Study investigates delayed onset of rainy season in the Amazon

Observations suggest that the transition from dry to rainy season in the southern Amazon is occurring nearly a month later than in the 1970s.  Brazilian and U.S. researchers are investigating the possible influence of pollution from Manaus in this process   

By Karina Toledo, in Washington, DC

Agência FAPESP – The transition from dry to rainy season in the southern Amazon region usually occurs between the months of September and October, and any delay in this time frame has a significant impact on the local agriculture, power generation and the flow of the region’s great rivers, on which the population depends for getting around.

The severe droughts that affected the Amazon region during 2005 and 2010 as well as the floods of 2009 and 2014 indicate a growing variability in the onset of the rains, which climate prediction models do not have the sensitivity to detect.

Obtaining a better understanding of the factors that influence this transition in order to improve existing mathematical models is the goal of a FAPESP-funded project coordinated by researcher José Antônio Marengo Orsini from the National Center for Natural Disaster Monitoring and Alerts (CEMADEN), in partnership with scientist Rong Fu from the University of Texas.

“We have observed an increase of nearly one month in the duration of the dry season compared to data from the 1970s. Existing mathematical models indicate that this delayed onset in the rainy season will further increase. We want to investigate whether the pollution plume from the Manaus metropolitan area is having an influence on this process,” Marengo explained.

The research is being carried out under the scope of the Green Ocean Amazon (GOAmazon) initiative, which brings together researchers from several Brazilian and U.S. universities and research institutions, funded by the U.S. Department of Energy (DOE), FAPESP and the Amazonas Research Institute (FAPEAM), among other partners (read more at:

Preliminary results were presented Tuesday (10/28/14) in Washington, DC during the FAPESP-U.S. Collaborative Research on the Amazon symposium.

“Evidence from the literature suggests that the transition from the dry season to the rainy season is influenced by external factors such as temperature anomalies on the ocean surface and moisture transport. But this transition is without a doubt triggered within the forest,” Fu said.

The researchers are working with two different models: the American model known as the Community Earth System Model (CESM) and the Brazilian Earth System Model (BESM). But according to Marengo, the models are still unable to precisely represent the impact of the extent of the drought in the southern Amazon region.

There are parameters that need to be improved, such as the inclusion of aerosols and the representations of low clouds. The idea is to use the entire range of data generated by the various GOAmazon experiments to feed and refine the models,” explained Marengo.

According to the researcher, the southern Amazon region suffers most from the delayed onset of the rainy system inasmuch as there is no defined dry season in the north. Besides the impact on the populations, scientists worry that prolongation of the dry season could cause permanent damage the forest.

“Humans adapt, but the forest may begin to dry out and become more vulnerable to fires. When the rains do come, it will be too late. Only by refining the models can we be sure about possible impacts,” Marengo said.

Cloud modeling

Another project carried out under the scope of GOAmazon and designed to improve climate prediction models was presented at the Washington, DC symposium by Tercio Ambrizzi from the Institute of Astronomy, Geophysics and Atmospheric Sciences (IAG) at USP, and his colleague Carlos Roberto Mechoso from the University of California, Los Angeles (UCLA).

“Our objective is to investigate how the aerosols produced in the Manaus area influence the process of cloud formation in the Amazon region. We compared the simulations that the different models are able to make using real data produced at the various GOAmazon research sites,” Ambrizzi said.

Once refined, these models can be incorporated into programs that design climate change scenarios, thus increasing the degree of reliability of the projections, said the researcher.

In all, the group is working with five different mathematical models, including one for global climate predictions, one for regional predictions, and one designed specifically for cloud formation. There is also a program that is able to map the trajectory of the clouds, from their initial development, through maturation and dissipation in the form of rain, using satellite imaging.

Using what is known as the Lagrangian model of particle diffusion, Ambrizzi’s group is investigating the details of where the moisture in the Amazon comes from and where it goes. Initial findings were reported in an article published in the journal Hydrology and Earth System Sciences.

“You can clearly see from the trajectory of the particles that the tropical north Atlantic and tropical south Atlantic regions are sources of moisture for the Amazon. These particles travel to the Southeastern region of Brazil where they become rain,” Ambrizzi said.

Research sites

Since early 2014, a vast amount of data regarding the chemical composition of aerosols and atmospheric gases, the microphysics of clouds, and meteorological parameters are being collected at the various research sites installed in the Amazon for the GOAmazon project.

What is known as site T3, located in Manacapuru, 100 km from Manaus, houses the Atmospheric Radiation Measurement (ARM) Facility – a mobile set of terrestrial and air instruments belonging to the DOE, developed for climate studies. The site receives the Manaus pollution plume after it has traveled long distances and interacted with solar radiation and particles emitted by the forest.

The T2 is situated in the municipality of Iranduba, located along the banks of the Rio Negro across from the city of Manaus, and it receives the pollution plume as soon as it is emitted. Also installed there with support from FAPESP is a container with equipment similar to that found in Manacapuru.

The infrastructure for collecting data generated by the GOAmazon initiative also consists of two towers installed in the city of Manaus. One is located at the headquarters of the National Institute of Amazonian Research (INPA) and known as site T1, and the other is located at an array of towers north of Manaus, known as T0, which includes the 320 m high Amazonian Tall Tower Observatory (ATTO). The T0 is situated on the side opposite that affected by the plume and thus represents Amazon atmospheric conditions unaffected by pollution.

“We are analyzing the data from stations before and after the Manaus pollution plume. Our initial findings are that without knowing the situation of the atmospheric chemistry before the plume, at T0, it is nearly impossible to interpret data collected at T3, the site of the ARM Mobile Facility,” said Paulo Artaxo, professor at the USP Physics Institute (IF) and creator of the GOAmazon initiative along with Scot Martin from Harvard University.

Comparisons between data collected at the various sites, said Artaxo, reveal that the chemical composition of the aerosols and trace gases observed in Manacapuru are strongly influenced by the Manaus pollution plume.

“We still need to analyze what the impact and consequences of this will be. With regard to the ozone, we have already observed a four-fold increase in concentration when comparing T0 and T3. It goes from 10 parts per million (ppm) to 40 ppm after the plume, reaching levels that could be harmful to plants. We have also noted the strong effect on the balance of atmospheric radiation, altering the quality of the radiation available for plants to use in photosynthesis,” he said.