What is the reason for precipitation decrease in deforested areas of Amazonia ?

What is the reason for precipitation decrease in deforested areas of Amazonia ?

Emanuel Gloor (School of Geography), Dominick Spracklen (School of Environment), Carly Reddington (School of Environment), Ben Hur Marimon University of Mato Grosso, Brazil, Beatriz Schwantes Marimon, Isabelle Bonini, Bruno l’Advocaat

e.gloor@leeds.ac.uk, d.v.spracklen@leeds.ac.uk, bhmarimon@unemat.br

Amazon forests — by far the largest tropical forests of the world — are often referred to as the lungs of the planet because they cycle huge amounts of carbon and oxygen. They host an incredible diversity of life and are of global importance for atmospheric carbon levels and climate. One important feature of forests is that they pump large amounts of water back into the atmosphere all year round, in contrast to pasture and grassland. Up to date approximately 20% of the Amazon forests have been converted to pasture (cattle) and crop farming (soy bean and maize). Thus while deforestation in Brazil has steadily decreased compared to two decades ago recently it has risen again with fields and pasture further and further encroaching into primeval forests (Figure 1). An important driver is global markets as the Brazilian economy is relying substantially on export of agricultural products.

Deforestation is expected to lead to changes in climate, both in temperature and precipitation, and both at large and local scale. Understanding the underlying mechanisms in full is important in a warming world. Famous modelling studies suggested that change in climate caused by deforestation leads to savannah vegetation in eastern Amazonia (e.g. Shukla et al. 1990). It has also been suggested that transpiration from forest vegetation is important to trigger the wet season in Amazonia, and in the deforested belt in Southern Amazonia studies find an increase in dry season length. It has also been demonstrated that temperature in deforested areas have increased more rapidly than in forested areas (Baker and Spracklen, 2019). A recent study by Isabelle Bonini and Ben Hur Marimon has analysed precipitation trends over the past few decades shown in Figure 1 (please see below).

The analysis reveals interesting patterns. Firstly precipitation trends tend to be negative in deforested areas and positive in forested areas. A possible explanation is that air parcels crossing deforested areas have picked up less transpiration than air parcels travelling over forests. Thus there is a decrease in water vapour in these air-parcels and a lesser chance for precipitation. Another possible explanation may be that the local contrast between forest and pasture may lead to a local atmospheric circulation with air upwelling over forests pulling low level air from the pasture which which leads to preferential rain over forests instead of pasture. Yet another factor is the increase in vapour pressure deficit caused by particularly rapid warming in this region over the past decades which will increase vapour pressure deficit and thus potentially down-regulation of plant stomatal conductance. Disentangling the various possible mechanisms underlying the precipitation trend patterns is the aim of this PhD project.

Figure 1: Precipitation time trends (red negative, blue positive) in forested and deforested areas of Mato Grosso, Brazil.  Courtesy Isabelle Bonini and Ben Hur Marimon.


We envisage to use two tools to answer the question of the project. Firstly a trajectory based approach similar to Spracklen et al. 2012 may help understand the effect of large scale cumulative effect along air-mass trajectories. Secondly a high resolution regional climate model (WRF) will be used to examine local atmospheric circulations set up by the mosaic of deforested and forest areas typical for the Southern Amazonia deforestation belt. WRF is being used routinely for the Amazon by Carly Reddington and will be available for the project on high performance computers at University of Leeds. Relevant climate data are available from INMET (Brazilian Meteorological Service).


You will work under supervision of a strong team of experts of Amazonian Climate and Climate modelling. As part of the project you will learn how to run climate models on high performance computers, how to visualize and analyse the outputs, how to compare them with observations and how to draw conclusions on mechanisms.

Potential for high impact publications

The problem to be investigated is important for predictions of future Amazon climate. As such the project is clearly of wide scientific and societal interest. It is expected that the student will be able to publish papers during the PhD and ahead of thesis submission.

Student profile

Applicants should have a strong interest in global environmental problems, a strong background in a quantitative science (math, physics, engineering, environmental sciences) and a flair for, and good familiarity with, programming and scientific computing.


Shukla, J., Nobre, C. and Sellers, P. (1990) Amazon Deforestation and Climate Change, Science,  247, 1322-1325

Spracklen, D. V., Arnold, S. R. and Taylor, C. M. Observations of increased tropical rainfall preceded by air passage over forests, Nature, 489, 282–286 (2012)