Fire Forecasts July-September 2023

Author: Kátia Fernandes

Fires in the Amazon are the result of human activities that (usually) begin with slash and burn practices to clear the land for the introduction of agro-pastoral activities. For most part, fires occur over non-forest landscapes but can escape to standing forest edges and burn the understory.

Drivers of fire ignition during the dry season depend on factors such as land use practices, market demands, existence and enforcement of environmental laws and the prevalent climate. Assuming that the anthropogenic drivers are similar from year to year, unintended fire spread depends on whether the dry season is unusually dry or wet, and predicting how intense the fire season will be is essential for prevention planning.  Fire activity as a function of climate can be determined based on the 20 plus years of continued active fires measurements from satellite and climate variables available from observations and seasonal climate forecasts from General Circulation Models.

Drivers of fire ignition during the dry season depend on factors such as land use practices, market demands, existence and enforcement of environmental laws and the prevalent climate. Assuming that the anthropogenic drivers are similar from year to year, unintended fire spread depends on whether the dry season is unusually dry or wet, and predicting how intense the fire season will be is essential for prevention planning.  Fire activity as a function of climate can be determined based on the 20 plus years of continued active fires measurements from satellite and climate variables available from observations and seasonal climate forecasts from General Circulation Models.

The seasonal fire forecast presented below follows a statistical model ( Fernandes et al., 2011 ) that uses sea surface temperature (SST) forecast from early as March to predict July-September (JAS) fire season anomalies. The forecast is updated monthly through June and the closer the forecast is from the July-September fire season, the more robust is the outlook. Anomalously warm SSTs in the north tropical Atlantic (NTA) shift patterns of precipitation northward reducing moisture inflow into the Amazon and drought follows. Thus, it is expected that a warmer NTA will translate into a more active fire season whereas a cooler NTA will result in milder (below average) fire activity (Fig. 1).

To predict the NTA, we use climate models which reliably predict sea surface temperatures (thus NTA) with a lead of 1 to 9 months. Once the NTA forecast is assessed, it can be used to determine July-September fire anomalies in the southern and western Amazon. map represent areas with no fires or with poor relationship to the NTA index.

The JAS 2023 forecasted anomalies in the NTA are similar to those of 2010, when the southwestern Amazon experienced a severe fire season. The red shades in Fig. 2 indicate moderately to well above average fire activity being the predominant pattern of the 2023 forecast. Gaps in the map represent areas with no fires or with poor relationship to the NTA index.

The statistical  fire forecast model  by  Chen et al. (2011 ), shows similar results pointing to high risk of an active fire season as seen in Fig. 3.

It’s important to keep in mind that the climate determines only part of the variability of the fire season. In the absence of human activities, fires would be non-existent in the Amazon rainforest biome even in dry years, while a concentrated effort to burn will likely result in more fires than usual in years with normal precipitation. Nonetheless, fire prediction based on seasonal climate forecasts can provide an advanced look of the background climate conditions that may favor (droughts) or inhibit (rainy) widespread fires.

The current 2023 fire forecast is based on tropical Atlantic conditions that are very similar to those of JAS 2010. Precipitation patterns shown in Fig. 4b as Standardized Precipitation Index (SPI), describes moderate to severe drought in most of the Amazon in 2010, which resulted in a very active JAS 2010 fire season, especially in parts of Bolivia and in the states of Mato Grosso and Pará in Brazil (Fig. 4b).

The north tropical Atlantic index forecast in June 2021, indicated the potential for average, or slightly below average precipitation from July to September. The observed precipitation pattern (shown in Fig. 4b as the Standardized Precipitation Index - SPI)), describes drought conditions in southwestern Amazon (such as the Brazilian states of Acre and southern Amazonas), where anomalously active fire occurrence (Fig. 4a) was also observed. But the overall precipitation pattern in southern Amazon describes normal (marked as white in Fig. 4b) to above-normal precipitation (shaded in green in Fig. 4b), which helped tamper fire activity in northern Mato Grosso, Tocantins and southern Pará (shaded green in Fig. 4a).

The predicted above-normal fire season of JAS 2023 describes a similar outlook as that of JAS 2010, one of most severe fire seasons on record in the southern and western Amazon in the last 20 years. However, the climatic conditions leading up to the dry season are distinct. Moderate drought was well established in most of the Amazon by June of 2010 as shown by 3-month SPI (April-June), suggesting that by the beginning of the dry season (JAS) the region’s vegetation was already water stressed.

As of June 2023 (3-month SPI April-June), normal to above normal precipitation has been registered in a large part of the western Amazon except for localized drought pockets in northern Rondônia and eastern Amazonas in Brazil, indicating that the vegetation is less water stressed than in 2010 and less likely to burn, especially in the early stages of the dry season. Nonetheless, the climate forecast points to an anomalously dry season (JAS) in southern and western Amazon in 2023 and it’s likely that fire occurrence will pick up by the mid to late dry season.

Fires in the Amazon can be broadly classified into (i) deforestation fires; (ii) agricultural fires that intentionally or accidentally escape into the understory of neighboring forests, (iii) fires used for agricultural management and, (iv) savanna and grassland fires ( Andela et al., 2022 ). Based on that classification, a team of NASA scientists working with SERVIR-Amazonia developed a  geospatial tool  that monitors, daily, the location (including delineation of protected areas) and type of fires in the Amazon (Fig. 6).  This resource allows you to filter, download, and toggle through detected fires and their respective statistics. Daily outputs of active and new fires can provide information for management decision-making as the fire season progresses both within and outside of the Amazon biome. Later versions of this Amazon Fire Dashboard will be available in Spanish and Portuguese, and receive updates to the UI functionality and methods documentation.

•  Fire Forecasts 2021 
•  Fire Forecasts July-September 2022 

We are interested in hearing your comments and questions  here 

SERVIR-Amazonia Copyright ©2023 International Center for Tropical Agriculture -CIAT