ago by Krishna Gans
CD in Wisconsin
October 19, 2021 3:05 pm

“Most of the material damage was to public infrastructure (BRL 484 million), housing (BRL 352 million), and retail stores and other services (BRL 290 million). Human-induced climate change is blamed for 41% of the total.”


Seriously? Are they kidding me? 41% and not 42%? Not 39%? Models are a religion, it’s all a matter of faith.

Geoff Sherrington
October 19, 2021 4:14 pm

So, there was heavy rain somewhere, not a new phenomenon. There is only so much rain to go around, so another place had less rain than usual. Expect another pseudo science attribution paper about a drought caused by climate change.
Sorry for my use of simple logic, but the human brain performs way better than a big computer programme. Geoff S

October 19, 2021 4:42 pm

Original Post at least reports the greater impact upon poor living on hillsides, a feature of Brazilian life – as is the possibility in Brazil of high rainfall events to prompt flooding.

October 19, 2021 5:39 pm

more probably the deforestation and run run off

all cities direct rain water to drainage systems that go to the river
once upon a time before hard surfaces the rain was slowed down in forest litter and by the grasslands etc so it allowed water to be mitigated
now we direct it to narrow waterways (people build on the river flats) that get caught behind man made things like culvets

Just look at the Queensland floods etc

again hysterical media who are nothing but talking heads

Mike Lowe
October 19, 2021 9:21 pm

“Using climate modeling for the region….” No point in reading any further!

October 19, 2021 11:08 pm

“it is unequivocal that human influence has warmed the atmosphere, ocean and land”

I think most people here would agree with this statement, however, it’s the degree of warming and the cause which is the issue.

Deforestation, urban concrete/roofs/roads, yes, CO2, not so much

IIRC, a study from Manchester University approx 15 years ago found the way to mitigate any warming (not required in Manchester I would have thought 😉 ) is to introduce green roofs.

The study stated in Brazil the introduction of green roofs could reduce local temperatures up to 10C, plus green roofs clean the air, attenuate rainwater and have a feel-good factor

I’d be happy to mitigate the global warming “crises” with a few sedum roofs.

David Guy-Johnson
October 19, 2021 11:40 pm

One word. Garbage

Peta of Newark
October 20, 2021 1:06 am

No matter
It will all stop when they’ve finished making the desert

Alasdair Fairbairn
October 20, 2021 2:47 am

Just another POLITICAL paper devoid of scientific ethics designed to be published in the run up to COP26.

Ed Zuiderwijk
October 20, 2021 3:13 am

Correction: peer-approved, not peer-reviewed.

October 20, 2021 5:01 am

So, morons building towns and cities in the middle of flood plains has nothing to do with the increase in damage from regular, normal flooding in the tropics? Got it.

very old white guy
October 20, 2021 6:42 am

Really good prep is never live on a flood plain.

October 20, 2021 6:44 am

I take it “extreme event attribution” is the new game where you start with a weather related disaster and the first one to come up with a reason to blame it on Climate Change wins.

Reply to  George Daddis
October 20, 2021 7:59 am

Yes … so this is how you play the game you find an event lets say Brazils covid 19 deaths and then you just make a claim like so.

“All Brazil’s covid deaths were caused by climate change”

Now anyone that disagrees with you is a denier because that is the rules of the game and that is now a climate change fact.

Last edited 1 year ago by LdB
October 20, 2021 7:21 am

The damage resulted from a lack of flood control and slope change to accommodate runoff water. That would not have worked with Noah’s flood.

Although man was given the hook in the report the heavy rain was caused by normal weather events coming together for a flood.

Bill Taylor
October 20, 2021 8:10 am

this has reached the level of insanity…..the climate is NOT a force, it has NO power and has NEVER caused any weather event of any type anywhere.

October 20, 2021 11:18 pm

Reminds me of a Dilbert cartoon strip.

%d bloggers like this:
Verified by MonsterInsights