Fire in the Amazon

In early August 2019 reports of increase fire counts across the Brazilian Amazon started reaching social media, especially Facebook and Twitter. Photos of fire destruction started to swarm the internet, followed by numerous news outlets sounding off data and information. Sides were chosen: on one hand some insist the Amazon is going up in flames, while on the other hand others say there is an over-reaction of the impacts. Like any complex issue, many have tried to simplified the current fire activities occurring in Brazil and elsewhere to one problem – either ‘bad’ farmers, current political rhetoric or indifference to the benefits of the Amazon rainforests.

Here’s a brief summary of what has happened so far:

Some states in Brazil (especially in the north e.g. Rondonia, Amazonas and midwest e.g. Mato Grosso) have seen an increase in fire activities in July and August. It’s natural to see an increase in fire activity during this time as it’s the main clearing season for farmers, not only in Brazil but across South America.

Image from NASA showing fire detection from MODIS between 15-22 August 2019

It’s also important to note that fire can be used either destructively or productively. For instance, many local and indigenous groups have utilize fire to clear forest for planting or revitalize landscape by clearing overgrown fields or pastures to promote new plant growth for cattle to feed. Though the Amazon rainforest have not co-evolved with fire like other biomes such as savannahs or regions of the world e.g. Australia, fire has been utilized for decades by numerous groups. In other words, not all fires are bad.

Fire activities have also increased in Bolivia and Paraguay (as well as in Colombia, Venezuela and Guyana). When fire activities in Bolivia and Paraguay combined with the fire activities in northern and midwestern Brazil, it generated the huge plumage of smoke across southern eastern Amazonia, generating images like the one below. Some of this led to São Paulo experiencing a smoke ‘blackout’ in the middle of the day. This has happened previously in past years due to the amount of fires that were lighted especially in Mato Grosso, the main agriculture centre of Brazil where farmers still use fire as a tool to clear lands. This, of course, has wider regional effect on (a) health, (b) carbon emissions and (c) damage to wildlife and neighboring communities.

Smoke from recent fire activities cloaks the sky in São Paulo. Image is from Washington Post 2019.

So, is there an increase in fire activity in 2019? To answer this question, we have to see what data is being used. Besides first hand account from people on the ground, there are two main types of data that scientists tend to utilize to examine fire activity: (a) active fire data: the number of fires that are occurring with detection based off of infrared radiation (heat), and (b) burned areas: this data comes from analysis of satellite images, looking for the change in how the surface looks after the fire has burned. 

The majority of this current detection is based on active fire data as satellite images are collected daily. Burned area data is available up to July 2019 (this is a monthly available product). As such we know active fires have increase significantly, and most likely know that the extent of area lost to fire may increase but we cannot confirm the extent (area) as yet. 

MODIS Fire Counts as examined by NASA
Fire Burned Areas across Brazilian Amazon from 2002 to 2019 according to INPE, Brazilian Space Institute. The 2019 plot does not contain August data as yet.

So why the fuss? What is surprising with this year’s activities is (a) where fire activities are occurring and the (b) the number of occurrences reported via active fires within such a short time. In regions where historically forest loss have been minimal, such as Amazonas state, we’re seeing an increase in activity here (see Pervasive rise of small-scale deforestation in Amazonia). According to recent data from NASA, the Brazilian Amazon burned up to about as much as the past 20-year average for the same time period, but more than during the past 10 yrs when deforestation declined. 

So what has led to this uptick in active fires? Besides clearing farmlands, many of the current supporter of the Brazilian government have become emboldened to clear lands for pasture and crops. The other side of this coin is the fact that there is high unemployment in many states and accessing lands is seen as a way to address this issue. Illegal gold mining (especially on resource-rich indigenous lands) has also seen an increase where fire is used to help clear lands prior to mining. Additionally, the significant cuts in budgets to environmental agencies, such as IBAMA, who historically helped monitor and combat fire events is also playing a role.

Climate change is also shaping activities in Amazonia. When a satellite detects a fire, it doesn’t tell us where the fire is occurring i.e. whether on degraded pasture, overgrown croplands or in forests. In fact, current satellites do not detect most fires underneath the canopy of standing forests. So when severe droughts or El Niño events like that of 2015/2016 occur, they can create the needed conditions for fire spread as trees dies during that period. With most drought events, there may also be a ‘lag’ in seeing the impact from these events i.e. we’ll see an increase in tree mortality 1-2 years after severe drought or El Niño events. Dead trees often equate to more fuel. We also know that according to precipitation data (how wet/dry a season is compared to previous years), 2019 is not as dry compared (with major drought years being 2005 and 2010) to other years so the increase in fire activities may be more human-related. 

Lastly, it’s important that we are cautious in how data and facts are utilize. The dry season is only just beginning and we really don’t have supporting data as yet on the extent (area) of forest loss, as oppose to pasture/cropland clearing.

About the Author: Dr Michelle Kalamandeen is a Remote Sensing and Systematic Conservation Planning Scientist currently doing a postdoc at the University of Cambridge, UK and Laurentian University, Canada. She was part of the Guyanese delegation at the UNFCCC COP21 in Paris and have established several protected areas in Guyana.

Published by Dr Michelle Kalamandeen

Dr Michelle Kalamandeen is a Remote Sensing and Conservation Scientist from Guyana, South America. She aims to get a more accurate picture of what the forest and aquatic landscapes looks like today and in the future, how they will respond to different environmental and land use changes and how this will impact on climate, biodiversity and the people living there. Michelle has extensive NGO leadership experience including 12+ years serving as an executive board member of the Guyana Human Rights Association. She is currently one of 170 experts from the Amazon working on the Science Panel for the Amazon ( and is part of the core Editorial Board of the new section in MDPI Remote Sensing called 'Discoveries in Remote Sensing' (

Create your website at
Get started
%d bloggers like this: