FIRES IN THE PARANÁ RIVER DELTA

Dymaxion Labs
4 min readOct 3, 2022

--

Over the last months, fires in the Delta and islands of the Paraná River, in Argentina, have had major consequences due to their devastating effects on the health of populations who live close to the area and on the local flora and fauna. Which areas were affected and why is it important to detect burned areas? In this article, we answer those questions using satellite imagery, of course.

The Delta of the Paraná River covers 243126 ha. of Santa Fe and Entre Ríos provinces, and it is considered one of the largest in the world. Thanks to the characteristic microclimate of the Delta, in this area there are ponds, streams, swamps and wetlands which, together, make up a biodiversity reservoir and play a key role in climate regulation.

As a consequence of these fires, many islanders suffered considerable material losses, such as their housing, livestock and various elements which are vital for them to keep carrying out their economic activities. In particular, the agricultural sector was highly affected.

The consequences of these fires also had a direct and indirect impact on the flora and fauna of the area, since the habitat of many species, their food sources and the necessary conditions for their survival suffered substantive adverse effects.

The fires wiped out the land. Source: La Nación newspaper

The inhabitants who live close to the area were not the only ones affected; cities such as Rosario, in the province of Santa Fe, were covered in smoke, reaching air quality levels way above the levels considered acceptable for human health. These effects also had an impact on the health system, where medical visits, mostly due to respiratory conditions, increased by 15% to 25%. The smoke also reached more distant cities, such as Buenos Aires.

Measurement of the burned area in the Paraná River Delta

To understand the current situation and take action, it is essential to have a clear image of which areas were affected and to what degree. To achieve that, we used the tools Satproc and Unetseg to detect the areas were fires occurred, as well as nearby water banks.

According to our analysis, by August 29 and September 4, of a total of 933686.952 ha analyzed, pertaining to the area that comprises the Delta and its surroundings, 108116,8 ha had been burned, that is to say, almost 12%. On the map, three main focal points can be observed.

To train the detection model, we used an NBR image, which is the result of the estimation of the Normalized Burn Ratio. This ratio is very useful to detect the impact on large areas of land and it combines Near Infrared (NIR) and Short-Wave Infrared (SWIR) bands.

NIR images from the Paraná River Delta

In normal circumstances, the NIR value for healthy vegetation reflectance is very high and, in contrast, the SWIR value for said reflectance is very low. When a fire occurs, these values change, enabling not only the detection of burned areas, but also the assessment of impacts and recovery levels.

Although we can guess which areas were burned based on NBR images, the predictive model obtained with Satproc offers the advantage of enabling the detection of another type of objects. This time, we included the “water bank” class to, for instance, determine how close they are to the areas affected by fires, in order to use them to fight fires or assess a possible volume decrease.

Satelital image of the Paraná River Delta that includes burned areas and water banks

Declines in biodiversity, soil impoverishment and economic losses are just some of the severe consequences of fires in the Delta of the Paraná River.

In light of this environmental disaster, we have the necessary tools to monitor its development and contribute to the generation of precise measures to stop the fire and implement recovery policies.

If you would like to learn more about burned areas detection or if you have any questions, contact us through our social networks or through this channel.

--

--

Dymaxion Labs
Dymaxion Labs

Written by Dymaxion Labs

Creating value from geospatial imagery.

No responses yet